Вы находитесь на странице: 1из 152

Chapter 1

INTRODUCTION

Background

Frank Biocca (1993), author of Communication in the Age of Virtual

Reality, has written that communication research is concerned with how humans

create techniques and technologies to turn each others thoughts' into each

others experiences (p.59). Biocca is referring to the ability of communication

technology to help translate thoughts into messages.

According to Biocca, communication research has taken the role of

spectator in respect to communication technology. In other words,

communication research has been content to cheer or boo new technologies

from the sidelines after their introduction. It is important to consider not only how

communication environments are socially constructed, but also how

communication researchers might engage in their construction. Social science

researchers have long considered the construction of social reality (Anderson,

Heath, Luff, & Moran, 1993; Gergen, 1986; Wertsch, 1991). Now, we have the

opportunity to become involved in understanding more about the social

construction of technology use.

As communication researchers, we should listen to Bioccas concerns.

We should become involved in the development as well as the criticism of

communication technology. In order to be able to design better communication

technologies, it is critically important that we understand more about how this

technology is actually used, including how it is learned.

Current computer interface designers are employing more user-centered

perspectives in their designs (Norman, 1986). Their goal is to produce human-

1
computer interfaces which blend into the world of the user in a seamless manner.

"The invisible computer" is a phrase that was coined at MIT's Media Lab to

capture the essence of this transparent interaction between user and computer

(Brand, 1987). Making computers which may be programmed and controlled by

the typical user is part of this goal.

One of the most promising developments along these lines is a new type

of computer programming technique known as Programming By

Demonstration. This is a method by which end users may program the

computer themselves. End users are people who use computers, but who are

not professional programmers and usually have not taken a programming

course. Briefly, it consists of demonstrating an action the user wishes the

computer to perform. The computer generates the software coding which

corresponds to that action. When the resulting program is run, the computer

performs the demonstrated action (Cypher, 1993).

Today, children in Americas classrooms are often exposed to computers

on the first day of kindergarten (Hoyles, Healy, & Pozzi, 1994). Schools are

endeavoring to provide as many computers in classrooms as budgets will allow.

However, in most classrooms the user-computer ratio is well below the one-to-

one correspondence experienced by many adults.

Therefore, by default, many school children tend to find themselves

sharing a computer with one or more peers. In doing so they must work

together or collaborate, when working on the computer. In this study,

collaboration is defined as a coordinated, synchronous activity that is the result

of a continued attempt to construct and maintain a shared conception of a

problem (Hoyles, Healy, & Pozzi, 1994, p.70).

2
Purpose of the Study

The dissertation research reported here explores how humans

communicate with each other when interacting with computers. This study

focuses on the interaction which takes place between two children who are

communicating with each other as they collaborate to program a personal

computer. In doing so, they will utilize a new type of software which allows the

user to program microworld simulations on the computer via demonstration.

Through content analysis, this research seeks a better understanding of the type

of communication which occurs between young learners as they interact with

each other and the computer. Therefore, the following research question was

asked: R1) What types of talk occur during the collaborative learning of

computer software?

The grounded theory approach of this study also lends itself to the

discovery of the function of collaborative communication in the three-way

interaction of children, the computer, and microworld simulation construction.

Thus, a second research question asked: R2) What functional role does

communication play during a collaborative microworld construction experience?

Collaborative activity is a well-documented feature of classroom

computer use (Fisher, 1993; Mercer, Phillips, & Somekh, 1991; Roschelle &

Teasley, 1995). In fact, communication activities during computer collaboration

are the essence of the collaboration. Monitoring this talk should be a good way

to understand more about the role of communication in the collaborative learning

of computer software.

3
Review of the Literature

A review of the computer assisted instruction (CAI) literature was

conducted to locate studies which focus on the human communication

component of collaborative computer interaction. A series of studies have

simultaneously considered both classroom communication and collaborative

learning with the computer. These studies are referred to as the Spoken

Language And Technology project or SLANT.

The SLANT project, conducted for two years from 1990 through 1991, was

based at The Open University and the University of East Anglia in the United

Kingdom. All research took place within Great Britains school system, but may

reasonably be applied to an investigation occurring within the American system

of public education. Its goal was to investigate collaborative learning at the

computer (Mercer, et al., 1991)

The following literature review is divided into three sections. The first part

considers the SLANT project and its findings. The second focuses on a

theoretical overview of collaborative learning at the computer. The third presents

Vygotskys Social-Cultural Learning Theory, a way of thinking about cognitive

development which is particularly well suited for the analysis of communication

during collaborative computer interaction (CCI).

The SLANT Project

The SLANT Project was begun as a part of an overall effort to improve the

quality of talk in England and Wales Classrooms (the National Oracy Project).

4
The SLANT project was aimed to develop knowledge and understanding that

can contribute directly to policy and practice (Mercer, et al., 1991, p.199).

Several studies have been generated from this research. The goals of the

SLANT project were as follows:

1) To contribute to knowledge about the development of childrens

exploratory and argumentative talk through classroom activities by:

a) identifying ways in which computer-based classroom

activities provide a context for exploratory and

argumentative talk;

b) describing activities which serve this function;

c) describing the range and quality of exploratory and

argumentative talk which arise through such activities;

d) providing information about the role of the teacher in

mediating and supporting such activities.

2) To make a contribution to educational policy and practice by

generating practical suggestions for how computers may be used

effectively to simulate exploratory talk and reasoned argument in the

classroom... (Mercer, Phillips, & Somekh, 1991, p. 195).

The SLANT project sought to investigate the possibilities for talk inherent

in computer mediated activity. Mercer, Phillips, and Somekh (1991) maintain that

computers can provide a vital, energetic presentation of multi-sensory

information. Computers capture the attention of children in a way that is not

approached in its effectiveness by other kinds of classroom collaborative

encounters. Children and teachers are viewed by the SLANT project as actively

co-constructing knowledge.

5
In the SLANT project, facts are seen as more than just objects in the

environment to be passively assimilated. Rather, facts are viewed as shifting

reflections of the world which are dependent upon a persons set of values.

Facts therefore become objects for discussion and judgment (Elliot, J., 1983,

cited in Mercer, et al., 1991, p. 198).

For example, on-screen cursor movement could be construed as one such

"fact". Cursor movement is a visual "fact" that is subject to "discussion and

judgement". Given this fact, one may believe that the cursor can be caused to

move only by moving the mouse. Or one may choose to believe that the cursor

is like any other on-screen artifact and is, ultimately, under the control of the

computer itself. The fact of an on-screen cursor movement does not clearly

support either belief. Negotiation of what is considered fact takes place during

collaboration. Therefore, collaboration creates an active crucible where real

world experiences are mixed with subjective impressions.

The SLANT project was also interested in how knowledge is negotiated

while it is being generated during a computer-based collaboration. The SLANT

project sought to distinguish ways in which a childs point of view is explained or

justified, how partners are persuaded into or out of certain courses of action or in

certain directions, and how consensus, when achieved, was accomplished

(Mercer, et al., 1991).

The SLANT researchers describe their analysis as the identification of

certain patterns and features of talk which seemed characteristic of childrens

conversations at the computer (Mercer, 1994 p. 26). No specific details were

given as to how specific patterns and features of talk were identified.

6
When looking at the physical conditions of collaborative computer

activities, one computer hardware specific variable was found important. In the

physical design of the hardware layout in most classrooms, students were forced

to sit side-by-side, giving control to the person who sat in front of the keyboard.

In two of the classrooms observed, children were encouraged to place the

keyboard on their knees, in the hope that a pair of children would pass the

keyboard between them (Mercer, 1994).

The researchers on the SLANT project describe software in terms of a

open-closed continuum, (Mercer, 1994, p. 28). The more closed a software

package the more highly structured its tasks and more pre-determined its range

of outcomes. The more open-ended the software, the more varied and wider

ranging the choices of interactions, and the greater possibility for exploratory talk.

It seems that in the case of open-ended software, such as word processing and

art programs, it is the children, not the computer, who take the initiating role in

the interaction. Close-ended software, on the other hand, like adventure games,

tend to position the child as respondent, reacting to action on the screen, rather

than having the chance to be proactive (Fisher, 1993).

A particularly useful subset of open-ended learning software environments

are commonly referred to as microworlds (Blaye, Light, & Rubtsov, 1992). This

type of software has been developed to invite exploration on the part of the

learners. Microworlds, otherwise known as simulations, require the learner to

first construct a representation of a given environment before being able to

perform an appropriate action.

Summary: The SLANT project took the initial steps towards understanding

the relationship between classroom computer activity and peer communication.

7
The SLANT project was concerned with teacher and learner talk during

collaborative computer interaction. This dissertation is concerned with the talk

that occurs between collaborators as they not only explore computer software,

but also the talk that occurs while they are programming the computer. It is

expected that software creation should present the entire gamut of discussion

topics, activities, and priorities which are unique to computer programming. If a

software interface is the current computer world one is inhabiting, the ability to

create that world is the ability to control every aspect of it.

Collaborative Computer Interaction and Learning Theory

Mercer, et al (1991) note that the more individualistic theories of cognitive

growth are increasingly being displaced by those featuring a constructivist or

more socially based model of cognition. An individualistic theory of learning has

been defined as one which is, predominately, if not exclusively, formed in terms

of individual actions and/or individual thoughts (Jones & Mercer, 1993, p. 19).

Both behaviorism and Piagetian approaches may be grouped under the

individualistic umbrella. Behaviorism is concerned with how individual behavior

is impacted by the environment. In the terms of behaviorist B. F. Skinners Law

of Operant Conditioning, appropriate behaviors are rewarded by the

environment, inappropriate behaviors extinguished (Skinner, 1938).

Piagetian theory is centered upon an individuals adaptation to lifes

complexity. Learning is termed as constructivist under Piagets approach, as the

emphasis is on the way a learner constructs his or her own understanding. As

opposed to operant conditioning, where the learner is shaped by the instructor

through selective reinforcement, the student here is promoted to the position of

active participant, who, through learning, is self-constructing his/her own lived

8
knowledge (Wadsworth, 1971). Piaget considered social interaction a source of

cognitive development whenever socio-cognitive conflict takes place. This

conflict occurs between childrens incompatible wrong answers, each partially

centered within a given childs social perspective. Piaget saw the social domain

as an arena in which to advance a de-centered solution that may be

internalized by each child as cognitive change (Wadsworth, 1971). In other

words, conflict leads to resolution via the negotiation of new view of the world.

Jones and Mercer (1993) maintain that the individualistic theories of

learning do not place adequate emphasis on the socialness of most learning.

Successful classrooms, they note, are characterized not as places of learning,

but rather as teaching and learning. Developing an understanding of computer

use in education requires that one theorize about and analyze teaching and

learning, not simply learning (p.20).

Teaching and learning is a communicative process whereby knowledge

is constructed, shared, debated, interpreted, and misinterpreted as teachers talk

with children, and children talk with each other (Mercer, et al., 1991, p. 19).

Teachers have increasingly recognized the need to consider classroom activities

which stress the orality of children. Children in classrooms are now being

encouraged to speak out more, justify their beliefs, evaluate information and form

and reform their opinions (Mercer, et al., 1991). The SLANT project discussed

earlier was created with this atmosphere in mind.

Beyond the peculiar circumstances of the [medical] examination room,

human problems are commonly solved through collaborative effort (Jones &

Mercer, 1993, p. 20). Computer oriented collaboration may take place on many

levels. In the present study, the emphasis is on the collaboration between two

9
co-acting users of the computer. In contrast, one may speak of a kind of meta-

collaboration which occurs between computer users and the designers of

computer hardware or software. Users of word processing software acquire the

skills necessary to accomplish their tasks by sharing the softwares designers

knowledge. They come to know the software, and thus come to possess

something of the creator of the softwares perspective on how words may be

manipulated on a computers monitor.

Summary: Theories of individualist learning, whether constructivist or more

classical, would seem to have little relevance when considering collaborative

computer interaction (CCI). If Mercers (1994) teaching and learning

supposition is the most appropriate for classroom learning, then another way of

thinking about cognitive development must be found. The late Russian

psychologist, L. S. Vygotsky, had a different philosophy, a position on child

development and learning which places communication and language at its core.

Vygotsys Social-Cultural Learning Theory

L. S. Vygotsky's ideas about a social-cultural approach to learning were

apparently at odds with the dominant Russian, individualistic learning theories

developed by Pavlov (Jones & Mercer, 1993). This was so much the case, that

Vygotskys work was banned by the Soviet state and did not significantly

resurface until the mid-1980s (Bruner, 1985; Newman, Griffin, & Cole, 1989;

Rieber & Carton, 1987; Vygotsky, 1978; Wertsch, 1985).

Vygotsky explained that the origin of the higher mental functions (i.e.,

voluntary action, voluntary memory, and rational, volitional, goal-directed

thought) were dependent upon two components (Rieber & Carton, 1987). First,

he proposed that the higher mental functions rely upon the mediation of

10
behavior by signs and sign systems, the most important of which is speech

(Rieber & Carton, p. 20). Second, Vygotsky argued that the higher forms of

human behavior have their roots in social interaction, in the individuals

participation in social behaviors (Rieber & Carton, p. 21).

Vygotsky emphasized the social aspects of cognition along two primary

axes. First, he focused on the central role that language plays in cognitive

development, problem solving and learning. Vygotsky theorized that when

children acquire language, it enables him or her to understand in new ways so

that a new cognitive instrument for making sense of the world becomes available.

Children are seen by Vygotsky to solve practical tasks with the help of their

speech, as well as their eyes and hands (Vygotsky, 1978, p.26). Second,

Vygotsky did not see the learner as the solitary being of the behaviorists or the

Piagetians, relegated to acting or reacting to some nameless environment. In

contrast, Vygotsky believed that Human learning presupposes a specific social

nature and a process by which children grow into the intellectual life of those

around them (Vygotsky, 1978, p. 89). He conceptualizes a childs learning

capability in a quintessential interactive fashion, via an incubator he called the

zone of proximal development or the ZPD.

In contrast to a childs level of actual development, the ZPD is a sphere of

cognitive activity which reaches beyond those acts which a child can perform

individually, to include those acts which a child can succeed at in collaboration

with, or under the guidance of a teacher or more capable peer (Rieber &

Carton, 1987, p. 211). Vygotskys research indicated that the zone of proximal

development has more significance for the dynamics of intellectual development

and for the success of instruction than does the actual level of development

11
(Rieber & Carton, 1987, p. 209). Vygotsky warns that, the teacher must orient

his [their] work not on yesterdays development in the child but on tomorrows

(Rieber & Carton, 1987, p. 211). The ZPD has growth, not conflict, as its central

concern. It acknowledges both the part others play in an individuals

development and the fundamentally social nature of cognitive change (Newman,

et al., 1989).

Vygotskys noteworthy declaration of learning as language dependent

could be said to have a certain obvious quality, yet its affirmation has profound

implications for both the role of the computer in the learning process and the role

of the teacher (or others) in the classroom during collaborative computer

interaction (Jones & Mercer, 1993). This communicative approach places less

emphasis on the relationship an individual learner has with the computer and

more on the computer as a communication medium through which a teacher

and a learner can communicate ( Jones & Mercer, 1993, p. 22.). In the present

study, the children themselves will help serve as each others teacher.

The idea of the computer as communication medium is a popular one in

human-computer interaction research. Computer systems are performing

messages, messages which themselves play the role of sender and receiver of

other messages (Laurel, 1990). Nicholas Nicroponte, director of MITs Media

Lab, has called computers the meta-medium (Brand, 1987). Computers are all

communication media rolled into one, or more succinctly, computers may be

viewed as metacommunication artifacts (deSouza, 1993, p. 756).

Vygotskys zone of proximal development does seem to be an interesting

way of theoretically describing the space in which collaborative computer

interaction takes place. It is apparent at first, as with many theories,

12
that the ZPD may be just another nice idea which might be very difficult to test in

the real world. However, one research team, although they never overtly say so,

has begun work on an experimental design which seems to capture the breadth

of the ZPD. Jeremy Roschelle and Stephanie D. Teasley (1985) call their

invention the Joint Problem Space.

Stephanie D. Teasley and Jeremy Roschelle (1993) have initiated a line of

collaborative problem solving inquiry which prefaces the notion that a shared

conception of a problem to be solved is central to collaborative problem solving.

They submit that social interactions in the context of problem solving activity

occur in relationship to a Joint Problem Space, (Roschelle & Teasley, 1995,

p.70). The JPS is a shared knowledge structure that supports interactive

problem solving by combining goals, descriptions of the current problem state, an

awareness of available problem solving actions, and associations which connect

goals, current problem state, and available actions.

Roschelle and Teasley (1995) maintain that the fundamental activity in

collaborative problem solving occurs via a socially negotiated, emergent set of

knowledge elements. This shared conceptual JPS is constructed via a meta-

framework which includes the sharing of language, situation, and activity. For a

joint problem space to be possible, collaborators must share the moment.

According to Teasley and Roschelle (1993), to build a JPS, collaborators

must be able to have ways of:

1. Introducing and accepting knowledge into the JPS.

2. Monitoring ongoing activity for evidence of divergences in meaning.

3. Repairing divergences that impede the progress of the collaboration.


(p. 236)

13
Additionally, a microanalytic methodology has been put forward by which

a joint problem space may be explored (Teasley & Roschelle, 1993). This

microanalysis combines pragmatics, conversation analysis, and protocol

analysis to investigate the JPS.Their analysis attempts to show how students use

the structure of conversation to continually build, monitor, and repair a JPS"

(Teasley & Roschelle, 1993, p. 236). Teasley and Roschelle (1993) discuss

several categories of discourse events which they have found useful in their

analysis. These consist of Turn-Taking, Socially-Distributed Productions,

Repairs, and Narrations. Turn-Taking was found to be the most pervasive and

general category (Teasley & Roschelle, 1993, p. 236). They looked for dialog in

which turn transitions are smooth, and the sequence of talk follows a

cooperative pattern (p. 237). Smooth transitions were those which followed,

from each participants point of view, a predictable pattern. It was realized that

cooperative problem solving includes interludes during which the partners are not

fully engaged with each other. Teasley and Roschelle (1993) suggest that in a

successful collaboration, such periods of withdrawal are usually followed by

periods of intense interaction," which allow for the mixture of personal

perspectives with the jointly-held problem-solving knowledge (p. 237).

Socially-Distributed Productions (SDPs), otherwise known as

collaborative completions are another category of turn-taking structure (Teasley

& Roschelle, 1993, p. 237). The IF-THEN form proved to be a especially

relevant type of collaborative completion. The distribution of the IF-THEN

across turns provides an opportunity for partners to accept or repair conditional

knowledge (Teasley & Roschelle, 1993, p. 237). One partner provides a

14
condition to be evaluated such as, What happens if I do this? and the other

partner has a chance to respond with an appropriate reaction. The socially-

distributed production may be a particularly powerful means of constructing

shared knowledge because it spreads interrelated goals, features, and actions

of a knowledge element across conversational turns (Teasley & Roschelle,

1993, p. 237).

Repairs or corrections made on behalf of past speech are the method by

which participants in talk can deal with problems or troubles in speaking,

hearing, or comprehension of dialog (Teasley & Roschelle, 1993, p. 238). As

Schegloff (1991) noted, repairs are a major means for the achievement and

consolidation of understanding and thereby the management of the mutual

intelligibility of the collaborative problem solving activity.

Narrations inform ones partner of the intentions corresponding to actions

(Teasley & Roschelle, 1993, p. 238). Such announcements afford the partner

the opportunity to highlight differences in mutual understanding.

Teasley and Roschelles technique of coding patterns of conversational

sequences, while an interesting approach to Vygotskys zone of proximal

development, seeks an understanding of the ZPD which is based more on the

forms, rather than the functions of collaborative discourse.

This dissertation study is more interested in investigating the function of

collaborative talk, not the process of its production. In other words, how

collaborative talk is generated is not the concern here, but rather, here the

concern is what purposes does collaborative talk serve its speakers.

15
Summary of Review of the Literature

The function of collaborative talk generated within a learning environment

is the central focus of this research. This dissertation study seeks to create a

"Zone of Proximal Development" in which students may learn from each other as

they explore the creation of computer microworlds using new self-programming

computer software (i.e., KidSim). The SLANT group's work on how knowledge is

negotiated during a computer-based collaboration, in particular, will serve as the

primary foundation to guide the present research (Mercer, et al., 1991).

16
Chapter 2

METHODOLOGY

This study is a discourse analysis of the recorded talk produced by three

pairs1 of 5th grade students as they built simulated environments using the

computer simulation program: KidSim.

The Subjects

The participants were six fifth grade public school students, collaborating

together in pairs. These students, pairs of best friends, were selected by their

teachers as good candidates for the study. The teachers were told to look for

students who could tolerate frustration, work well together with others, and enjoy

working with computers. These characteristics were singled out to help screen

for students who would be expected to concentrate on the task at hand."

Research time is valuable, and with a study such as this one, access was limited.

Exact subject pairings were dependent upon parental approval, but the

three pairs of children were best friends, one pair of boys and two pairs of girls.

Mixed sex best friend pairs were difficult to find in an elementary school setting.

Friendships have been shown to occur between persons that are

cognitively similar (Infante, Rancer, & Womack, 1990). Extending this argument,

best friends may be thought of as friends who possess the potential to be more

cognitively similar to each other than they are to other persons in their lives. The

best friend dyad, by its very nature, is thought to increase collaboration more

than might be otherwise possible.

17
The Setting

The site of this study was a public grade school in a small Midwestern city.

This site was chosen to provide a familiar location for the study. The choice of

location helped to encourage the creation of a learning context which supports

the exploration of Vygotskys "Zone of Proximal Development (Fisher, 1993,

p.112). Here, new software may have a reasonable possibility for assimilation,

while limiting any threatening environmental characteristics. For example, this

location eliminates both the threats of interacting in unfamiliar surroundings and

the tendency to have that under the microscope feeling which might occur in a

more typical laboratory setting.

The interaction took place in one of two rooms located at the school. The

first four interaction sessions for each pair took place in a small private office.

The final session took place at a small table along one wall of the schools library.

Only the students and the researcher were present at these sessions. The

student collaborators were seated in front of the computer with the computers

keyboard placed between them and the mouse positioned so that it was easily

reached by either student.

Interaction Sessions

Each student pair participated for five total interaction sessions. Two,

unrecorded training sessions of one hour in length served as training for students

on KidSim. These training sessions were conducted using KidSim's instruction

manual, "Getting Started with KidSim'" by Apple Computer (1995). This manual

includes a step by step tutorial on building a simulation of an aquarium. Each

student pair was guided through this tutorial. Beginning with the third session,

each student pair worked together to jointly build one microworld over the next

18
three meetings. These work sessions each lasted 50 minutes, and the remaining

10 minutes was used to interview each student separately, 5 minutes per child.

Both the interaction sessions and the interviews were video recorded. The video

camera was placed above and to the right or left of the researcher/observer who

was sitting approximately six feet in front of the participants.

Microworlds, as mentioned in Chapter 1, are the simulated environments

one builds with KidSim [See Appendix A for screenshots (i.e., pictures of the

monitor output) of an example microworld.]. Personal observations by this author

involving students the same age as those in this study, indicated that 3 one-hour

periods are sufficient to completely build and run a rudimentary microworld.

Data Collection

This study uses a grounded theory approach for data collection.

In particular, Glaser and Strauss's (1967) grounded theory technique will

be utilized via discourse analysis to discern general patterns and themes which

may exist within a recorded series of talk exchanges. Grounded theory, or the

"discovery of theory from data" is a systematic method of data analysis which

attempts to build "conceptual categories of data" using a technique known as the

"constant comparative method". The defining rule of the constant comparative

method is: "while coding an incident for a category, compare it with the previous

incidents in the same and different groups coded in the same category" (Glaser &

Strauss, 1967, p.106). The constant comparative method is an iterative process

in which each stage is subject to re-visitation and revision. Glaser and Strauss's

(1967) method consists of four stages:

19
1) Comparing incidents applicable to each category- Here, "each incident

is coded into as many categories of analysis as possible, as categories emerge

or as data emerge that fit an existing category" (p.105).

2) Integrating categories and their properties- As coding continues, "the

constant comparative units change from comparison of incident with incident to

comparison of incident with the properties of the category that resulted from initial

comparisons of incidents" (p.108). In other words, incidents are first compared to

each other to discern initial categories. Next incidents are then compared

against the categories themselves, to see if they fit an existing category or

whether a new one should be created. Memos should be made along the way to

record ideas, particularly when "conflicts in the emphasis of the analysts thinking"

occurs (p.107).

3) Delimiting the theory"- Delimiting occurs on two levels: the theory and

the categories" (p.110). As the theory takes shape, there is a sense that major

modifications become fewer and fewer as each new incident is analyzed.

Interrelated categories are reduced via an examination of "underlying uniformities

in the original set of categories" to formulate a theory with a "smaller set of higher

level concepts" (p.110).

4) Writing Theory- At this stage the analyst possesses coded data, a

series of memos, and a theory. "The discussions in the memos provide the

content behind the categories, which become the major themes of the theory

later presented in papers or books" (p.108).

A transcription of an audio recording of the interaction sessions was

made. Content analysis, particularly the analysis of "referential units", was used

to confirm the reliability of the categories of incidents generated by the constant

20
comparative method. Krippendorff (1980) notes that "referential units" are

identified "by particular objects, events, persons, acts,to which an expression

refers" (p. 61). In this study, the referential units of interest are talk sequences

which revolve around a single computer activity. Multiple functions of the talk

were coded where appropriate. Classification of talk are linked to on-screen

activity in order to show what it is referring to via a picture of that activity.

All talk was coded in three possible ways. First, as a Talk Activity

category similar to the conversational sequences that the SLANT project found

in their study; Second, as a type of human-computer interaction or Computer

Activity; Third, as a Functional Activity or a special sequence of human-

computer-human interaction generated by the unique nature of the collaborative

use of microworld construction software. The transcriptions were analyzed by

the researcher using the qualitative analysis program known as Q.S.R. NUDIST

(Non-numerical Unstructured Data Investigation Searching and Theorizing), to

help isolate patterns within the recorded talk.

The SLANT data identifies three kinds of conversational sequences

which highlight educational goals:

1) During Disputational talk, speakers challenge or disagree with the other

speakers views but do not attempt to justify their challenge by building on

previous utterances or offering new information (Mercer, 1994). An example of

Disputational talk would be the statement: Thats stupid," in response to

another's action or comment without giving a reason for the challenge.

2) During Cumulative talk, speakers contribute to the discussion by taking

up and continuing a previous speakers utterance, without explicit comment

21
(Mercer, 1994). Cumulative talk is illustrated by the phrase: OK, now see what

this does," as a response to another 's action or comment.

3) During Exploratory talk, hypotheses are proposed, objections are made

and justified, and new relevant information is offered (Mercer, 1994). Exploratory

talk is illustrated by the phrase: Thats weird, does it do that only when you click

there? as a response to another 's action or comment.

SLANT researchers propose that education should develop and

strengthen a childs ability to use talk as a facility for reasoning and become a

willing receiver of the reasoned arguments of others while making decisions or

drawing conclusions (Mercer, 1994). It is their belief that collaborative computer

activities which create abundant amounts of exploratory talk help to further this

aim.

During each interaction session, the researcher took careful notes that

served as a guide to events to be examined during the probe interview. A five

minute post-hoc probe interview was conducted privately with each participant at

the end of each day's interaction. Specifically, the researcher noted any

problems that created obvious conflict between the subjects, any situations

where one of the subjects seems to disengage from the task, as well as any

other aberrant or unusual behavior of the subjects. The researcher also noted

times of euphoric response to their activities to help later clarify and elaborate on

what the subjects were thinking at the time.

The following probe questions are typical of those used in the interviews:

P1. What did you mean when you said _____________________?

S1. Please tell me why.

22
P2. Why did you stop [drawing that character? ] )

(Why did you [get mad at __________ ?] )

(Why did you [say _________ ?] )

[Probe questions one & two will be repeated as often as necessary.]

P3. What did you do today that was the most fun?

S3. Have you been able to do [this] on other days?

P4. Did anything go wrong today?

S4. What could be done to make it better?

P5. What didn't you get to do today that you want to do next time?

Five interaction sessions were scheduled during a two week period for

each pair of subjects. Data collection took place during normal class hours via

special arrangements with each students teacher. A camcorder was set to shoot

a 2-shot of the participants as they faced the computer. A second VCR was set

to tape the computer output via a special computer/video interface box2 . A

microcassette audio recorder simultaneously recorded the same audio as the

VCRs. The researcher was present at all times, within sight of both participants.

The researcher was available to answer questions at all times and sought

to maintain a comfortable atmosphere by being as natural and undemanding as

possible. It is common, in this type of interaction, to include a facilitator (usually

the teacher) to help the child collaborators progress while learning the program

(Clements, Nastasi, & Swaminathan, 1993). However, once the initial training

period is over, the facilitator is encouraged to keep interjections at minimum, only

assisting when the situation merits. Researcher/student talk was coded only

when it related directly to the computer activity at hand.

23
The Hardware

As this investigation is concerned with human-computer interaction, choice

of computer was considered a critical part of the final outcome. A Macintosh IIsi

was chosen both for convenience sake (it is owned by the author), and due to its

WYSIWYG (What You See Is What You Get) GUI (Graphical User Interface).

This particular Macintosh was 5 years old. It employs a Motorola 68030 CPU

which was enhanced to run at 25Mhz. The Macintosh GUI, world famous for its

ease of use, allows for the quick launching of software in an easy to learn

environment (Williams & Hartley, 1990). All of the computer's control functions

may be accessed via mouse movements, eliminating the need to memorize any

of the command-line codes demanded by some other systems. Simple software

adjustments are only a "point-and-click-away" on the button and menu driven

Macintosh "Desktop" (Williams & Hartley, 1990). A GUI environment is a

necessary first step for many, if not most, "Programming by Demonstration"

software designs (Cypher, 1993).

The Software

Over 100 million people today can now use personal computers (Smith,

Cypher, & Spoher, 1994). Children as young as two years old can use a mouse

and a paint program. Yet very few people know computer programming. Until

recently, most of the power associated with computer programming has been in

the hands of the professional programmer. KidSim, or Kids Simulations, the

software chosen for this research, consists of an environment that allows children

to create their own simulations via end-user programming or programming by

demonstration (Cypher & Smith, 1995). They may build their own characters and

24
create rules that specify how the characters are to behave and act [Please see

Appendix A for an extended discussion of KidSim.]. All of this is accomplished

without a conventional programming or scripting language (Cypher & Smith,

1995, p. 27).

A simulation program was chosen over many other possibilities (e.g.,

games or educational applications) as it provides an open-ended quality, with no

particular goal prominent in its execution. Open-endedness sets few preset limits

upon the user and encourages maximum spontaneity. Both of these qualities

serve to lessen interaction lulls and offset early boredom.

Children working in groups with computers undergo a complex set of

conditions which may affect their collaboration (Hoyles, et al., 1994). One of

these conditions is the software selected for collaborative interaction (Fisher,

1993; Harel, 1990; Reilly, 1992). Fisher (1993) has noted that programming

computers as a collaborative activity may be the most beneficial use of software

for childrens overall learning. Computer programming involves a complex set of

cognitive resources on all levels (Papert, 1980).

Programming environments lie at the extreme end of the open-closed

software interaction spectrum (Light & Mevarech, 1992). Learners who act

simultaneously as programmers, have additional control over the machine. The

student collaborators tell the computer what to do and how to do it. KidSim is

the first simulation software which is user-programmable via simple

demonstrations. KidSim seems to be an ideal candidate in the quest for open

ended, exploratory talk generating software.

25
Summary of the Methodology

Glaser and Strauss' (1967) grounded theory technique will be employed to

analyze discourse generated by pairs of fifth grade best friends as they learn the

"programming by demonstration" microworld construction environment known as

KidSim. The KidSim software will be "running" on a Macintosh computer while

the students are audio and video recorded working with the software. Glaser and

Strauss' "constant comparative method" will be used to construct categories of

talk and computer activities. These categories will then be reviewed via further

analysis to create focused descriptions of the collaborative function of this

discourse.

26
Chapter 3

RESULTS

Introduction

Over a two-week period, three pairs of fifth grade best-friends spent five

hours becoming acquainted with the KidSim computer simulation program. This

section presents the analysis of their conversations as they collaborated to build

their own microworlds". Extensive excerpts of the children's talk will illustrate in

what ways their dialog enabled their collaboration.

Communication between the collaborators served to support three distinct

kinds of interaction activities: The first two of these, Talk and Computer Activities,

seek to answer research question number one: R1) What types of talk occur

during the collaborative learning of computer software? The third category,

Functional Activities, attempts to answer research question number two:

R2) What functional role does communication play during a collaborative

microworld construction experience?

Talk Activities are defined as human to human interactions which were

captured in talk sequences. Most varieties of Talk Activities are found throughout

all Computer Activities. Computer Activities are defined as human-computer

interactions within the microworld construction environment (i.e., KidSim running

on the Apple Macintosh computer). Finally, Functional Activities focus on what is

being talked about rather than forms of talk that are produced or the current

microworld construction step. Therefore, Functional Activities may be thought of

as "meta activities" or activities which occurred a level of abstraction above those

established by the first two. Each of these three varieties of collaborative activity

will be discussed within its own subsection.

27
Six varieties of Talk Activities were identified using grounded theory as the

investigative methodology. These were: Help Talk, Cumulative Talk, Query Talk,

Disputational Talk, Solo Talk, and Exploratory Talk. (Talk during the post-hoc

interviews was not a part of this coding.) These Talk Activities functioned as the

base level of communication which occurred between the collaborators.

Beginning with a most familiar form of communication, their speech, the

collaborators maintained a "bridge" to each other in order to jointly construct their

KidSim microworld. (An overview of KidSims features, layout, modules, and

design is found in Appendix A.)

Sixteen types of Computer Activities were examined using content

analysis. The four most prevalent of these, Creating a New Piece, Rule Writing,

Modifying a Piece's Appearance and Running the Clock Forward are presented

in detail during the Computer Activity discussion. The Computer Activities

functioned as the "construction bricks" which these KidSim collaborators used to

build a working microworld.

Three special Functional Activities were found to occur during KidSim

microworld construction. Role Sharing, Applause, and Object Negotiation were

found to be special human-computer-human interactions that were encouraged

by the act of collaborative microworld creation. One of these, Object Negotiation,

may be thought of as the most unique finding of this research. During Object

Negotiation, the KidSim collaborators created a portion of a microworld via a

"synthesis" of their own world views. Such a "synthesis" is an activity which is

particularly relevant to the creation of microworlds.

28
A total of 4,475 conversational turns were coded for type of Talk and

Computer Activity. A turn was defined as one unit of non-interrupted speech.

Speaker turns which follow themselves in the transcription indicate at least five

seconds has passed since the previous utterance from the same person. Each

unit was given a unique number within an interaction pairs transcription. As far

as possible, complete thoughts were coded together, so that antecedents and

responses about the same topic were coded into a single computer activity

category. All names have been changed to protect the privacy of the

participants.

Talk Activities

Table 1 provides a list of the six varieties of Talk Activities, their

percentage of occurrence, and the typographic coding key as they appear in all

transcription excerpts. A second coder reviewed twenty percent of the total

4,475 conversational turns, yielding an intercoder reliability of .803.

Table 1
Percentage of Talk Units for All Talk per Talk Activity (N=4475)

TALK ACTIVITY TYPE % per n


w/ subcategories subcategory

Help Talk (Researcher Initiated) 20.0% 895


(Student Initiated) 21.8% 976
Cumulative Talk 30.2% 1351
Query Talk 18.9% 848
Solo Talk (Engaged) 3.7% 165
(Disengaged) 0.8% 34
Disputational Talk 3.3% 146
Exploratory Talk 1.3% 60
Total 100% 4475

29
Help Talk is signified by blue type. Help Talk is defined as any talk

spoken by or dialog with the researcher (e.g., JD). Help Talk consists of two

facets: Student and Researcher Initiated. Often, while interacting with KidSim,

these young collaborators would become "stuck" and need help from JD to

continue. Help came both in the form of answers to collaborator questions and

volunteered suggestions from the researcher. In this way, the collaborative

environment was more similar to the give-and-take which occurs during typical

classroom teaching as opposed to a "laboratory experiment" based approach.

Help Talk occurred most frequently during this research (41.8% of all talk

units as shown in Table 1). Help Talk may be initiated by either the researcher or

one of the participants. Student Initiated Help Talk is defined as all questions to

the researcher. Student Initiated Help Talk is illustrated by the following

sequence:

J: (to JD) All of the elephants go right or just that one?


JD: All of them, it's good for all of them.
J: Oh, okay.

Researcher Initiated Help Talk is defined as any unsolicited comment made by

the researcher. Researcher Initiated Help Talk is illustrated by lines 1034

through 1036 of the following sequence:

1033 M: Now there's still giraffes in the way. But I think I


see a problem. Because, even now, because this
rule might only work if there's two giraffes in the way,
then one.
1034 JD: Remember every, according to KidSim,
everything that's inside the window is important. So
you need to decide what to do with that.
1035 A: Take that little thing and put it right there over
him. That will ignore it, right?
1036 JD: That's pretty clever. Now remember "I Don't
Care Squares" work on the left side of the rule. It's
saying if there's something here you don't care about.

30
Questions addressed directly to JD are so indicated (i.e., to JD). JD is always

coded as Help Talk and never in any other talk category. Help Talk may occur

singly or as a sequence of turns.

Help Talk is the primary collaborative activity between the students and

the researcher/facilitator. Researcher Initiated Help Talk occurred primarily as a

form of supplemental training or as a reminder of KidSim details which had

already been discussed. The researcher/facilitator often chose to interject at

moments in which a child's frustration toward the software was building. Student

Initiated Help Talk verified the progress of student learning and confirmed the

recently acquired attributes of the software.

Cumulative Talk is represented by plain, unmodified, type. In Cumulative

Talk speakers contribute to the discussion by building upon a previous speakers

utterance, sometimes giving a suggestion, without elaboration. Cumulative Talk

serves as a "presence check" in which each collaborator is letting the other one

know that they are "paying attention" and are "tuned in to" the current activity.

Cumulative Talk may be overtly positive or neutral in tone. Cumulative Talk does

not include questions and answers (see Query Talk).

Cumulative Talk accounted for 30.2% of all talk sequences (see Table 1).

Cumulative Talk is illustrated by the following sequence:

A: So, okay.
J: Those are nice ears.
A: Kind of looks funky, I don't know...

It always occurs in a sequence of two or more turns. Cumulative Talk functions

as the collaborative activity in which each speaker indicates to the other that she

is "mentally present" and aware of their exchanges. The momentum of the

collaboration is maintained and supported through a dialog of Cumulative Talk.

31
Query Talk is shown within the excerpts as italic type. Query Talk is

defined as cumulative talk which includes questions and answers; it is an overt

request for information. Query Talk serves as an explicit indication that the

speaker is concerned about some aspect of the Computer Activity.

Query Talk occurred in 18.9% of all talk sequences (see Table 1). Query

Talk does not include self-answered questions (e.g., "What? Oh, now I see.") or

questions during Disputational Talk, but the antecedent statement may be

included to clarify the impetus of a question. Query Talk is illustrated by the

following sequence:

J: I like the head. I like the ears.


A: Are you just saying that?
J: No, I like it. Looks good.

Query Talk may occur in a sequence of two or more turns. Query Talk functions

as the collaborative activity which indicates that more information is needed

before the current operation may proceed. Query Talk announces that learning

is stalled; it is a yellow flag which says: " Information or Confirmation Needed."

Solo Talk is noted as underlined type. Solo Talk is defined as un-replied

statements or statements which may or may not require a response, to which

none is given. Solo Talk may also take the form of someone thinking out loud.

Solo Talk was found to occur in 4.5% of all talk (see Table 1). Solo Talk

can be either engaged or disengaged. Engaged Solo Talk is Solo Talk about the

current computer activity. Engaged Solo Talk is illustrated by the first line of the

following sequence:

J: Then it won't connect to the nose.


J: Brown hair. Brown eyes. I suppose you're making your
girl first?
A: Uhm hmm.

32
Disengaged Solo Talk is Solo Talk which is "off topic" or not pertinent to

the current computer activity. Disengaged Solo Talk is illustrated by the first line

of the following sequence:

319 M: Hit the lollipop that's flyin' in the sun (mumbles).


320 C: Don't we need the.. empty space box?
321 JD: Yeah, yeah. The... "I don't care square".

This cycle of engagement/disengagement confirms the findings of Teasley and

Roschelle (1993) reported in Chapter 1. Cooperative problem solving settings do

seem to include interludes during which the partners are not fully engaged with

each other. Solo Talk always occurs singly or as a succession of turns from one

speaker. Solo Talk functions as the collaborative activity which indicates that a

collaborator is "going it alone" for the moment and may serve as a subtle plea for

more interaction from the other partner.

Disputational Talk appears as red type. Disputational Talk, is talk in

which speakers challenge or disagree with the other speakers views, but do not

attempt to justify their challenge by building on previous utterances or offering

new information. Disputational Talk is an angry, direct form of disagreement

which serves to tell the other that: "I don't like the course of action you are

currently taking." Disputational Talk is overtly negative in tone. Disputational

Talk is coded to include the antecedent unit. Disputational Talk, occurring in

3.3% of the interactions, is illustrated by the following sequence:

J: That look better?


A: Not really.

The following phrases are also examples of Disputational Talk: That's

stupid," or I don't like this." Such statements are made in response to another 's

33
action or comment without giving a reason for the challenge. Disputational Talk

may occur in a sequence of one or more turns.

Disputational Talk functions as the collaborative activity which most clearly

says: "I don't agree with you." The fact that immediate justification for the

disagreement is missing during Disputational Talk may indicate that the source of

the objection may be more emotional than logical. A collaborator may use

Disputational Talk to put into words the fact of the disagreement without

immediately being able to articulate why they disagree.

Finally, Exploratory Talk is printed in green type. Exploratory Talk may

include hypothesis proposals, spoken objections with justifications, new

perspectives or information. During Exploratory Talk, students are learning from

one another and testing their ideas, and the communication is often positive in

tone. Exploratory Talk serves as a clear indication of "learning-in-progress" and

gives the collaborators a way to demonstrate "what-they-already-know" to

themselves and to each other.

Exploratory Talk was found to occur in 1.3% of the total talk sequences

(see Table 1). Exploratory Talk may occur in a sequence of one or more turns,

alone or during other types of talk. Exploratory Talk is illustrated by the following

excerpts from each pair of collaborators:

Anne: No, 4. No. Put 8, because, ahmm, we want to do


them about the same size as the animals, don't we?

Marie: Thats weird, does it do that only when the clock is


running?

Jenny: We have to push this first, that's the way it works on


my computer.

34
Exploratory Talk functions as the collaborative activity which signals that one

partner is willing to share their knowledge with the other. When Anne is speaking

of "No. 4 or 8" she is referring to the magnification size of a character in the paint

module. By calling into question the relative size of the current character, Anne

is adding the perspective of size comparison to the piece creation process.

Marie mentions the game clock, to focus her partner's attention on the way a

piece's appearance changes as the game is running. Jenny brings new

knowledge to the collaboration by noting how "things work" on her computer.

Each of these examples illustrate the new knowledge or perspective found in

Exploratory Talk.

Summary of Talk Activities

The six types of Talk Activities delineate the collaborative communication

processes which form the foundation for both improved interpersonal and

human-computer interaction. Help Talk, Cumulative Talk, Query Talk, Solo Talk,

Disputational Talk, and Exploratory Talk each have a specific function during the

learning of KidSim. Coding and analysis of these six Talk Activities allows the

researcher to scrutinize how communication influences and supports the learning

process. The division of talk into these six types provides a reference point

which connects the peer collaboration to the Computer Activities. Collaborative

talk contextualizes each particular Computer Activity.

Computer Activities

Computer Activities were the second type of collaborative activity isolated

via grounded theory. Computer Activities form the core of the human-computer

interactions isolated by this study. By constructing a clear sequence of these

35
Computer Activities, a pattern of Talk Activity may be built which illustrates how

collaborative communication functions throughout the complete cycle of

microworld construction. KidSim microworld creation involves sixteen categories

of Computer Activities. Table 2 provides a complete list of the categories of

Computer Activity. A second coder was not used in the analysis of Computer

Activities. This researcher believes that longitudinal knowledge of an entire

KidSim interaction session is necessary to adequately code for some Computer

Activities. No sample of the video or audio taped interactions could guarantee

enough information for an adequate analysis. Computer Activity coding remains

a largely subjective process in this particular context.

For example, talk produced during Creating a New Piece would be very

similar to talk produced while Modifying a Piece's Appearance. Only by being

present during the complete interaction, could a person be expected to identify,

within the context of the simulation, each Computer Activity. Talk Activity Coding

does not share this burden. All relevant information to code for Talk Activity type

is contained within a representative sample of the complete transcription.

36
Table 2
Distribution of Talk Units per Computer Activity Category (N=4528 4)

COMPUTER ACTIVITY % of Talk Units n

Rule Writing 23.3% 1057


Creating a New Piece 22.2% 1006
Modifying a Piece's Appearance 17.7% 800
Running the Clock Forward 13.9% 628
Planning the Microworld 5.8% 261
Setting a Piece on the Gameboard 4.6% 207
Rule Grouping in a Sub-Routine 4.2% 189
Vacuuming a Piece 1.7% 76
Modifying a Piece's Position 1.6% 73
Naming a Rule 1.0% 47
Naming an Appearance 1.0% 47
Modifying a Piece's Energy 0.9% 39
Naming a Piece 0.7% 31
Naming the Microworld 0.5% 23
Vacuuming a Rule 0.5% 23
Running the Clock in Reverse 0.5% 21
Total 100% 4528

37
Computer Activity Analysis

A detailed discussion of KidSim Computer Activities will now be presented.

The emphasis will be placed on the four most prevalent of these: Creating a

New Piece, Rule Writing, Modifying a Piece's Appearance, and Running the

Clock Forward. Each type of Talk Activity will be considered in decreasing

frequency of appearance within a particular Computer Activity. The four primary

Computer Activities appear in Bold Type.

Anne and Jenny's approach to Creating a New Piece will be considered

first, as they created by far the widest variety of pieces. Subsequently, Mark and

Cary, the second student pair that became part of this study, illustrate the use of

"I Don't Care Squares", one of KidSim's most useful tools during Rule Writing.

Third, Amber and Marie, show a certain "flair" in Modifying a Piece's

Appearance, while creating the most technically sophisticated of these three

microworlds. Running the Clock Forward, actually watching a microworld

"animate", became the fourth computer activity reviewed in detail.

Creating a New Piece is addressed first, as it is the fundamental activity

of microworld creation in KidSim. Creating a New Piece (22.2% of total Talk

Units) was found to include only slightly less talk than Rule Writing (23.3%).

Throughout the discussion, screen images will be included within the text to

illustrate portions of the current microworld. Any relevant post-hoc interview

comments will be included at appropriate moments throughout each activity

discussion.

38
Creating a New Piece

Pieces or Game Pieces are the objects that make up a given microworld.

Pieces may or may not be animated, but they must always be constructed or

drawn. By pressing the button the Appearance window is opened,

allowing a new piece to be created (see Appendix A, Figure A2). New piece

creation is the most fundamental and perhaps the least complex of the five

primary KidSim Computer Activities isolated by this study. As shown in Table 3,

Cumulative Talk was the most prevalent Talk Code type exhibited during

Creating a New Piece (37.3%), followed by Query Talk (32.9%). This talk was

largely goal directed (Cumulative) mixed with self-checks (Query). Creating

pieces is very much like drawing, an activity with which most children are quite

familiar. A result of this familiarity was a decrease of Help Talk (12.6%) as

compared to later phases of microworld production (see Rule Writing, p. 51).

Table 3
Percentage of Talk Units per Talk Category for Creating a New Piece
(N= 1006)

TALK SEQUENCE CODE % of Talk Units n

Help Talk (Researcher Initiated) 7.0 70


(Student Initiated) 5.9 59
Cumulative Talk 38.4 386
Query Talk 32.1 323
Disputational Talk 8.1 81
Solo Talk (Engaged) 5.4 54
(Disengaged) 0.3 3
Exploratory Talk 3.0 30

39
Cumulative Talk during Creating a New Piece (38.4%)

The moment this author met his first pair of students, Anne and Jenny, he

knew he was in for an interesting experience. Anne was very outgoing and quite

talkative while her best friend Jenny was rather shy and quiet. These two

decided to build ANNE and JENNYs ZOO as their KidSim project. The following

first few lines of dialogue on their first day of interaction reflects the fact that they

are willing to point out "discrepancies" in the drawing to each other.

Jenny begins by using the mouse to draw the bars of a cage.

5 A: That one's bigger.


6 J: Hmmm, Oh, well.
7 A: Never mind, go ahead.

Anne refers to two different line widths when commenting on Jenny's

drawing (unit 5). This is a good example of Cumulative Talk because Jenny (unit

6) hears Anne's objection, but deems it a minimal problem. Anne realizes that

Jenny has a good point and decides to let Jenny continue drawing rather than

protest further (unit 7). Jenny was able to continue creating her piece because

she acknowledged Anne's comment. Anne, having been heard, is able to agree

with Jenny that "the problem" is a minor one, not worth really complaining further

about.

Cumulative Talk is normally accompanied by a continuation of activity as

opposed to Disputational Talk which tends to temporarily stop progress. Anne

and Jennys' Cumulative Talk serves as a safe way to continue toward their goal,

without interruption, early in the interaction. Anne and Jenny are just getting to

know KidSim and may be reluctant to delay construction at this beginning point.

40
142 A: Let's just make those line type things that I did.
143 J: Okay. Go to clear. This time I'll do it.
144 A: No, I have a better idea! It's with the lines. Look
145 J: OK.

Here Anne makes a suggestion which is met with an affirmative response

from Jenny (units 142-143). Anne wants Jenny to draw wide, straight lines to

serve as the bars of the cage and clarifies her idea by demonstrating for Jenny

how to draw using the line tool (unit 144). Each child's talk is cumulative as it

acknowledges the presence of the other and confirms to the other that they are

interested and involved in the current activity. Cumulative Talk is an indicator of

forward progress. Regardless of the potential for deferences to occur,

Cumulative Talk is inherently unthreatening. It is a relaxed, low-key form of

dialogue, that may even lead to laughter. Anne believes that the "tiger's stripes"

must be drawn with the free-form paintbrush tool, but Jenny realizes the line tool

makes straighter lines (units 243-244). They both laugh when they realize that a

tiger's stripes do not always need to be straight.

243 A: Oh, you have to be on paintbrush.


244 J: No it don't. So it don't look more cruddy. That
doesn't matter, their, emm, lions, their stripes aren't
always.... (laughter)
Each partner is "tuned in" and actively contributing to the current process

whenever Cumulative Talk is present.

259 A: That looks like a great tiger!


260 J: Uhm hmmm, it's not cruddy as I thought it would
be.

Creating a New Piece as a whole is enabled by Cumulative Talk.

Cumulative Talk keeps the process of piece creation moving forward, promotes

collaboration between partners, and leads to the creation of cooperatively

41
engineered pieces. Cumulative Talk proved successful for Anne and Jenny by

providing them with a method for effective, comfortable collaboration.

Query Talk during Creating a New Piece (32.1%)

Unlike Cumulative Talk, which indicates a condition of "all is well", Query

Talk, often serves as a kind of "self-check" or request for confirmation from the

other partner. Query Talk is a friendly way that one partner solicits input from the

other. Queries openly invite participation from the other collaborator.

Queries "ask" directly for input. A person asking for input is often in the

process of self-monitoring. They are wishing to check their own perceptions

against those of their partner.

9 J: Does the mistakes look good or should I erase


them?
10 A: Erase them.

Jenny's Query Talk (unit 9), is a request by Jenny as to how "extra marks"

on the fence drawing should be handled in the future. Queries indicate a

willingness to solicit the advice of the other partner.

215 A: Want me to just draw the face?


216 A&J: (giggles)
217 J: Draw a whole body.

Anne starts to draw a tiger, but feels uncertain about how much of it to

draw (unit 215). Anne is checking with Jenny to see if it is "OK" with her that she

draw the whole body. Jenny's response, accompanied with joint laugher (unit

217), hints that they are starting to realize that they are better off if Anne draws

the more complex shapes (i.e., a tiger's body). Anne's query to Jenny indicates

to her that her input is valued.

42
In the following excerpt, Anne consults with Jenny as to "how to make a

tiger." Jenny replies with a question, indicating she is not quite sure about her

answer (unit 227).

226 A: You know how to make a tiger?


227 J: You make stripes?

Query Talk allows the questioner a chance for non-threatening, invited

feedback. By using a query, instead of waiting for a reaction to occur, Jenny is

able to confirm an addition to the piece immediately (unit 232).

231 A: Beautiful tail. Make its tail come around.


232 J: Like that?
233 A: Yeah. Now I want to put the stripes. It more looks
like a lion, but oh well.

Anne is quick to compliment Jenny (unit 231). Now Anne feels it's her turn

to put in "the stripes" (unit 233). Anne's response is important as it allows both of

them to confirm what the other sees as her individual skills. Anne and Jenny are

beginning to understand, through their use of Query Talk, to which duties each

might be better suited. Jenny, probably the better artist of the two, has taken

responsibility for the more complex features of a drawing. Anne decides to do

the less challenging straight lines. Neither of these duties is truly easy, as

drawing with a mouse, in itself, is always a challenge. Query Talk indicates the

other's input is still important, regardless of who is currently doing the drawing.

Through the use of Query Talk, Anne and Jenny's roles were better

defined. They found a way of collaborating which facilitated piece production by

sub-dividing their "creative expertise." They also found, in the same solution, a

way to respect each other's talent and contribution.

Anne and Jenny often sought confirmation from each other about their

drawings. The resulting sharing of duties had a significant boost to overall piece

43
productivity. In this excerpt, Cumulative Talk is mixed with Query Talk in almost

equal quantities.

218 J: You have to go to paint.


219 A: Yeah.
220 A: Let me use the paintbrush, Jenny. So, push down
on shift... Oh no I don't.
221 J: Yes or no?
222 A: No.
223 J: Have you've been pushing down on the.... I'm just
going to make the head. (Laughter)
224 A: You want to go ahead and make it for me?
225 J: Yeah.

Anne and Jenny's creation is more collaborative here, with each partner

consulting the other about the current drawing. The smooth progress of

Cumulative Talk is interspersed with brief queries for confirmation or requests for

help from the partner (units 221 & 224). Both were able to stay continuously

involved in the process. As a result, they were able to produce a larger variety of

pieces than any other pair of students.

Cumulative and Query Talk were found to be primary, foundational types

of collaborative communication during Creating a New Piece. When both of

these varieties of talk were present in high quantities, piece production

proceeded at a sure pace, with both partners making contributions. Jenny's

questioning of Anne's color choice (unit 246) allows Anne to correct a subtle

mistake (unit 247). By using a query, instead of becoming disputational, Jenny

allows Anne to "save face." Jenny causes Anne to think about her selection,

rather than telling Anne that she is wrong.


245 A: Hi, again, oopsy...
246 J: Why are you using red?
247 A: I used this color. There. Now I just have to...
(whispers).
248 J: Make it like that.
249 A: Only 15 minutes.

44
Query Talk may allow a certain degree of "discord" to come through

without the questioner appearing to be in complete disagreement (unit 246).

Query Talk is a good way to register disagreement without becoming

disputational. By continuing the mix of Cumulative and Query Talk, with the

occasional aside of a Solo Talk remark, Anne and Jenny are able to progress

through their drawing with little friction.

Help Talk during Creating a New Piece (12.9%)

New piece creation is the first Computer Activity each KidSim user

encounters. At this early point in a child's interaction with the program, some of

the program's tools may be unfamiliar to them. Help Talk is a way to receive

advice about some aspect of the software. KidSim has many features which may

benefit the user if they are aware of their existence. The following Help Talk

concerns the Flip/Rotation Control Buttons (see Appendix A, Figure A2):

164 JD: Just to remind you all...


165 A: Huh?
166 JD: The little arrows down there allows you to flip
things left to right, top to bottom, and turn 'em around a
quarter turn. See the three sets of arrows down there?
167 A: Um hmm, right here.
168 JD: ...in case you want to move something around.
169 A&J: Oh, OK.

JD wanted to review these particular controls, as Jenny indicated earlier

that she might want to turn over a piece sometime in the future. This example

of Researcher Initiated Help Talk reinforces the notion to the collaborators that

they are in a non-judgmental atmosphere. Their microworld building activity was

45
not a test of KidSim proficiency, but rather an exercise to observe the process of

collaborative communication.

Recall that Help Talk can appear in two ways, as either Student or

Researcher Initiated. While interacting with KidSim, these young collaborators

sometimes became "stuck" and needed help from JD to continue. Help came

both in the form of answers to collaborator questions and volunteered

suggestions from the researcher.

Not all Help Talk was treated as appropriate for a response from the

researcher. Any question from a student which was qualitative in nature was

immediately resubmitted to the asking party for further contemplation.

118 A: (To JD) Do you think grass...is this okay?


119 JD: Do you mean blocks of green? It is up to you.

The researcher tried to remain neutral, but Anne was looking for a different

answer. Anne wanted a third party (JD) to take her side in an extended

argument. Student Initiated Help Talk used in this fashion is not simply a request

for more information, but rather a solicitation for an opinion and support.

Help Talk gave these student collaborators critical information about

KidSim when it was requested, as well as, supplemental information, as it was

deemed useful. Help Talk most certainly brought the researcher further into the

collaborative activity of the KidSim users. It is believed that this researcher input

gave the students some information necessary to continue their collaboration,

information without which their creative process could not continue.

Help Talk during Creating a New Piece was largely supplemental in nature

as compared to communication with the researcher concerning other Computer

Activities. Creating a New Piece is in many ways simply another form of

46
drawing, an activity with which most fifth grade children are very familiar. Help

Talk was only needed to increase the children's awareness of KidSim's drawing

tools.

Disputational Talk during Creating a New Piece (8.1 %)

Disputational Talk might be thought of as the antithesis of Cumulative

Talk. When present, it is a sure sign that all is not going well. Disputational Talk

is disagreement without immediately giving any reasons for the dispute.

Disputational Talk has a kind of "How dare you!" air about it. Disputational Talk

signals that collaborative microworld construction has halted until the dispute is

settled.

The following example illustrates Anne and Jenny's first disagreement.

Jenny does not agree with Annes actions, but does not give any evidence to

support her objection (Unit 110). Jenny does not think that the Paint Brush

tool that Anne is using at the moment colors in large areas very well.

110 J: You can't do that.


111 A: Uhh?
112 J: Just do that.
113 J: No!
Jenny is trying to convince Anne to use the Paint Can tool to expedite

the filling in of an area with paint (unit 112). Anne does not like the effect of the

Paint Can (unit 113).

Disputational Talk lets the other person know that "a problem exists"

without being specific as to what is the actual objection. The use of Disputational

Talk is a good way to voice a current concern without having to put immediately

into words just what is the area of concern. Children, in particular, may be

especially prone to statements which approximate the phase, "Just because."

47
Disputational Talk gives them a way to "ask" for more time to think about the

exact nature of their disagreement. The pressure to provide a

complete justification for one's way of thinking is lessened for the moment.

114 A: What?
115 J: Can.. That thing.
116 A: This thing?
117 J: Right there... See

Jenny can still not express what it is that she wants Anne to do. She

disagrees with the way Anne is drawing (unit 110) , but is unable to put this

objection into words. Instead, Jenny is relying upon pointing to the computers

monitor to explain her objection (unit 117). Jenny (and probably Anne also) did

not know the name of the tool she was urging Anne to use. Whenever the name

of an on-screen icon is unknown to a user, it is often difficult to communicate

about that object. Pointing to the object is a poor substitution for talk, as many

small icons may occupy the same area on the screen.

Anne and Jenny's disagreement escalates from the use of the Paint Can

to an extended discussion about how the piece they are creating, grass, should

be drawn. Anne disputes Jennys block-like grass and starts to create a

different look.

120 A: NO!
121 J: We have to make it professionally.
122 A: Just go like this. Hold the shift key down.
123 J: How's the shift key even help? I have a better
idea.
124 A: That looks good green grass.

Jenny's Disputational Talk (unit 121) introduces the concept of

professionalism by implying that Annes effort is anything but professional.

Jenny thinks block-like "grass" is more professional looking. Anne continues by

commanding Jenny to use the Macintosh technique of holding down the shift

48
key (to help constrain lines into being straight, either vertically or horizontally),

without giving any reason why this is necessary (unit 122). Jenny queries that

she sees no advantage in this effort (unit 123), and goes on to demonstrate a

solution of her own. Jenny disputes Anne's actions by beginning a different

course of action of her own. Anne continues the dispute by ignoring Jenny's

question and her claim that she has a better way (unit 124). Anne now

announces that she has drawn good green grass (unit 124).

Jenny does not agree. Both Jenny and Anne have very different ideas of

what grass should look like. Each person feels that they have the right to tell the

other how grass should look. Why might this be so? Grass belongs to a group

of things in the world generally referred to as common objects (Norman, 1990).

Common objects are a part of each individuals personal history. Most children

growing up in small-town America probably have some personal experience with

lawns and grass. These are common objects whose look they are certain

about.

We each come to own our own personal view of how these things look

and act. If something is presented to us and it is given a label which belongs to a

class of common objects we expect to look a certain way, and behave likewise.

Otherwise, we reject the label, saying something to the effect of: Gee, that

certainly doesnt look like a tree to me.

Jenny does not like Annes idea of green grass:

125 J: Here. Can I show you my idea of green grass?


126 A: See? That looks good!
127 A: I don't like straight lines.
128 J: What?
129 A: I don't like straight lines.

49
Jenny and Anne are having a battle of ideas. Jenny sees grass as being

a series of rather straight lines, whereas Anne conceives of grass as being more

random, less rigid. Through Disputational Talk Jenny and Anne can continue to

note their displeasure with the current situation. However, as long as they elect

to respond to each other in only a disputational fashion, they make little progress

toward the resolution of their problem. Persisting on this course, Jenny and Anne

will never be able to complete any more pieces.

159 A: I don't like it how you did it.


160 J: No, I'm not going to do it that way. See, just make
it like this.
161 A: ok. [ almost whispered ]
162 J: OK?
163 A: OK.

Once again, Disputational Talk is used to raise an objection (unit 159),

without giving any reason for that objection. Jenny counters with a disputation of

her own and demonstrates her better idea visually (unit 160).

This is an example of how disputes during Creating a New Piece are

often resolved. One person demonstrates to the other what they had in mind

visually, in lieu of vocalizing a defense. The "electronic slate" of the KidSim

gameboard gives these collaborators an "interactive tablet" on which to illustrate

their ideas. Disputations are often settled in this nonverbal fashion. However, in

this case, Annes first soft spoken ok is only half-hearted agreement. Jenny

queries to confirm Anne's response (unit 162).

After the "grass incident", Anne and Jenny learned that Disputational Talk

is not very productive and only leads to prolonged disagreement. Cumulative

Talk such as this later excerpt (units 234-236), continued the flow of the

construction while at the same time allowing minor changes to occur.

50
234 J: Oh well.
235 A: The stripes are going to be red.
236 J: Red? They're usually black.
237 A: Oh, yeah. That color black or that?

Jenny may now understand that it is not enough to object to Anne's

actions disputationally. Jenny begins to give reasons for her objections (unit

236). Jenny causes Anne to pause and consider her question, "Red?", going

beyond a disputational comment, by mentioning the way tiger stripes appear in

the "real world." A disputational reply would tend to impede further progress until

a resolution of the crisis could occur.

During that day's post-hoc interview, Anne puts into her own words how

she feels about Jenny's grass:

471 JD: What didn't you like about Anne's, rather, Jenny's
grass?
472 A: It looked too plain. It had no squigglies. It was
just plain green.
473 JD: Do you like yours better?
474 A: Yeah.
475 JD: Why?
476 A: Because it looks like more like grass.
477 JD: Grass it should have squigglies?
478 A: Yeah.

Anne clearly prefers "squiggly" grass, because it "looks more like grass".

Jenny, during her interview, had a different idea:

545 JD: Why didn't you like Anne's grass?


546 J: There's some things I just don't like what she does!
Um, it was funky.
548 JD: What was funky about it?
549 J: I mean, it just sprayed all over the place and it
didn't look like grass.
551 JD: What, what should grass look like?
552 J: Grass is like that and it's like all colored in, and
she just had it going all over the place.
554 JD: Just too crazy?
555 J: Yeah.

51
Jenny likes grass that is "all colored in" (unit 552), not "sprayed all over the

place" (unit 549). Anne and Jenny clearly see grass differently. During the

creation of their microworld, one is able to observe how these children see and

interpret the natural world. They are communicating to each other and to us, as

observers, how best to represent what they "picture" when they see their world.

By building microworlds a person must examine how they perceive the ''real"

world. In order to copy reality at some level, they must first try and express how

reality looks and feels to them. Collaborative communication allows them to

share a "new reality" of their own making: the microworld.

Disputational Talk allows Anne and Jenny to confront each other about

certain things in which they are in disagreement. First, they can not agree upon

which tool to use for drawing grass (the Paint Brush or Paint Can). Later, they

have difficulty agreeing upon the "correct" appearance of grass. Despite these

two issues, Disputational Talk during Creating a New Piece amounted to only

eight percent of the total talk. Disputations may be a rare occurrence for best

friends in general. It may take something like the clashing "personal histories"

mentioned earlier for best friends to become involved in a extended dispute.

Solo Talk during Creating a New Piece (5.7%)

Solo Talk is literally "talking to one's self", whether intentional or not.

Any statement which is "not-replied to" by the other collaborator is considered

Solo Talk. Solo Talk, is classified as either engaged or disengaged. Engaged

Solo Talk is Solo Talk about the current computer activity. Disengaged Solo Talk

is Solo Talk which is "off topic" or not pertinent to the current computer activity.

52
The following excerpts of Engaged Solo Talk occurred as students were

Creating a New Piece.

46 A: That thing doesn't help really. [that thing is


the lasso].
219 M: I see a couple. Make some taller ones.

These two examples from Amber and Marie highlight Solo Talk which is a

suggestion by one collaborator to the other. Solo Talk gives the partner which is

not currently controlling the mouse a way to still have some input into the creative

process. Amber suggests that Marie use a different tool (unit 46). Marie tells

Amber about some stray lines in her drawing (unit 219). The fact that these

statements are "not-replied to" may indicate that the "mousing" partner is too

engaged in what they are doing to respond verbally. Otherwise, the speaking

collaborator may simply be unheard or is ignored by their partner for the moment.

Anne and Jenny provide some more examples of Engaged Solo Talk

suggestions.

71 J: You just have to go to OPPS.


74 J: Keep going.
87 A: Just leave it.
182 A: Oh, just put one right there.
228 A: It's not as straight...
264 J: That right there.
265 J: Just make one.
375 A: Give it a small tail.
725 A: Click it there.
798 A: We need it more redder, don't you... or is it
lipstick?
837 A: Yeah, that looks good. Just leave it like that. I like
that. That looks dingy. Sorry. Oh cool! Yeah, I like
that.

Many of these statements do not have a verbal response, but are

responded to non-verbally via their partners on-screen actions. In this sense,

Engaged Solo Talk , is similar to Cumulative Talk, in that it does often elicit a

53
response from the partner, albeit a non-verbal one. For example, Anne in units

182, 375, & 725 is making suggestions to Jenny as to what she should do at the

current moment. Anne's comments do not require verbal feedback; an on-screen

response is all that is necessary.

Another variety of Engaged Solo Talk includes a note of "self-

congratulation" on the part of the speaker. Solo Talk exhibited in this manner

may be taken as a confident sign that the speaker is pleased with the current

outcome (units 8, 126, 183, and 270). This self-congratulatory Solo Talk usually

does not elicit any kind of reply from the other partner.

7 A: Never mind, go ahead. (Long Pause)


8 J: Got that line fine.
9 J: Does the mistakes look good or should I erase
them?

125 J: Here. Can I show you my idea of green grass?


126 A: See? That looks good!
127 A: I don't like straight lines.

182 A: Oh, just put one right there . (Long Pause)


183 J: That looks like an animal.
184 A: Wait. Don't we need another cage? (Brief Pause)
Hello? Thank you.

269 J: Emmm, no eraser. (Long Pause)


270 A: Okay, that looks good.
271 A: Just put it on the thing and it looks perfect.

Self-congratulatory Engaged Solo Talk allows the speaker to make

"pronouncements" about one's current efforts without demanding that their

partner agree or disagree with them as a query might ask them to do. Engaged

Solo Talk lets the other collaborator "in" on how the speaking partner feels the

current computer activity is going.

54
Solo Talk can also occur which is disengaged or truly unconnected to the

current computer activity. Most of these Disengaged Solo Talk turns which

occurred during Creating a New Piece were comments about the current time of

day.

249 Anne: Only 15 minutes.


216 Mark: " 2:55".
267 Mark: One after three.

Disengagement should not always be characterized as a negative

sign as far as the interaction is concerned. Talk about the current time of day is

an indication of one of two things: 1) either the speaker is bored with the current

computer activity, in which case time is passing too slowly or 2) it is meant as a

warning to their partner that the current computer activity must be "sped up" if it is

to be completed by the end of the interaction session. A disengaged partner is

the most likely one of the two to notice what "time it is" and thus has the

opportunity to comment on this observation. A warning of "time remaining", such

as Anne's use of the word "only" in unit 249, serves to spur on the activity and

heighten the urgency for completion. Otherwise, Mark's reports of the time of

day (units 216 and 267), spaced just six minutes apart, may be an indication that

he has resorted to watching the clock as a way of suppressing boredom.

The idea that Mark was looking for some means of distraction at certain

points during new piece creation is supported by reviewing these further

examples of Disengaged Solo Talk.

272 M: I told you! Told you!...


274 M: ...not-to-mess-up.
275 M: Jumbo jet.
277 M: Okay, jumbo jet person.
280 M: (whispering....Erase, erase, erase. Jumbo, jumbo
jet, time to erase, erase, erase).

55
Marks' comments illustrate a partial detachment from the current computer

activity. The subject of his talk, the Jumbo Jet, is a piece that his partner Cary

was drawing at the time. In lieu of assisting Cary in the creation of the piece,

Mark withdraws into a kind of fantasy state where he begins talking to himself

about a "jumbo jet person" (unit 277). Mark does notice a problem with an extra

scribble in the drawing (units 272-274), but he does not articulate his solution to

Cary. Mark elects instead to whisper a hint of a suggestion (unit 280).

Solo Talk is serving many functions for these collaborators as they create

new pieces. If a speaker is engaged in the current activity during Solo Talk, their

comments provide the other collaborator with a sense of how they feel the

current creation is progressing. These comments may be in the form of

suggestions via the "non-mousing" partner or "self-congratulations" from the

"mouse user". Non-engaged students used Solo Talk as self-entertainment, to

express boredom or to make a note of time remaining in the interaction session.

Exploratory Talk during Creating a New Piece (3%)

Exploratory Talk allows one member of the collaborative pair to bring the

other's attention to a particular aspect of the drawing process by introducing new

information or a new point of view. This "focusing of attention" allows a new way

of seeing or thinking about the creation to be entertained. In this excerpt, Anne

discovers a more positive, less confrontational way to approach the impasse'

regarding the "look" of grass.


130 J: That's not going to be a straight line.
131 A: Look how it is up there?
132 J: Hmmm...
133 A: Do you like it up there? How it is? I don't. But
that's what it's gonna look like.
134 J: Okay.

56
Anne uses Exploratory Talk (unit 133), as a way of explaining her

objection to Jenny's drawing instead of continuing to be disputational. Anne

points out how Jenny's grass is going to look on the gameboard, once it has

been demagnified outside of the Appearance Window. Anne is trying to get

Jenny to see the "grass" she is drawing from a new perspective. Exploratory

Talk, in this situation, serves as a "pressure relief valve", a way to cool down an

argument before things "go too far." Usually consisting of factual, logical

statements or requests, Exploratory Talk, lets some form of proof be introduced

into a discussion. Exploratory Talk is a new option here, setting the facts out "on

the table" for all of the world to see. Once introduced, the "evidence" can be

considered on its own merit, disconnected from the party that provided it. Anne

hopes that by providing Jenny with the "gameboard view" of her drawing, Jenny

will agree with Anne that the drawing must be altered.

157 A: You don't want it too big, it might look bulky.


158 J: See why I don't want it like this is because it is
gonna to be like that on our screen and we can't turn
it over like that.

Jenny's Exploratory Talk (unit 158) is similar to what Anne said earlier (unit

133). Jenny is trying to get Anne to see how the grass will look from a different

perspective, out "on the screen". Jenny believes that the way the finished grass

will look on the gameboard is too linear. Annes grass, in Jennys opinion, will

not look good rotated to sit on its side. Rotation is an option which Jenny seems

to think is important (Rotation of an object is an advanced feature

of drawing which few of these KidSim users employed).

173 J: I like them little spikes.


174 A: I like that grass. (?) I want to do something. You
have that little line right there?

57
175 J: Yeah. Where did that come from? I think you went
over too far and it went out of the screen. We can get
rid of it later.
176 A: They're like bushes.
177 J: I like that.

At last, Anne and Jenny's collaboration has created a new piece: grass.

Jenny offers some more Exploratory Talk (unit 175) in an effort to explain why an

extra line appeared on Annes drawing. Anne suggests that the grass they have

been drawing looks more like bushes. Perhaps this change of idea about

what they have drawn is just the compromise that was needed. Bushes, while

certainly common objects, have a much broader range of appearances than

does grass. Bushes can have spikes and at the same time have more

professional looking straight lines in them. Bushes are less universal than

grass. Perhaps the more ubiquitous an object is, the greater the demand to be

faithful to that object's appearance.

Exploratory Talk, may not always succeed as an attempt to inform the

other partner about a new perspective. In this example, Jenny has an idea about

how to create stripes for their zoo's tiger.

250 J: Don't make it all on his face.


251 A: Ahhm, there are stripes.
252 J: But, we have make it right, yeah? And do his feet
like that, and then we'll reverse and reverse there
and have the feet striped.
253 A: Huh?
254 J: Nothing. Oh we have to do something on the tail.

Jenny uses Exploratory Talk (unit 252) to verbalize how she can use the

technique of "reversing" or flipping an area of a drawing over to create the mirror

image of an object. However, Anne does not understand her idea. Rather than

embarrass herself or her friend, the more quiet Jenny chooses instead to change

58
the subject. Jenny may be well aware of the concept that she is trying to explain

to Anne, but she may not have the words or the will to do so. A collaboration

between best friends may carry the blessing of familiarity as well as the burden of

preset limits. Friends know when and when not to push the other.

Summary of Creating A New Piece

Through a cycle of Cumulative Talk, interrupted by the occasional

Disputation or Query, Anne and Jenny gained a sense of each other's

perspective about certain things in the world. Collaborative communication

allowed them to negotiate not only how certain objects should be drawn, but also

who would be drawing them. Anne became the primary illustrator for structures

and simple shapes. Jenny tackled the more complex objects and served as a

kind of "quality control supervisor" for the entire process. On-screen visual

demonstrations gave each collaborator a way to express their ideas without

having to verbalize them in advance.

Cumulative Talk, the majority of talk during Creating a New Piece

(38.4%), signaled a period of calm progress in which each collaborator was

aware of the current activity and contributed positively to it. Query Talk (32.1%)

was the second most prevalent kind of talk during Creating a New Piece. Query

Talk functioned both as a plea for confirmation as well as served as a subtle way

of voicing disagreement. Disputational Talk (8.1% of Creating a New Piece talk)

proved to be counterproductive. After the "grass incident," later Disputational

Talk was squelched early, after only 2 or 3 turns. Disputational Talk places a

barrier to further communication whereas Exploratory Talk, while occurring only

3% of the time of Creating a New Piece, served as an invitation to more

59
discussion. Exploratory Talk also played the role as a kind of peacemaker,

bringing in "facts" or "visual demonstrations" to help settle an argument and

continue the drawing. Solo Talk (5.7% of Creating a New Piece talk)

contributed little to Anne and Jenny's collaboration and mostly served as a form

of self-narrative by Anne and Mark. Solo Talk is perhaps a largely personality-

driven phenomena, because some people simply talk more to themselves.

Finally, Help Talk (12.9%) was not as prevalent during Creating a New

Piece as it would be during later Computer Activities. The act of drawing is

adiversion which most children are very familiar with by the time are in the fifth

grade. The collaborative setting of this study gave them a chance to draw

together, interactively, with the need to learn only a few simple tools. The

necessity for Help Talk was lessened as a result.

Anne and Jenny's mixture of Cumulative & Query Talk created a total

of nine pieces for their zoo. In addition to the Cage, Grass, Tiger and Sidewalk

detailed above, they drew an Elephant , a Fence , a Monkey ,

a Girl , and a Hamburger . Each of these creations became a part of

ANNE and JENNY's ZOO as shown in Figure 1. Through a cycle of cumulative

plus queried discourse, Anne and Jenny were able to produce more characters

than either of the other pairs. Anne and Jenny's willingness to talk about most

any detail of every piece resulted in a fluid creation process. Their collaboration

became a unified creative force through copious amounts of give-and-take.

The less "collaborative" types of talk (Disputational and Solo) did not

dominate Anne and Jenny's discussion. After an initial path of disputational

"dead ends," their cycle of Cumulative and Query Talk, interspersed with an

60
occasional Exploratory remark, functioned as a very efficient way to Create a

New Piece. Anne and Jenny's interaction pattern allowed each person to find a

niche in the collaborative communication process, permitting them to create a

greater variety of pieces than any other pair in this study.

Figure 1
"ANNE and JENNY's ZOO"

61
Rule Writing

The discussion of KidSim's four primary Computer Activities now

continues with the most prevalent in terms of Talk Units (23.3%), Rule Writing.

Once a piece in a KidSim world is created and placed on the gameboard, it must

be given rules in order to animate it and provide for the possibility of interaction

with other pieces. Animation of KidSim pieces is accomplished through Rule

Writing. Help Talk during Rule Writing has a higher rate of occurrence (61.6%)

than any other type of Talk Unit (See Table 4). This finding is reasonable as

Rule Writing proved to be one of the most difficult skills for these new KidSim

users to master. Rules are created via the Rule Window (see Appendix A,

Figure A4), which is accessed through a piece's "Brain" (see Appendix A,

Figure A3). Table 4 lists the distribution of Talk Units for Rule Writing.

Table 4
Percentage of Talk Units per Talk Category for Rule Writing (N= 1057)

TALK SEQUENCE CODE % of Talk Units n

Help Talk (Researcher Initiated) 20.3 215


(Student Initiated) 41.3 437
Cumulative Talk 21.9 231
Query Talk 11.2 118
Disputational Talk 1.9 20
Solo Talk (Engaged) 2.4 25
(Disengaged) 0.9 9
Exploratory Talk 0.2 2

By pressing the button, a new rule is generated with identical left

and right sides or before and after states. KidSim's system of "programming by

demonstration" records any interaction with the computer while the Rule Window

62
is open and interprets it as a series of changes. These changes are then

reflected in the right (after) side of the rule. The talk for accomplishing Rule

Writing will be illustrated with examples from both of the remaining pairs of

students that participated in this study. These students and their "worlds" will be

introduced first, followed by a discussion of the most frequent types of talk

exhibited during Rule Writing.

Mark & Carys' Fantasy World

The second set of fifth graders observed for this study were two boys,

Mark and Cary. Of these two, Cary is the quieter one. He's also taller and wears

glasses. Mark is quite talkative, inquisitive and more prone to distraction than his

"best friend," Cary. They designed and built "Fantasy World." Compared with

Anne and Jenny's ZOO, Mark and Carys' Fantasy World is visually less

populated, consisting of only five pieces:

a Hot Air Balloon , a Jumbo Jet , a Hang Glider ,

a Tower , and a Tree . Fantasy World's scenario involves

Jumbo Jets and Hang Gliders popping holes in Hot Air Balloons, deflating them

and causing the Balloons to fall. The Jumbo Jet also crashes into a Tree and

burns.

Amber & Marie's Animal Life

The third pair of students observed for this study were two fifth grade girls,

Amber and Marie. Marie was probably the most physically and mentally mature

student that participated in this research. Her best friend, Amber, while often

more silly than Marie, worked closely with her friend to build their microworld, the

63
most technically sophisticated of all three KidSim worlds represented. They

designed and built "Animal Life" (see Appendix A, Figure A1).

In comparison to the two other microworlds presented here, Amber and

Marie's "Animal Life" is visually very simple, consisting of only three pieces:

Pokey the Giraffe Suzy Swan and the Wasaki Mountains


, ,

. Animal Life's scenario involves Suzy Swan swimming across an

area of water located in front of the Wasaki Mountain range. Meanwhile, Pokey

the Giraffe walks behind the mountains as he swishes his tale up and down.

Amber and Marie's Animal Life went through many transformations before

it assumed the final state shown in Appendix A. Early in their interaction, a

boulder named Bob (no picture is available) became a "test piece" for movement.

Originally Bob was created to serve as a kind of "island" in the water area in front

of the mountains. Bob was then programmed to move, only to be deleted later.

Amber and Marie's experience with Rule Writing is noted here to serve as

a contrast to Mark and Cary. Compared with the two boys, Amber and Marie

used talk to collaborate together much better when writing rules for their

microworld, "Animal Life". Subsequently, Amber and Marie created the most

technically complex of the three microworlds produced during this research.

Help Talk during Rule Writing (61.6%)

Help Talk supported and strengthened the process of Rule Writing. The

researcher (JD) realized that when new rules were necessary, students often got

64
"stuck", unable to continue microworld construction. In order to maintain their

progress, Help Talk was offered to serve as a guide through a current problem.

An analysis of this Help Talk reveals that early on, this communication

consisted of direct, sometimes in-depth, responses to student questions, often

teaching or reinforcing a particular aspect of KidSim. Later, a mixture of brief

"look-here" hints with other types of talk was noted. A student pair was not

simply told how they should proceed as their microworld building proficiency

increased. Rather, the students were encouraged to "investigate" in a certain

direction to see if they could discover their own solutions.

In this first Rule Writing excerpt, Mark and Cary are creating a rule that

causes the Jumbo Jet to hit the Hot Air Balloon.

317 C: Okay, write a new rule?


318 JD: But this... You want a "hit it" rule first? Yeah, go
ahead, since you're nearby.
319 M: Hit the lollipop that's flyin' in the sun (mumbles).
320 C: Don't we need the.. empty space box?
321 JD: Yeah, yeah. The... "I don't care square".
322 M: "I don't care square".
323 C: Right there?
324 JD: Not quite.
325 C: Right there.
326 JD: No.
327 M: Above yours.
328 JD: Left side.
329 C: Oh yeah. Right here.
330 JD: Yeah.

Cary is doing the Rule Writing here, while Mark tries to help (unit 327).

Often a sequence of Help Talk will begin, as it did in unit 317, with a request from

one of the student collaborators. Cary's question is typical of a "confirmation

only" Help Talk request. Cary only wants his current line of thinking confirmed

and is not really asking for new information. This pattern of "student request/

65
confirmation by researcher" continues throughout this exchange. The Solo Talk

(unit 319) is typical of one of many "mumbled" chants that Mark verbalized.

Often when Cary was drawing Mark would sing or chant. These outbursts were

sometimes tied to the activity at hand. In this case, Mark is singing about the

"lollipop in the sun" as Mark believes the Balloon looks like a lollipop.

331 M: Why do you need the "I don't care" square?


332 JD: That, that rule says that if there's ah, ahh... What
are you calling this, the lollipop, the balloon?
333 M&C: (giggle)
334 JD: If there is a balloon in front of the jet, ahh, that's
all you care about - you don't care if there is
something just above the balloon. So you can have
two balloons stacked up, it wouldn't matter.
335 C: So we've got to move it over?
336 JD: Yeah.

JD continues to try to help Mark and Cary understand the tool known as

the "I Don't Care Square" found in the top area of the Rule Window (see
,
Figure 2). Placing an "I Don't Care Square" on the before (left) side of a rule,

causes KidSim to ignore any objects which a moving piece might encounter, if

they fall within this square.

66
Figure 2
Rule Window with Jumbo Jet diving
(note: the "I Don't Care Squares" in the left portion of the rule.)

Amber and Marie also required help with the "I Don't Care Square". Early

in their Rule Writing experience they were given an on-screen illustration of this

tool's function.

818 JD: By using this "I Don't Care Square" it means that,
it doesn't care! It'll ignore whatever is on the left side.
819 A: There's something on the right side, too.
820 JD: There's something on the right side?
821 M: Should we just cover the whole thing?
822 JD: Yeah, cover it all.
823 A: This rock is just doesn't care.
824 JD: So if there is anything that is right near it like
that...
825 M: If there is anything on top of it though...
826 A: Bob is so insensitive.
827 M: He can't eat anything, though...
828 A: What is he supposed to eat?
829 JD: Well, he doesn't care about anything at this point.
Let's just see what happens. We may erase this rule.

Marie is quick to point out a problem with placing "I Don't Care Squares"

on all sides of Bob (unit 827). If Bob ignores everything in his path he will not be

67
able to recognize when he is near food. Amber points out that this may not be an

issue for a rock (unit 828). Although they choose not to have any of their pieces

actually eat anything, it is interesting that Marie understood the implications of

having Bob (the boulder) truly not "care" or sense anything in "his" environment.

JD's remark in unit 829 points to the shift to the "Let's see what happens" type of

response which dominates later Help Talk.

An analysis of the transaction recordings indicates that Help Talk changed

from complete discussions of program functions into hints about how to proceed

as a student pair's proficiency with KidSim increased. Students were given the

opportunity to "figure out" their next move on their own, whenever the researcher

felt that they could. Whenever the students got stuck on a given point, JD simply

tried to gently push them in the direction of a solution (units 1705, 1708, 1710).

1705 JD: Only one way to find out.


1706 M: It won't. It's gonna stop there.
1707 A: I hope not. Do we have to go down here? Okay,
he's moving from there to there.
1708 JD: Look what your top one says.
1709 M: He's moving there to there, but then that one
says he's moving there to there.
1710 JD: So what should you do?
1711 M: Get rid of the bottom one?
1712 JD: Yeah.
1713 M: Thank you. It's not been working.

Marie is talking about the rules which control how the giraffes walk "along"

the mountains (unit 1709). Since the giraffes move in the same direction, one

step with each click of the game clock, they can never run into each other.

Consequently, other rules are not needed and may be deleted (unit 1711).

Writing rules in KidSim is the primary "programming by demonstration"

technique and requires users to learn the more specialized features of the

software. Compared to piece creation, Rule Writing is a much more complex

68
activity. Writing a rule which incorporates movement involves using the Spotlight

Control (see Figure 3) on the gameboard to open up squares. The Spotlight

surrounds the piece which is selected for Rule Writing. Grabbing a handle, with

the mouse, on the spotlight or right side of the rule window (see Figure 4), allows

squares to be revealed into which the piece may be moved.

Figure 3 Figure 4
The Spotlight Right Half of the Rule Window

Rule Writing requires a different, more proactive kind of communication

with the researcher. Unlike Creating a New Piece, the essence of which is

drawing, Rule Writing is not a familiar activity. It becomes familiar through

collaborative talk by verbalizing the rules which create a piece's programming.

More help from the researcher was needed when the students began to create

rules for their pieces. Help Talk from JD provided supplemental training

throughout the interaction to reinforce various parts of KidSim's interface.

Along with their occasional breakthroughs, Amber and Marie had some

periods of frustration, particularly when learning one of the key lessons about

working with rules in KidSim (units 1187-1189).

69
1187 M: Amber click on the thing. It won't....
1188 A: And then what do we do?
1189 M: You have to write rules for every single thing!
1190 JD: It's a simulation. You have to tell it how this world
works.
1191 A: You have to tell it to make the grass sit still.
1192 M: No, you don't! It does that automatically.
1193 JD: Well see, you don't have to tell it how to do
everything.
1194 JD: Okay...

Marie said it well when she discouragingly proclaimed, "You have to write

rules for every single thing!" (unit 1189). JD's Help Talk (unit 1190) acted to

reinforce Amber and Marie as they were trying to master a basic lesson of

KidSim simulation construction: "all actions must have a rule."

1301 JD: Okay. What does your rule say so far?


1302 M: "I'm a rock. Hi."
1303 JD: "Hi, I'm a rock." What do you want it to say?
1304 A: "Hi, I'm a rock. I bounce off giraffe's heads."
1305 JD: It always starts off with if. If...
1306 M: If I'm a rock (which I am),
1307 A: And there's....
1308 M: I'm wantta bounce off this giraffe's head if I come
close, and then go back up and then come back
down.
1309 JD: Listen to what you just said and describe that as
a rule.

Help Talk decreased in quantity as Rule Writing progressed. This

reduction was accompanied by an increase in different types of talk. Sequences

of Cumulative and Query, with occasional Solo, Exploratory or Help remarks,

formed a pattern which allowed microworld construction to continue, minor

problems to be addressed and new information to be examined.

Cumulative Talk during Rule Writing (21.9%)

Cumulative Talk was a distant second to Help Talk during Rule Writing in

terms of overall frequency of occurrence. Cumulative Talk is a type of "verbal

70
presence", an ongoing response to the on-screen action. Whenever such a

pattern of speech occurs, the students are reinforcing and guiding each other.

The existence of Cumulative Talk is usually a good indicator of normalcy or a

"smooth" period of interaction. Rule Writing did not exhibit many of these

periods.

KidSim's "graphical re-write" rules, its actual programming language, are

new concepts to these students. They are not merely transferring old skills to a

new medium as with drawing. Rule Writing required learning and applying new

tools and techniques, ones which had not been thoroughly assimilated yet.

Subsequently, when Cumulative Talk was present in an extended Rule Writing

sequence, this was an indication that the tools and techniques in use at the time

were well understood.

On the second day of their interaction, Mark and Cary had become

much more comfortable with "I Don't Care Squares".


1785 C: NEW RULE...for that dude, that dude, that dude.
1786 M: We want 'em to go down diagonally. We don't
want 'em to go forward.
1787 C: No. Uh oh! We've got to put an "I Don't Care
Square".
1788 M: Yes we do.
1789 C: We don't care if there's anything right there, right
there, right there, right there, right there...or right
there.
1790 C: That won't work with that plane right there!

Cumulative Talk units 1785-1790 are typical of those exhibited by Mark

and Cary when Mark was participating in the collaboration. Here the boys are "in

sync" and each is contributing to the interaction. Mark's point that they want the

plane to go diagonally not forward (unit 1786) is meant with a quick, confirming

"No." by Cary (unit 1787). Cary, in turn, realizes that another "I Don't Care

Square" is needed (unit 1787) to which Mark responds "Yes we do" (unit 1788).

71
The majority of the rules for "Fantasy World" were written or planned by Cary.

Mark drew at least half of the time but did not take a great interest in rule

creation.

The next talk sequence is representative of much of the Cumulative Talk

which took place during Amber and Marie's Rule Writing. Here the girls are

again working to have their giraffe walk along the mountain range. Note that

Amber is concerned that a rule will not be enough to keep the giraffe moving

forward (1703).

1697 A: Right there.


1698 M: And so it moves over, now...
1699 A: Not very far over either.
1700 M: Well, it should be moved to the next one.
1701 A: I know, I'm trying.
1702 M: And how, but we don't want it to stop there; we
want it to keep moving.
1703 A: It'll keep going though, won't it?
1704 M: I don't know.
1705 JD: Only one way to find out.

This sequence illustrates another verbalization of on-screen actions which

give both a context and a purpose for the conversation. Marie reinforces Amber

by "commenting" about the current activity, checking to see if everything is going

smoothly (units 1698, 1700 and 1702). Amber "moves" the dialogue ahead by

giving Marie a "play-by-play" account of each on-screen action (units 1697,

1699 and 1703). Each is letting the other know that they are "with them" and that

they are connected to what is going on. Cumulative Talk allowed each student to

communicate their collaborative connection the other.

72
Query Talk during Rule Writing (11.2 %)

Amber and Marie used 25 percent of their total talk units for Rule Writing.

This is the same percentage as the other two student pairs combined.

Differences in amount of total talk not withstanding, Amber and Marie wrote

many more rules than the other students. They simply tried to do more with

KidSim. Having written a greater number of rules, they also had a greater

number of failed attempts. The following excerpt illustrates their use of Query

Talk (units 927-928) & Help Talk (units 929-930) to help solve a simple problem.

925 A: Okay. NEW RULE. Now... If, hmmm.... If Bob


here...
926 M: Gets on top of that little thing then...
927 A: This ducky, am I doing it right?
928 M: Ahmm, yeah, because he can go backwards.
The duck can go backwards, I guess.
929 A: Well, wouldn't we have to have a ducky rule for
that?
930 JD: I think you would. I'll just answer your question.
How about moving Bob, since we're doing a Bob rule
here.

Note that Unit 929 was coded as Query Talk because Amber (via eye-

contact) is asking JD to confirm her supposition about the "ducky rule". JD also

reminds her of another course of action (i.e., moving "Bob", the rock). JD's

response is (unit 930) hoped to serve as catalyst for further thought about how

objects and rules interact. Amber correctly guesses that in order to have the

duck (i.e., Suzy Swan), move, one must be writing a rule for that duck and not

another piece (unit 929). This can be a confusing point, as the appearance of a

second piece may be changed when writing a rule for the original piece (see

Modifying a Piece's Appearance). Amber and Marie were able to solve their

problem, through a cycle of Cumulative Talk and Query Talk (units 925-928), with

a confirmation from the researcher (units 929-930).

73
Solo Talk during Rule Writing (3.3%)

Mark became less "engaged" as the second and third day's sessions

played out. The primary "evidence" of his disengagement came in the form of

Solo Talk. Much of Mark and Carys' interaction time with KidSim was spent

writing rules, which Mark took less interest in. Rule Writing is an exacting,

rigorous, and often frustrating activity when compared to drawing. Rule Writing

requires a higher level of concentration. Mark "acted out" many times to keep

from being bored or to merely attract attention. Here are some examples of the

many Solo Talk comments Mark made while Cary was using the mouse.
491 M: (yawn) I'm tired. I wanna go home and go to bed.
562 M: (yawning) I don't care.
921 M: (yawn) (whispering- This is taking forever.)
1011 M: (makes noise = OCK!)
[Mark repeated this noise twice more later.]
1966 M: I'm just "I Don't Care Square's" sidekick.
1974 M: ...(singing!) So we want "I Don't Care Squares"
and we want 'em now! We want them now! We want
them now! We want them now! We want them now.
(repeating ever softer)
2015 M: (singing) I want my "I Don't Care Squares", I want
my "I Don't Care Squares"...
Unit 1966 is particularly telling here. By referring to himself as "I Don't Care

Square's sidekick", Mark is saying that he really isn't involved and that he is

inconsequential as far as the activity goes. Mark is simply Cary's sidekick.

Mark may have become bored with KidSim. Cary did not seem to actively

ignore Mark as much as Mark chose not to participate in Rule Writing. Cary did

exhibit many periods in which he was silently "working" with the mouse. Mark,

during these periods simply chose to verbalize the Disengaged Solo Talk

remarks listed above.

KidSim invites different types of talk at every step of the design process.

KidSim, a synthetic environment, does not require collaboration, but certainly

74
encourages it. When two persons are creating together with KidSim each may

support the other as they solve problems and generate new ideas. The use of a

shared workspace, the computer screen, plus the necessity of a shared tool, the

mouse, requires participants to negotiate their interaction time with the computer.

However, forced collaboration can not guarantee equal participation. Mark's lack

of interest, demonstrated through his Solo Talk, did not seem to concern Cary

very much. Cary was able to construct rules and test them with only occasional

assistance from Mark. Cary did attempt to get Mark involved in Rule Writing, as

illustrated within the Disputational Talk section which follows (units 478 &479). A

functioning "Fantasy World" mattered more than involving Cary with talk (see

Figure 5).

Figure 5
Mark and Carys' "Fantasy World"

Disputational Talk during Rule Writing (1.9%)

Disputational Talk only occurred in twenty of the over one thousand talk

turns that were recorded while Rule Writing. This low rate of incidence is most

likely another result of non-familiarity with the Rule Writing procedure itself. Rule

75
Writing demands the acquisition of several new pieces of information about

KidSim. These young users, new to the software, did not have a broad enough

knowledge base to question each other's actions to any great degree.

Program related Disputational Talk which did occur was limited to software

issues which were well understood. For example, Anne (unit 1217) believes that

pasting an object during the writing of a rule is the next required action. Jenny,

who may have a deeper understanding of KidSim, objects with a disputational

"hunh-uh", and declines to explain why Anne is in error.

1217 A: You're supposed to go to Paste.


1218 J: hunh-uh. (meaning "No")
1219 A: ahh-huh. (meaning "Yes")

Jenny, the less vocal member of this pair, is busy trying to copy an object,

and knows that the action "COPY" needs to be performed prior to "PASTE".

Jenny chooses to proceed by carrying out what she thinks is the correct action

rather than responding directly to Anne.

Mark and Carys' only example of disputational speech involved a simple

disagreement over who would write the next rule (units 478-479). Prior to this

point in their discussion, Cary had written all of the rules. Now he feels it is

Mark's turn to take a "shot" at Rule Writing.

478 C: Now write a new rule.


479 M: Not me.

Mark elects not to observe his best friend's "command" (unit 478). It was clear to

this researcher that Mark never became too comfortable with the act of writing

rules. Perhaps he did not want to embarrass himself in front of Cary or JD (the

researcher). The more likely reason may be that Mark simply preferred to let this

76
"more capable peer," his best friend, continue with the creation of rules, while he

continued to watch in a somewhat detached fashion.

Disputational Talk was a very rare occurrence during Rule Writing. The

third pair of KidSim collaborators observed for this study, Amber and Marie, did

not exhibit any Disputational Talk during Rule Writing whatsoever.

Exploratory Talk during Rule Writing (0.2%)

Exploratory Talk occurred only twice while rules were being written. Marie

was doing the talking on both occasions. The possible reason for this scarcity

was, once again, the complexity of Rule Writing. Writing KidSim rules involves

new, somewhat complex processes for these students. Learning how rules are

created and implemented is in itself a challenge. The ability to make exploratory

remarks may only be possible once some threshold of understanding of an action

is achieved. Rule Writing may be too foreign an activity to elicit any significant

amount of exploratory speech. Marie uses Exploratory Talk to help both herself

and Amber understand how precise Rule Writing must be (unit 1033).

1029 M: All right, so do it up here. So what should I just....


1030 A: Pull it down.
1031 M: Pull it down... two squares?
1032 A: Yeah.
1033 M: Now there's still giraffes in the way. But I think I
see a problem. Because, even now, because this
rule might only work if there's two giraffes in the way,
then one.
1034 JD: Remember every, according to KidSim,
everything that's inside the window is important. So
you need to decide what to do with that.
1035 A: Take that little thing and put it right there over
him. That will ignore it, right?
1036 JD: That's pretty clever. Now remember "I Don't
Care Squares" work on the left side of the rule. It's
saying if there's something here you don't care about.

77
KidSim is designed to let pieces cycle horizontally or vertically in a loop

across the screen (i.e., a piece which moves off the right side of the screen will

reappear on the left or a piece which moves off of the top of the screen will

reappear at the bottom). Amber and Marie decided to take advantage of this

feature by making their rock, Bob, bounce off of the head of a giraffe and then

loop over the top of the screen. Help Talk during this sequence (units 1034-

1036) seeks to be supportive, and provide reinforcement, without "leading" the

interaction. This way, Marie is encouraged to keep "exploring" by hinting that she

is thinking in the right direction.

1037 M: Well...
1038 A: (makes sound- whyy?)
1039 M: That means only... I don't know. 'Cause this
means that, uhmm, that if we put one over this and
there's a giraffe on that side but not on this side, it
won't work. And if we put one on that side and
there's not.... we need one straight in the middle.

The girls are writing a rule which causes the rock to interact with the

giraffes. Marie sees a problem with the way the rock is oriented in relation to the

giraffes (unit 1033). She uses Exploratory Talk to point out that if two giraffe's

are within the Spotlight, both of them must be present for a rule to work (unit

1033). Amber has an interesting suggestion (unit 1035), but as Marie correctly

reasons with Exploratory Talk (unit 1039), using "I Don't Care Squares", will

merely lead to a situation in which only giraffes appearing on either the left or the

right side will be affected by the rule. Exploratory Talk allows Amber and Marie

to investigate how important a piece's position is on the Gameboard prior to Rule

Writing. Marie, although actually more accomplished than the others, still found

Rule Writing to be daunting, particularly at first.

78
Summary of Rule Writing

Help Talk (61.6%) accounted for the overwhelming majority of the

communication recorded as these students collaborated to write KidSim

interaction rules. The justification for this finding may be attributed to at least

three factors: program complexity, student ability, and training effectiveness.

Once the fundamentals of Rule Writing were absorbed, often in the last

half of the second day's interaction, each student pair initiated and completed

some rules on their own, with little, if any, Help Talk. At this point in their

interaction, each team began to exhibit the more familiar pattern of Cumulative

Talk cycled with Query Talk, illustrated by Anne and Jenny during Creating A

New Piece.

KidSim is a deceptively powerful simulation program which only looks

simple on its surface. Underneath is a complex visual programming language

which requires the use of layers of "graphical rewrite rules". These rules

translate on-screen programming actions into game piece interactions. To be

able to build a simulation in KidSim with any degree of sophistication, this

language must be understood and utilized efficiently.

Anne and Jenny employed little movement in their "Zoo" and thus had less

need to master Rule Writing. Mark and Cary plus Amber and Marie, on the

other hand, did introduce lots of action into their worlds, and therefore needed to

understand Rule Writing as early as possible. Cary, Amber, and particularly

Marie, did seem to be up to the task. Mark, however, was more easily distracted

and tired of Rule Writing long before the others did. Rule Writing is not an easy

skill to acquire, but as three of these six students demonstrated, it was not too

difficult for fifth graders, given the appropriate time and training.

79
Compared to Help Talk, other types of talk contributed in a much less

prominent way to Rule Writing. Once the basics of creating rules was

understood, a cycle of Cumulative Talk (21.9%) mixed with Query (11.2%) did

prevail. Solo Talk at 3.3% came next in frequency, due largely to Mark and

Cary's input. One of the most influential types of talk found in Creating a New

Piece, Disputational Talk, was present only 1.9% of the time. Rule Writing is a

new, perhaps unique, activity to these students. They expended most of their

energy trying to learn the process, with little time left over to dispute what they

were trying to do or what they had accomplished.

Rule Writing is less flexible than piece creation, inviting a discussion of

procedure instead of style. Rules were written because of emphasis on

movement, not appearance. As a result, the talk changed from a negotiation

about how something should look, into an investigation about how it should

move. In order to conduct this investigation, these KidSim users soon realized

that verbalizing rules proved more difficult than drawing pieces.

The learning of Rule Writing and its application required the assimilation

of KidSim interaction knowledge gained while Creating a New Piece. But the

types of talk which dominated previously do not generalize to the new activity.

Talk which successfully "got pieces created" was present in much lower

quantities as rules were written. The difficulty of learning Rule Writing may

diminish the more interactive types of talk (i.e., Cumulative and Query) while

increasing talk which spurs reaction (i.e., Help).

80
Modifying a Piece's Appearance

Third in overall prominence among KidSim Computer Activities (17.7%) is

Modifying a Piece's Appearance. Once a piece is created, placed on the

gameboard, and given animation rules, it is often necessary to change the

piece's appearance in order to accomplish some particular interaction between

pieces (e.g., Jumbo Jets which crash into trees must change appearance into

Jumbo Jets which are on fire). Table 5 lists the distribution of Talk Units for

Modifying a Piece's Appearance.

Table 5
Percentage of Talk Units per Talk Category for
Modifying a Piece's Appearance (N= 800)

TALK SEQUENCE CODE % of Talk Units n

Help Talk (Researcher Initiated) 19.9 159


(Student Initiated) 24.6 197
Cumulative Talk 28.4 227
Query Talk 17.6 141
Disputational Talk 3.4 27
Solo Talk (Engaged) 4.3 34
(Disengaged) 1.3 10
Exploratory Talk 0.6 5

Modifying a Piece's Appearance is much like Rule Writing with respect

to task complexity. The highest number of Talk Units for Modifying a Piece's

Appearance was also found to be Help Talk at 44.5%. KidSim users who wish

to change a piece's appearance are required to use a series of interaction steps

which are unique to the program. These interactions must be learned before

pieces may be modified. Help Talk contributed greatly to this learning. To start

81
illustrating Modifying a Piece's Appearance, let us return to Anne and Jenny as

they are writing a rule to provide their elephant with "lunch" in the form of a

hamburger.

Help Talk during Modifying a Piece's Appearance (44.5%)

1210 A: Well, what I was saying is don't you have to go to


Appearance sometime?
1211 JD: Yeah, oh yeah...
1212 J: Do alt...SELECT ALL.
1213 JD: ...just so it's done before the rule's finished.
1214 J: COPY...
1215 JD: You can't fix rules once they're closed.
1216 J: ...PASTE.

Modifying a Piece's Appearance starts with a three step process which

creates a copy of the current appearance of a piece. These steps are mentioned

by Jenny in units 1212, 1214, & 12165. Student Initiated Help Talk by Anne (unit

1210) confirms her belief that appearances can only be changed while a new rule

is open for writing6. Without the researcher's confirmation at this point, the

students would be forced to try various alternatives while hoping to find the

appropriate sequence. Unfortunately, interaction time was limited. The

researcher felt it was best to assist students with their "worldbuilding"

rather than letting them become stalled in a process of trial and error.

During Modifying a Piece's Appearance, as with Rule Writing, the

researcher tries to help focus the student's actions. Researcher Initiated Help

Talk can be the "lens" that lets the students look at their work in a new way.

82
Anne and Jenny have a question about whether their burger should be devoured

in one or more bites. The researcher tries to help clarify their discussion (unit

1229).

1227 JD: Do what you want to do to it.


1228 J: But wouldn't you clear it if you want it to make it
disappear?
1229 JD: Oh yeah. Do you want to make it disappear
entirely?
1230 A: No.
1231 JD: It's up to you.

Jenny's response (unit 1228) to JD's neutral statement (unit 1227)

indicates that she is aware that clearing a piece will make it disappear. Jenny

does not see why this is a problem. JD, however, tries to hint at a course of

action which allows the piece to disappear more slowly (unit 1229). Anne

immediately grasps the essence of JD's suggestion (unit 1230). Anne and

Jenny, through Researcher Initiated Help Talk, have been given an alternative

way of proceeding. This may not have occured to them on their own in the

limited interaction time which they had.

Later in their interaction with KidSim, Anne and Jenny discover a problem:

the location on the Gameboard where their burger was just a second ago has

been replaced by the standard "blob of clay" (indicating a generic new

appearance). Researcher Initiated Help Talk is used to stop the work on the

computer for a moment and examine the situation (units 1240-1245). JD begins

this examination by having Jenny open the "lunch of elephants" appearance

change rule (unit 1240).

1240 JD: Now just a second. Click on lunch for elephants.


1241 J: I cleared that.
1242 JD: I know. I don't think it will let you have
something completely empty. It doesn't know what to
do about it. In other words, you can't have nothing.

83
1243 A: So we have to draw all our hamburger over
again?
1244 JD: No. Hit clear. Try pasting again. I think it's still
in memory.
1245 J: You're right.

Anne and Jenny were trying to do something completely new with their

"L.E." rule. By calling for the complete disappearance of a piece's appearance,

they had inadvertently stumbled into a condition for which KidSim does not seem

to be programmed. This is an interesting discovery. Making an object vanish

could be very useful in some simulations. In the real world one can think of

countless situations where objects might disappear from view (i.e. trash cans,

clouds, windows, etc.). Another way must be found to delete appearances. JD's

Help Talk continues with another suggestion (unit 1246).

1246 JD: What I would suggest...take a bite out of it.


1247 A: I'll go....and then clear up that whole side... like
that?, Is that what you mean?
1248 JD: It's up to you, but see, we can't make it
completely disappear because the computer won't
work with it.

Help Talk has allowed Anne and Jenny to continue to progress as they

add more complexity to their Zoo. Modifying a Piece's Appearance, an

advanced form of KidSim Rule Writing, required new information about the

program and new ways to integrate that information with what the children

already knew. Anne and Jenny did not yet possess enough of a base

understanding of KidSim in order to be able to program the complex task of

making a piece disappear on their own. Researcher Initiated Help Talk, applied

judiciously, served to guide the students as they constructed their world. Units

1355 to 1362 illustrate this guidance with an excerpt of Help Talk which suggests

84
that the students read through the rules in order to be able to understand what is

occurring.

1353 JD: What's going on?


1354 A: Nothing. You have to make a rule.
1355 JD: What should this rule say?
1356 A: This rule says they eat it whenever an elephant...
1357 J: No it just stops because, we need a I DON'T
SQUARE.
1358 A: No. You have to make a rule saying that the
elephant needs to eat it.
1359 JD: It took one bite out of it.
1360 A: Uhm hmm.
1361 J: Now it needs to take another one.
1362 JD: Or maybe this time it could finish it. I have
feeling there's a way to make that burger disappear.

Learning KidSim, in the collaborative fashion presented in this study, gave

all involved, students plus researcher, new perspectives on how the program

functioned. Help Talk was the conduit though which this

awareness was transmitted.

Cumulative Talk during Modifying a Piece's Appearance (28.4%)

We will now return to Amber and Marie's interaction to illustrate how

Cumulative Talk contributes to Modifying a Piece's Appearance. The primary

effect of Cumulative Talk seems to be a signal that "all is well" and both children

are "connected" to what they are doing on the computer screen. Amber and

Marie use Cumulative Talk in a particularly complementary way. One party,

Amber (unit 490) tries a color and Marie reinforces this choice (unit 491).

490 A: No, that's white, though. Yeah.


491 M: It looks pretty.
492 A: There. It's not blue.
493 M: That color is too deep.

Marie comments on Amber's next color selection giving an opinion why the

current color choice is "not blue" (unit 493). Each child is involved here. Both

85
are contributing to the interaction. The next several examples illustrate

sequences of Cumulative Talk which are very interconnected. These two best

friends actually begin to finish each other's sentences.

1798 M: Ummp, yeah there's a little...


1799 A: I know it..
1800 M: There's another direction... there's...
1801 A: Trash can.
1802 M: Ahmm...
1803 A: ...and then...

Amber and Marie are so "caught up" in their interaction with KidSim that they

spend little time verbalizing complete thoughts. Instead, the physical

collaboration they are accomplishing via on-screen visual demonstrations,

closely connects their Cumulative Talk.

1888 M: But then its tail will be... see, that's the end of its
tail right there.
1889 A: I know. It'll be up, and then it'll move down, up,
down, up, down, up, down, up.
1890 M: We're going to have to make more than one rule
'cause its going to have to be, like... it's got some of
its yellow in it. Just for it to move up and then move
down and then move up and then down...
1891 A: Okay.

The interactive environment of KidSim allows these student "programmers" to

test ideas in a fluid, complementary way. Amber and Marie's Cumulative Talk

builds idea upon idea. Marie begins this excerpt by pointing out the Giraffe's tail

(unit 1888). Amber complements by "verbalizing" the actions the tail will take

(unit 1889). Marie continues by mentioning that more than one new appearance

rule will be needed (unit 1890). Amber agrees (unit 1891). Each person is

contributing and eager to support the other in their joint goal. Cumulative Talk

during Modifying a Piece's Appearance is a sure sign that all is "going well".

86
Query Talk during Modifying a Piece's Appearance (17.6%)

Query Talk was ranked third in overall occurrence during Modifying a

Piece's Appearance. Defined as an overt request for information, Query Talk

was often used during Modifying a Piece's Appearance to make a

"confirmation request". Anne and Jenny illustrate this variety of Query Talk in

units 46-47 and 67-70.

46 J: Like that?
47 A: Yeah.

67 J: Keep the mouse button down.


68 A: About right here?
69 J: Um hmm..
70 A: I kept it down.

Query Talk is the best way to obtain missing information, even if the

question is anticipated by one's partner (units 28-29).

28 A: I know that, but how do...?


29 J: Okay, go to new piece I'll show you.

Queries may also signal that a negotiation for "mouse control" is taking

place. Modifying a Piece's Appearance, a special form of Rule Writing, is

fraught with doubt for these students. Gaining possession of the computer's

mouse, and thus KidSim itself, is the only way to directly affect the simulation's

outcome. Students who have direct control of the mouse may feel more

involved, less perplexed by the task of modifying appearances. Talk which gains

this control is usually in the form of a direct request.

83 A: You want me to do it?


84 J: (Laughs) Thank you.
85 A: I think I'll do it. Pick one. Got it now?
86 J: Yeah. (Whispers something)

87
Anne volunteers her services to Jenny when she seems to be struggling

with the mouse to select one of several overlapping pieces on the gameboard

(unit 83). Anne is using Query Talk to have a chance at "running the show".

Jenny, speaking in a sarcastic tone (unit 84), declines the offer, while Anne

seeks a confirmation of Jenny's actions (unit 85).

Mark and Cary also used Query Talk as a way to transfer control of the

mouse (units 1610-1615).

1610 C: Don't you want to do it?


1611 M: Draw?
1612 C: Yeah.
1613 M: Draw what?
1614 C: The airplane crashing?
1615 M: Okayyy.

Cary wants to voluntarily give control to Mark (unit 1610), but Mark does

not really know what he wants to do (unit 1613). Cary suggests to Mark that

drawing "the airplane crashing" is the next appearance change that should be

created (unit 1614). By asking Mark to be more involved in their collaboration,

Cary is perhaps hoping to stave off some of the boredom Mark exhibited earlier.

In the following two excerpts, Marie wants control of the mouse. At first,

her request came in the form of a query (unit 2317), followed by simple

statements (units 2319 and 2337), but when this subtle approach did not work,

she began to make a series of more direct queries (units 2337, 2339, and 2345).

2317 M: May I do something? Let me try something.


2318 A: No, I just wanted its feathers to move!
2319 M: Let me try something.
2320 A: I don't want it dead! (giggle)

2335 M: Hey, get, throw, time... Here, I have.. can I do


something real quick?
2336 A: (No.) Oh,...
2337 M: Let me try something.
2338 A: ...I still need one more, at the bottom.

88
2339 M: Can I try something Amber? I have an idea.
2345 M: Can I try something?
2346 A: Uhhh!
2347 M: Thank you.
2348 A: Forget it.

Marie finally succeeds in transferring control of the computer (unit 2347).

Query Talk allowed her to directly approach the goal she was chasing. Marie's

show of patience to simply re-ask the same question paid off for her at last.

Solo Talk during Modifying a Piece's Appearance (5.6%)

Solo Talk is defined as un-replied statements or statements to which no

response is given regardless of whether a response might seem appropriate.

Solo Talk which occurred as pieces were modified often consisted of instances

where one member of the collaboration pair made a comment which served as

"verbal coaching" for their partner. Jenny illustrates this use of Solo Talk in units

71 and 74.

71 J: You just have to go to OPPS.


72 J: Wait, don't start yet. Now get it where you want it.
73 A: Um hmm
74 J: Keep going.
75 A: (Long silence) Got it down?

These remarks were coded as Engaged Solo Talk. Engaged Solo Talk is

Solo Talk about the current computer activity. Jenny is urging Anne to "keep

going", to "keep moving forward" with their new task of piece appearance

modification. Anne later directs the same style of remark towards Jenny (unit

88) and Jenny responds with a Solo Talk remark of her own (unit 89).

87 A: Just leave it.


88 J: That part's going to be crooked. That part's going
not goin' to be totally straight. That looks good
enough.
89 A: You ready?

89
Engaged Solo Talk also took the form of a series of helpful "cues" or

"directions" for the partner whose turn it was to control the mouse as they

maneuvered pieces around the gameboard. Marie (unit 1896) and Cary (373,

715 and 716) each exhibited this use of Solo Talk.

1896 M: Why don't you take the eraser and erase that little part
right there? I don't like that.

375 C: Here do this. Go to the... Try to get on the very, very


edge. Go up on the top. No, over more, that way. Down
until it touches the square.

715 C: Try to make that one just the same too. Yep, just leave it
like that. Move that balloon and change its appearance. Try
that.

716 C: Another appearance. Now push play. Hmm. Move the


plane up a little bit, move that thing down.

Cary, in particular, used Solo Talk during Modifying a Piece's

Appearance, to give guidance to his partner. Solo Talk gave Cary a way to stay

engaged in the computer activity while he was not actually controlling the mouse.

Solo Talk also gave these collaborators a way to make "off topic" or disengaged

remarks. Disengaged Solo Talk is Solo Talk which is not pertinent to the current

computer activity. Note these solo comments from Mark which he made as he

gazed at the clock (unit 267) or simply "off into space" (units 452).

267 M: One after three.

452 M: Uh oh, spaghettos.

Mark, in these excerpts, is not immediately involved with his partner or the

present computer activity. Mark seemed to be little interested in piece modifying

or Rule Writing in general. The complexity of these tasks may have been the

90
reason. Most of Mark's Disengaged Solo Talk occurred during Computer

Activities which were less connected to the simple task of drawing or in which

drawing only played a minor role. Mark may have not felt comfortable enough

with Modifying a Piece's Appearance to become more than an observer for the

majority of the interaction time.

Solo Talk during Modifying a New Piece gave engaged partners KidSim

interaction input regardless of who was actually controlling the mouse.

Somewhat ironically, Solo Talk also gave disengaged partners a means to

express themselves externally to the activity at hand. One conclusion that might

be drawn from this observation is that the greater the complexity of a current

activity, the higher the probability that Solo Talk will be present. Each individual's

reaction to the challenges presented to them determines whether their Solo Talk

serves a supporting or a distracting role.

Disputational Talk during Modifying a Piece's Appearance (3.4%)

Occurring for only 27 units out of Modifying a Piece's Appearance's total

of 800, Disputational Talk continued to influence each collaboration in much the

same way as it did in the previous activities of Creating A New Piece and Rule

Writing. Disputational Talk is found at moments when a disagreement is

occurring, or more precisely, at moments when a disagreement is present in

which one or both partners neglect to give any reasons for their statements.

Let us return to Mark and Cary for an example of Disputational Talk during

Modifying a Piece's Appearance.

362 C: Now draw. Make the balloon look like it just got
hit by an airplane.

91
363 M: (whining) How am I going to do that? That is like,
impossib....
364 C: You've gotta draw it.
365 M: You draw it. You can draw, ten times better.

Cary is "ordering" Mark to draw the airplane's appearance change (units 362 &

364). Mark would rather have Cary do it instead (unit 365). Cary may think he is

doing his best friend a favor by encouraging him to do some drawing, an activity

he knows Mark likes. However, Mark does not give Cary any reason why he

must create the plane's next appearance. Cary does not understand that

Modifying a Piece's Appearance to make it look like a plane that is crashing is a

rather difficult thing to do.

366 C: Oh. (to JD) Don't we go to clear?


367 JD: Well, do you want to start from scratch?
368 C: Yeah, don't you wantta start from scratch?
369 M: I don't care, as long as you draw it...
370 C: You gotta draw..
371 M: ...all you need is appearances.
372 C: You gotta draw one.. Wait you missed it, if you
aimed it at all.

JD is attempting, through Help Talk (unit 367), to answer a question that

Cary has posed (unit 366). JD wants to have the boys think about whether or not

they really want to start from scratch at this point. It would be much easier for

them to modify a piece's current appearance rather than create one from scratch.

Mark may have shown some understanding of this procedure in unit 371 when

he said, "all you need is appearances". Yet, Cary continues to be disputational

by insisting that Mark has "gotta draw " (units 370, 372). Mark eventually gives in

to Cary's demand and does begin to make the next appearance of the airplane.

Disputational Talk has again stalled the collaborative process. Cary's

insistence that Mark draw at least one new appearance made Mark feel very

uncomfortable (see unit 363). Mark did not become as proficient with KidSim as

92
did Cary. By forcing Mark to draw, instead of inviting him to, Cary's Disputational

Talk may have become an obstacle to their later communication. Many of Mark's

disengaged Solo Talk remarks came after their disagreement over who should

draw the crashing "Jumbo Jet".

Exploratory Talk during Modifying a Piece's Appearance (0.6%)

Exploratory Talk amounted to less than one percent of total talk units

during Modifying a Piece's Appearance. Here are two examples of Exploratory

Talk from Cary as he gives some advice to Mark (units 392 and 396).

392 C: Make like, little lines. Yeah. Like, just make like, I
don't know what that thing does. Oh yeah, it cuts and
you can move it.
393 M: Huh, where did it go?
394 C: What's that?
395 M: It disappeared already.
396 C: You don't want that - that cuts out things. You
don't want to cut it out. Try that pencil, make little
lines in the air. Like air's coming out of it.
397 M: Whoa, swishhh...
398 C: Make it come out of the middle of the hole too.
399 M: Swishhh...
400 M: Yeah!

These examples are centered on Cary trying to explain to Mark how various

KidSim tools work (see Appendix A, Figure A2). Cary is going beyond giving

simple commands to Mark (see Solo Talk units 375, 715 and 716, p.90). Cary's

"thinking out loud" through Exploratory Talk is helping to teach Mark to identify

each tool and how each is used.

This is a classic example of Vygotsky's Zone of Proximate Development at

work. Here the more capable peer (Cary) is helping his partner (Mark)

understand something which is only a little beyond what Mark already knows.

Cary is pulling Mark into the sphere of knowledge which he (Cary) already is

93
comfortable. Perhaps Cary's earlier insistence that Mark "now do the drawing"

actually helped Mark to comprehend the software a little better by giving Mark a

chance to control the mouse more often.

Summary of Modifying A Piece's Appearance

Modifying A Piece's Appearance, like Rule Writing, proved to be a

difficult skill for these students to master, albeit to a less complex degree than the

latter. Some understanding for this finding may be found in the relationship

between Modifying A Piece's Appearance and Rule Writing. When a piece's

appearance is modified, at least two actions must occur which encompass the

appearance change. First of all, a new rule must be opened which specifies

"when" an appearance changes. For example, "the giraffe's tail moves up after

every other forward step." This first portion of piece modification was familiar to

the students as they had already learned about Rule Writing. Secondly, in order

for this change to complement a piece's current appearance, the original

appearance must be copied before it can be modified. This second concept, in

particular, was a difficult one for these children to learn.

All three pairs of collaborators struggled to remember the ordering of steps

necessary to copy a piece's original appearance so that it could be modified.

The Help Talk (44.5% of total talk units), which occurred early in Modifying A

Piece's Appearance, may be partially attributed to a reinforcement of these

copying steps. As found while Rule Writing, a period of retraining/reinforcing

Help Talk was necessary before piece appearance changes could proceed.

Anne and Jenny only applied appearance changes to two of their pieces,

yet they were also the one's who created the sophisticated change of a burger

94
with multiple bites taken out of it. Their pattern of Cumulative Talk mixed with

Help Talk, acquired during Rule Writing, allowed them to tackle the difficult

challenge of making a piece disappear. Mark and Cary, in contrast, had a high

number of appearance changes. Cary, in particular seemed up to the challenge,

using lots of Student Initiated Help Talk to ask many questions. Mark, by this

point in their interaction, may have felt a little overwhelmed and tried to avoid

creating new appearances altogether. Amber and Marie created only three

pieces in the end, but two of these incorporated appearance changes. Their talk

included some Exploratory remarks from Marie and a high level of productive

Query mixed with Cumulative Talk. Amber and Marie may have achieved the

greatest sustained period of collaborative progress of any of these three pairs at

this later point in their overall collaboration. Their steady work interval came as

they experimented and finally understood how to make the giraffe's tail move up

and down while it was simultaneously stepping forward behind a mountain range

[see Appendix A for details on Amber and Marie's "Animal World"].

Modifying A Piece's Appearance, while less daunting than Rule

Writing, was still a complex process for these pairs of fifth graders to master. A

Help Talk majority percentage of 44.5% suggests that much in the way of

guidance or additional training was needed as pieces were modified. As the

basics of appearance modifying was understood, a cycle of Cumulative Talk

(28.4%) mixed with Query (17.6%) did prevail. Solo, Disputational, and

Exploratory Talk were only minor in occurrence during appearance modification

at less than ten percent total across all three talk types.

It is interesting to note that Query Talk during Modifying A Piece's

Appearance occurred at a level that was over 6% higher than that encountered

95
during Rule Writing. Perhaps this is due to the greater familiarity that the

students had with writing rules. They were able to amass a greater quantity of

questions concerning how rules operated once Rule Writing became less of a

foreign activity to them.

When they first began writing rules, these students may have been too

overwhelmed to have a sufficient level of understanding in order to be able to ask

many questions. Amber and Marie, in particular, seemed to be more comfortable

with rules as their interaction with KidSim progressed. Note how much more

confident they "sound" later in the interaction as Amber and Marie quickly

answering each other's queries (units 2131 to 2139).

2131 M: We'll get a new rule. What should we make her


do?
2132 A: Ahm...
2133 A: Oh, I got it!, I got it!, I got it!, I got it!, I got it!
2134 M: Then, tell me!, tell me!, tell me!
2135 A: Go to APPEARANCES. This is cool. Okay, see
watch...
2136 M: What?
2137 A: ...go to APPEARANCES. Okay, now, we're going
to make her feathers go up and down, up and down,
up and down...
2138 M: Her feather?
2139 A: Feathers. Go to NEW.

Familiarity formed a foundation which supported Amber and Marie's eagerness to

experiment with appearances.

Running the Clock Forward

The fourth of KidSim's four primary Computer Activities (13.9% of Total

Talk Units), Running the Clock Forward, is reviewed next. Running the Clock

Forward is probably the most enjoyable computer activity for the students. After

96
a piece in a KidSim world is created and given rules, the game clock must be

activated. Running KidSim's game clock is actually playing the simulation itself.

Table 6 lists the distribution of Talk Units for Running the Clock Forward.

Table 6
Percentage of Talk Units per Talk Category for Running the Clock Forward
(N= 628)

TALK SEQUENCE CODE % of Talk Units n

Help Talk (Researcher Initiated) 29.8 187


(Student Initiated) 18.2 114
Cumulative Talk 34.6 217
Query Talk 9.7 61
Disputational Talk 0.0 0
Solo Talk (Engaged) 4.5 28
(Disengaged) 1.8 11
Exploratory Talk 1.6 10

Running the game clock is the "big payoff" for all of the intricate work of

creating pieces and writing rules that went on before. Only by allowing the

simulation to run can one see if his or her actions worked out as planned. The

game clock is activated forward in time by either clicking on its green "forward to

the right" VCR-style button or on the clock face itself (see Figure 6).

97
Figure 6
The Game Clock
(located in the lower right portion of the Main Interface- see Appendix A, Figure 1A)

Help Talk during Running the Clock Forward (48.0%)

As was the case with Rule Writing and Modifying a Piece's

Appearance, Running the Clock Forward was found to have the highest

percentage of Help Talk Units (48.0%). One explanation for this finding may be

that Running the Clock Forward was a time of ebullience, so enticing, that the

researcher was often asked, or volunteered, to comment about how the

microworld was progressing. The researcher sometimes took these

opportunities to mention a problem with the current "state of the world" that might

need attention (unit 1051).

We now return to Anne and Jenny as they play, "Anne and Jenny's Zoo."
1049 J: Yeah! That looks cool now!
1050 A: Yeah, I like that!
1051 JD: But they're not facing the right way, are they
always?
1052 J: No.
1053 JD: How about writing a rule then, making them face
the right direction?
1054 A: You get to do this.
1055 J: Okay.

Here the two collaborators are sharing their excitement about the way the

microworld is running (units 1049-1050). JD points out a problem with the

98
appearance of one of the Zoo's animals (unit 1051): the elephants are not

necessarily facing the direction they are moving. Anne and Jenny had been

working on this one for awhile. More Help Talk (unit 1053), gives the students a

way of solving their predicament. Help Talk allowed the researcher (JD) to again

remind the students that writing a new rule was a good way to solve problems in

KidSim.

In the next sequence, Amber and Marie are having a problem with a piece

which stops moving at an unplanned moment. Help Talk gives the researcher a

way to encourage the children to examine the situation for possible solutions.

765 JD: Why is it doing that?


766 A: Because it can't do it.
767 M: It falls down because we didn't tell it . It could
come back up.
768 JD: No, it, KidSim works in a circle. So it would come
back up, but what is going on?

Amber and Marie have a piece ("Bob", the rock) which should cycle from

the top to the bottom of the screen and then "bounce" off giraffes at the top to

reenter again at the bottom. Instead, "Bob" leaves the bottom of the screen and

just stops at the top. As everyone is watching "Animal Life" run, JD focuses the

students attention on the problem (unit 765). Marie proposes a solution which

includes the idea that "we didn't tell it. It could come back up." (unit 767),

implying that no rule was yet written to cover this situation. While a good guess

on her part this was not the correct answer in this case. Marie has learned at this

point in the process that writing a new rule is often the best way to resolve a

KidSim interaction problem. As JD points out in his response, if "Bob" could be

made to go off the top of the screen, "it would come back up". JD again asks for

the students to think about what is going on in this situation (unit 768). JD is

99
asking them to focus their attention, to look at the entire problem and see what

else might be occurring on the screen.

769 A: Red?
770 JD: It's red because it can't fire anymore. It's telling
you: "I can't work this rule anymore." We don't have a
rule that describes what's going on on the screen, do
we? Look where the rocks, what are you calling this,
Bob?... Look where Bob is right now.
771 M: He's in two places, between the two...
772 JD: Okay, one problem with Bob is...
773 M: He's between two...
774 A: He'll go, he'll get the duck... oh that doesn't have a
rule yet, though.
Amber realizes that one rule isn't firing (unit 769), as shown by the "Red"

indicator light in the Brain Window (see Appendix A). Through Help Talk (unit

770), JD asks the students to notice where "Bob" is currently. JD starts to reveal

the complete nature of the impasse in unit 772, when Marie interrupts (unit 773)

and then Amber interrupts her (unit 774) with the correct solution that a rule is

missing. This sequence of interruption followed by interruption was very

characteristic of Amber and Marie whenever they were in a period of great

discovery. Amber's excitement in particular would simply be too overwhelming

for her and she would just have to let her "thoughts be known". JD continues his

probe of the current problem by confirming Marie's discovery that a rule was

missing (unit 775).

775 JD: So are there any rules that have "Bobs" and
"Swans" in them?
776 M: No, so we got to make one?
777 JD: And what would be the other problem?
778 M: That he couldn't go past this.
779 JD: Right. And he is looking so far ahead of where he
is right now, that he is already running into the
giraffes. In other words, he goes off the bottom of the
screen, but he can't come back on because he is
looking out 2 or 3 squares in front, and all he sees are
giraffes, so he says: "Forget it." So write rules.

100
Marie now realizes why "Bob" was not acting in an expected way: a rule
which includes "Bobs" and "Giraffes" had not been written yet (units 778 & 779).
Help Talk served as a way to scrutinize problems by focusing the students'
attention on each of several particular errors.

Cumulative Talk during Running the Clock Forward (34.6%)

The second highest percentage of talk units during Running the Clock

Forward was for Cumulative Talk (34.6%). Cumulative Talk has been defined as

a period of steady progress in which each speaker contributes to the discussion

by building upon a previous speakers utterance. A high percentage of

Cumulative Talk is not surprising, giving the nature of Running the Clock

Forward as a computer activity. Running the Clock Forward was largely a

time for watching and therefore encouraged the collaborators to "trade"

observations about how well their "worlds" were performing. The following

example from Amber and Marie illustrates this kind of "give and take".

754 M: Let's just see what it looks like.


755 A: They sink too fast.
756 M: They do sink too fast. I want to see something.
757 A: Shhh drop...
758 M: I know it's not...
759 A: See if the other one is still down below.
760 M: Okay,...
761 A: (makes sound- ahh haa!)
762 M: ...It, it like, disappears,...
763 A: (makes sound- ahh ahh!)
764 M: ...because it can't, it didn't come back up, it...

Marie starts their world "running" and Amber notes that "Bob" the rock is

sinking too fast (units 754 - 755). Marie confirms Amber's observation and starts

to investigate why this is happening. Cumulative Talk allows them to reflect

observations off of each other to further their collaborative effort. Amber begins

to say that the rock should drop slowly (unit 757) when Marie interrupts her.

101
Marie then is interrupted by Amber (unit 759) giving a suggestion as to how to

proceed. Amber's dual "ahh haas" (units 761 & 763) show her excitement with

their mutual discovery of the rock's disappearance. Their conversation continues

with the Help Talk discussed on page 99.

Mark and Cary also exhibited "interrupted speech" during Cumulative Talk

while Running the Clock Forward. In the following sequence, Cary has just

started the clock moving.

1163 C: ...and then play. Let's watch the planes hit. Pop!
1164 M: Ha, Ha, it worked!
1165 C: Now we gotta have the planes go down into the
tower and....
1166 C: This one hits...
1167 M: ...needs to hit the tree and kafummph
1168 C: ...and then this one will hit the tree and this one
will hit the tower.
1169 M: And then we want those branches to fall down
and that one run into the tower.
1170 C: That would be kinda hard.
1171 M: Yes, oh it could happen.

The boys are obviously enjoying watching their creation at work.

Cumulative Talk allows them to express their current satisfaction as well as

providing them an outlet to channel ideas for the future. As Cary begins to

explain his idea for one of the plane's movements to Mark (unit 1166), Mark

interrupts by completing Cary's thought. The shared collaborative space of the

computer screen allows Mark to anticipate Cary's words by noticing to what

object Cary is pointing. Cary "returns the favor" by interrupting Mark in his next

turn (unit 1168). Mark adds to Cary's last remark as he embellishes it (unit

1169). Cary responds with a judgment about the difficulty of implementing

Mark's last suggestions (unit 1170). Mark responds positively, "...it could

happen". The last four lines of this excerpt, in particular, exhibit the cooperative

102
give and take associated with Cumulative Talk. One collaborator makes a

remark about the way their world is progressing and the other takes this thought

and follows with a complimentary or continuing remark. Each collaborator adds

to the mix of observations and suggestions necessary for mutual progress.

Query Talk during Running the Clock Forward (9.7%)

A popular question the students often asked of each other while Running

the Clock Forward often took the form "What shall we do next?" This is a

logical development considering that the moments spent running KidSim's clock

are often ones of reflection. These are times when the intense concentration

needed to build a microworld is replaced by the simple pleasure of watching

one's characters interacting together on the screen. The following excerpts

illustrate the persuasiveness of this open ended question.

2253 M: Now we got to have... Now what are we


gonna do? 'Cause we've got it to do everything we
wanted to.
2254 A: I want her feathers to go up and down.
2255 M: Oh, okay.
2256 A: So, can I do her feathers?
2257 M: Well, you got to do his tail, too.
2258 A: Part of it?
2259 M: No, all of it.
2260 A: Ahh... Okay... CUT, EDIT, COPY...

Marie takes a moment while the clock is running to ask Amber "Now what

are we gonna do? " (unit 2253). Amber replies that she wants to add "feather

movement" to the tail of their duck (unit 2255). Marie agrees, but also notes that

the tail should be made to move also (unit 2257). Amber asks if only part of the

tail should be made to move (unit 2258). Marie responds answering that all of

the tail should be made to move (unit 2259). Amber after a brief pause, accepts

103
Marie's suggestion (unit 2260).

Through this process of open-ended question, suggestion, and

confirmation the collaborators are able to negotiate the elements of their

developing world. Marie seeks out Amber's input , responds to any suggestion

offered and then continues by either passing or taking control of the computer

(via the mouse). Amber begins to build along this new path (unit 2256), once the

future course of a pair's construction has been determined.

The period of relaxed observation which accompanies Running the Clock

Forward allows each student a chance to think about future additions to their

KidSim world. Indeed, another of the sixteen Computer Activities identified for

this study, Planning the Microworld often occurs at the same time as the game

clock is running. Query Talk during Running the Clock Forward gives the

collaborators an explicit way to inquire about future plans. Mark and Cary

illustrate another example of this phenomenon.

1462 Cary: Whoa. Okay, lets go ta... Now what do you


want to do?
1463 M: About anything else we could draw?
1464 C: Yeah, we gotta draw the plane crashing. But
what's it going to be crashing in to?
1465 M: How are we going to draw that? We got to make
a rule.
1466 C: What is he gonna to crash into first?
1467 M: See, ahmm, like that plane would crash into that
tree, and then that plane would crash....
1468 C: But we don't know which way, whenever it's
gonna hit because it goes down in so many places.
1469 M: Well, do you have an idea of what to draw?
1470 C: A plane crashing into a tree. That way, one of
them will end up crashing into one, so... You're right,
it's tricky.

Cary begins this excerpt by asking his friend, Mark, about what they should do

next (unit 1462). Mark responds to Cary's question with a question when he asks

104
what else remains to be drawn (unit 1463). Cary suggests that they draw "the

plane crashing", but also follows-up with another question about what object the

plane should be hitting as it crashes (unit 1464).

This selection of Query Talk during Running the Clock Forward highlights

Cary and Marks' struggle to create a new appearance for their "Jumbo Jet". They

are able to visualize some likely problems that might occur when they add a new

"behavior" (the plane crashing) to their world. As they watch their world "run",

they have a chance to start thinking about just how many different objects the

crashing plane could hit (units 1466-1468). Cary and Mark do not realize at this

point that the appearance of the crashing plane itself could be made the same for

any object that it hit. At the same time, they also understand that the appearance

of any object that the plane might hit would also need to be changed (unit 1470).

Query Talk lets them directly probe into the many possibilities raised through

Cary's simple suggestion to "draw the plane crashing".

Query Talk, by definition an exchange of collaborator questions and

responses to each other, has been shown to be a useful resource for these

students while Running the Clock Forward. Through their queries they are able

to investigate any problems with their "worlds" as they now exist or start making

plans for the future. The visual reference of a running microworld gives these

collaborators a mutual environment where they can see and speculate about how

well their creations are performing.

Solo Talk during Running the Clock Forward (6.3%)

Solo Talk appears in approximately the same proportion (6.1% of total talk

units) while Running the Clock Forward as it did during Creating A New Piece

105
(6.1%) and Modifying a Piece's Appearance (6.0%). The tendency for a

collaborator to employ Solo Talk may be more of a function of individual

differences, and less the consequence of a current computer activity. Mark and

Cary's comments well illustrate this notion as they accumulated a total of thirty

three Solo Talk units during Running the Clock Forward compared to the

combined total of only seven units for the two remaining pairs. The following

examples illustrate Solo Talk that took place while the KidSim game clock was

running forward. The first excerpts of Solo Talk focus on Cary.

711 C: Hit play. Move it down more, yeah right there.


Try that.
811 C: Got to go back to appearance.
814 C: We didn't stop the clock.
832 C: Start the clock.
1547 C: It's malfunctioning. We've gotta stop the clock.
Stop the clock. Okay, I gotta make something.
There, it's stopped. Go to NEW RULE.
Click on RULES. Click on PROPERTIES. "X" it.
Now click on (whispers...).
1416 C: Too bad we can't get circles.
1933 C: We don't want the tree.
2011 C: I can't get it.
2014 C: I can't grab him.

Cary is employing Engaged Solo Talk in these excerpts. In each case he

is making a comment about some aspect of the current computer activity (i.e.,

Running the Clock Forward). Cary (unit 711) refers to starting the game clock,

"hit play", followed by a suggestion to Mark to move one of the game pieces as

the clock is running. Mark and Cary were alone in employing this "technique" of

moving pieces by hand as the game clock was running. This was their way of

trying to control where the pieces were on the gameboard. Mark does not

respond verbally to Cary's suggestion, but merely complies by moving the piece.

106
Solo Talk gives Cary a way to voice his observations about the state of

their microworld. Sometimes these remarks are examples of simply "thinking

out loud" (units 811, 814, 1416, 1933, 2011, and 2014) and at other times they

appear to be "commands" directed at Mark (units 711, 832, or 1547). In either

case, Cary's remarks are always pertinent to the current on-screen activity. For

example, as Cary is watching "Fantasy World" run he is wishing out loud that

there was a way to make the planes fly in a circle (unit 1416). Mark, on the other

hand, often used Solo Talk to disengage from interacting with KidSim and Cary.

Mark's use of Disengaged Solo Talk, as reported in previous sections of

this chapter, served as a distraction for him while Cary continued to work with

KidSim. The following excerpts are typical of Mark's Disengaged Solo Talk which

take place during Running the Clock Forward

562 M: (yawning) I don't care.


1011 M: (makes noise = OCK!)
1014 M: (makes noise = OCK!)
2015 M: (singing) I want my "I Don't Care Squares", I want
my "I Don't Care Squares"...

In each of these examples, Solo Talk serves as a form of escape for Mark,

escape away from interacting with Cary and KidSim. Mark begins with a yawn

and an overt comment about his boredom (unit 562) , continues later by making

loud sounds (units 1011 and 1014), and finally chooses to begin singing (unit

2015). Each of these comments illustrates how "disconnected" from the

collaborative effort Mark sometimes became. However, this was not always the

reason for Mark's use of Solo Talk.

Sometimes Mark, much like his friend Cary, made Solo Talk statements to

simply express his thoughts about actions on the computer screen. In the next

set of excerpts (units 1245, 1438, 1447, and 1449), Mark is watching the "Jumbo

107
Jet" move on the screen while commenting about its proximity to the Hot Air

Balloon (with whom he would like to see the jet have a collision).

1245 M: Okay, airplane move...


1438 M: Move the plane back up.
1447 M: Oh! No! It just missed it! Oh, no, just missed it!
No! No!
1449 M: Pop, oh, just missed it again.

Solo Talk during Running the Clock Forward gave engaged partners a

way to comment upon their microworld's progress. Solo Talk for engaged

partners allowed them to express their observations out loud in perhaps an

attempt to communicate these thoughts or desires to their partner. Otherwise,

Solo Talk during Running the Clock Forward for a disengaged partner again

served primarily as a form of distraction. The greatest utilizer of Disengaged

Solo Talk across all three pairs of students and Computer Activities was Mark.

Of the thirty five total Disengaged Solo Talk units coded throughout this study,

Mark was responsible for thirty of these. Solo Talk continues to serve a

supporting or a distracting role depending upon the level of engagement the

speaker is at that moment experiencing.

Exploratory Talk during Running the Clock Forward (1.6%)

Exploratory Talk amounted to less than two percent of total talk units

during Running the Clock Forward. The first example of Exploratory Talk

comes from Marie (unit 1993).

1992 A: It went up and stayed up.


1993 M: Yeah. Because we didn't tell it to go back down.
We just told it... from the tail up or from... See, all the
tails went up and we didn't tell it to go back down and
come up and go back... does that mean
we'll have to make a rule for each wag?
1994 JD: I think you only have to tell it... try it.

108
As they watch their world "run", Marie is trying to explain to Amber why the

tails of their giraffe "stayed up". Exploratory Talk gives Marie a way to "teach"

Amber why things are occurring the way they see them happening on the screen.

Cary also uses Exploratory Talk to teach his partner, Mark (unit 1890).

1888 C: Uh oh, this thing stopped. (to JD) How come it


won't catch on fire?
1889 M: Because we ain't got fire out there.
1890 C: I don't think it will fly with fire, I'm mean just
saying that, because we haven't made a rule for it to
do anything. We haven't made a rule for it to go
down. We've got to make a rule for it whenever it hits
something and turns on fire.
1891 JD: You're in charge.

Here Cary "explores" why the plane is not behaving the way he and Mark

expect it to. Cary understands the problem, and relates it clearly to Mark when

he says: "We've got to make a rule for it whenever it hits something and turns on

fire." Cary, at this point in their interaction, understands the importance of

rulemaking. Mark, on the other hand, really does not seem to understand how

KidSim pieces may be made to change appearance as he states that fire must

be "out there" on the gameboard for a piece to start burning (unit 1889).

Exploratory Talk's direct connection to Vygotsky's Zone of Proximate

Development is seen once more. Here the more capable peers (Marie and Cary)

are helping their partners (Amber and Mark) understand actions which appear as

their worlds are "running". Marie and Cary are attempting to pull Amber and

Mark into the sphere of knowledge which they already inhabit.

Summary of Running the Clock Forward

Running the Clock Forward, unlike any of the other three primary

Computer Activities (i.e., Creating A New Piece, Rule Writing or Modifying a

109
Piece's Appearance), was chiefly a time of reflection, a time to absorb how

one's microworld was progressing. Talk during this period was dominated by a

combination of Help Talk (48.0 %) and Cumulative Talk (34.6%). As the students

watched their worlds "move", many situations occurred where they requested or

received comments from the researcher about what was happening. Otherwise,

they were often complimenting each other on their progress through Cumulative

Talk. If a question about a particular situation did occur, it was settled without

dispute as no Disputational Talk was recorded while the simulations were "in

action".

Running the Clock Forward was the "reward" for all of the work which

had gone on before. These students enjoyed seeing their worlds in action. And

no wonder, they built some rather impressive places as shown in Figure 3.

110
Figure 3

"ANNE and JENNY's ZOO" "FANTASY WORLD" by Mark & Cary

"ANIMAL LIFE" by Amber & Marie

111
Summary of Primary Computer Activities

A detailed review of "peer programmers' " interactions with KidSim has

been presented. Four primary Computer Activities encountered when using the

KidSim program were isolated and categorized: (1) Creating a New Piece, (2)

Rule Writing, (3) Modifying a Piece's Appearance, and (4) Running the

Clock Forward. The three pairs of collaborating KidSim users spent 77.1% of

their Total Talk Units4 engaged in one of these four primary activities (see Table

2, p. 36). Each one of the four will now be summarized.

(1) Creating a New Piece. The greatest amount of talk during Creating a

New Piece was Cumulative Talk at 38.4%. This signaled a period of calm

progress in which each collaborator was aware of the current activity and

contributed positively to it. Query Talk (32.1%) was the second most prevalent

kind of talk during Creating a New Piece. Query Talk functioned both as a plea

for confirmation as well as served as a subtle way of voicing disagreement.

Disputational Talk (8.1% of Creating a New Piece talk) proved to be

counterproductive. After the "grass incident," later Disputational Talk was

squelched early, after only 2 or 3 turns. Disputational Talk places a barrier to

further communication whereas Exploratory Talk, while occurring only 3% of the

time in Creating a New Piece, served as an invitation to more discussion.

Exploratory Talk also played the role as a kind of peacemaker, bringing in "facts"

or "visual demonstrations" to help settle an argument and continue the drawing.

Solo Talk (5.7% of Creating a New Piece talk) contributed little to Anne and

Jenny's collaboration and mostly served as a form of self-narrative by Anne and

Mark. Solo Talk for Cary and Marie gave them a way to have input into the

creative process at moments when they were not controlling the mouse. Finally,

112
Help Talk (12.9%) was not as prevalent during Creating a New Piece as it

would be during later Computer Activities. The act of drawing is a diversion

which most children are very familiar with by the time they are in the fifth grade.

The collaborative setting of this study gave them a chance to draw together,

interactively, with the need to learn only a few simple tools. The necessity for

Help Talk was lessened as a result.

(2) Rule Writing. Help Talk (61.6%) accounted for the overwhelming

majority of the communication recorded as these students collaborated to write

KidSim interaction rules. The justification for this finding may be attributed to at

least three factors: program complexity, student ability, and training

effectiveness.

KidSim is a deceptively easy simulation program which only looks simple

on its surface. Underneath is a powerful visual programming language which

requires the use of layers of "graphical rewrite rules". These rules translate on-

screen programming actions into game piece interactions. To be able to build a

simulation in KidSim with any degree of sophistication, this language must be

understood and utilized efficiently.

Anne and Jenny employed little movement in their "Zoo" and thus had less

need to master Rule Writing. Mark and Cary plus Amber and Marie, on the

other hand, did introduce lots of action into their worlds, and therefore needed to

understand Rule Writing as early as possible. Cary, Amber, and particularly

Marie, did seem to be up to the task. Mark, however, was more easily distracted

and tired of Rule Writing long before the others did. Rule Writing is not an easy

skill to acquire, but as three of these six students demonstrated, it was not too

113
difficult for fifth graders, given the appropriate time, training and motivation.

Rule Writing proved to be a complicated process for these students. Help

Talk became a necessary "support system" for the KidSim collaborators. The

students certainly would have failed to produce functioning microworlds without

frequent suggestions and explanations from the researcher during Rule Writing.

Natural curiosity and intuition may have led them to discover some of the Rule

Writing features of KidSim. However, the overall complexity of the task,

combined with the limited amount of training time, created a climate which gave

rise to large amounts of Help Talk as rules were initially written.

Compared to Help Talk, other types of talk contributed in a much less

prominent way to Rule Writing. A cycle of Cumulative Talk (21.9%) mixed with

Query (11.2%) did prevail once basic rule creation was understood. Solo Talk at

3.3% came next in frequency, due largely to Mark and Cary's input. One of the

most influential types of talk found in Creating a New Piece, Disputational Talk,

was present only 1.9% of the time. Rule Writing is a new, perhaps unique,

activity to these students. They expended most of their energy trying to learn the

process, with little time left over to dispute what they were trying to do or what

they had accomplished.

(3) Modifying a Piece's Appearance. While less daunting than Rule

Writing, Modifying A Piece's Appearance, was still a complex process for

these pairs of fifth graders to master. Some understanding for this finding may be

found in the relationship between Modifying A Piece's Appearance and Rule

Writing. When a piece's appearance is modified, at least two actions must occur

which encompass the appearance change. First of all, a new rule must be

opened which specifies "when" an appearance changes. Secondly, in order for

114
this change to complement a piece's current appearance, the original

appearance must be copied before it can be modified. This second concept, in

particular, was a difficult one for these children to learn. A Help Talk majority

percentage of 44.5% suggests that much in the way of guidance or additional

training was needed as pieces were modified.

As the basics of appearance modifying was understood, a cycle of

Cumulative Talk (28.4%) mixed with Query (17.6%) did prevail. Solo,

Disputational, and Exploratory Talk were only minor in occurrence during

appearance modification at less than ten percent total across all three talk types.

(4) Running the Clock Forward. Unlike any of the other three primary

Computer Activities, Running the Clock Forward, was chiefly a time of

reflection, a time to absorb how one's microworld was progressing. Talk during

this period was dominated by a combination of Help Talk (48.0 %) and

Cumulative Talk (34.6%). As the students watched their worlds "move," many

situations occurred where they requested or received comments from the

researcher about what was happening. Otherwise, they often complemented

each other on their progress through Cumulative Talk. If a question about a

particular situation did occur, it was settled without dispute as no Disputational

Talk was recorded while the simulations were "in action."

Overview of Transitional Computer Activities

Although less time consuming than the four primary Computer Activities,

and usually serving as a transition between them, is a group of twelve

Transitional Computer Activities. (1) Planning the Microworld, which did not have

a distinct period human-computer interaction of its own, often took place while

115
Running the Clock Forward or prior to each day's actual interaction period. The

eleven others, while they are not necessarily less important, did occur much less

frequently. Four of them are quite utilitarian in nature: (2) Setting a Piece on the

Gameboard, (3) Modifying a Piece's Position, (4) Vacuuming a Piece, and (5)

Vacuuming a Rule. These four acts are things that sometimes must be done for

work on the simulation to progress.

Three other activities in this transitional group are completely optional.

A KidSim "world builder" may decide whether or not to: (6) Name a Rule, (7)

Name an Appearance, or even if they want to (8) Name a Piece. Unnamed items

tended to be quite common in the three worlds represented

here.
Only one activity of these eleven is required and it need be performed but

once: (9) Naming the Microworld.

The three remaining activities of the transitional group have a more

advanced nature: (10) Modifying a Piece's Energy, (11) Running the Clock in

Reverse and (12) Rule Grouping in a Sub-Routine. The first two were rare in

application. Only Amber and Marie created and modified the Energy level of one

of their pieces.

The second activity of the final three, listed above, (11) Running the Clock

in Reverse, can be an interesting diagnostic tool and is often enjoyable to watch.

It was also little used by these KidSim programmers. Following a suggestion

from the researcher, Anne and Jenny were the only pair to watch their world run

backwards. (12) Rule Grouping in a Sub-Routine was employed by all three

pairs of KidSim users to some extent. However, this complex procedure for

"placing rules into execution groups" was not emphasized during KidSim training.

116
Functional Activities

The first two types of collaborative activities isolated in this study, Talk and

Computer Activities, are rooted in discrete, pre-defined patterns of human talk

and computer interactions. The third type of interaction identified in this study,

Functional Activities, emphasizes the "subject" of the talk itself.

Functional Activities are special interaction periods where the KidSim

collaborators are empowered via negotiation or confirmation. The Functional

Activities of Role Sharing, Applause, and Object Negotiation were found to be

special human-computer-human interactions that were encouraged by the act of

collaborative microworld creation. Each of the three Functional Activities will now

be presented individually. The Functional Activities appear in Bold, Italic Type.

Role Sharing

Role Sharing, the first of the two Functional Activities which employed

negotiation between collaborators, occurred as participants arbitrated shared

control of the computer's mouse. The use of a joint workspace, the computer

screen, plus the necessity of a shared tool, the mouse, required the students to

negotiate their interaction time with the computer. One student in each

interaction pair seemed to consistently "drive" the direction that the current

Computer Activity was taking. The other student actually controlled the mouse

117
more often. One member of each pair seemed to have a clearer goal of what

they were trying to jointly accomplish. Role Sharing was implemented by all

three pairs of students at various points in their interactions: one plans, the other

"mouses". An example of Role Sharing occurred as Anne and Jenny completed

the drawing of their first animal. Anne is currently using the mouse.

261 A: What color the eyes are going to be?


262 J: Oh, brown?
263 A: Yeah.
264 J: That right there.
265 J: Just make one.
266 J: You should've just make one eye.
267 A: I did.
268 J: Oh. Sorry. That looks better now.
269 J: Emmm, no eraser.
270 A: Okay, that looks good.
271 A: Just put it on the thing and it looks perfect.

One student, Jenny in this case, seems to have a more complete "vision"

as to how their character should look. Jenny became the teacher or "more

capable peer" mentioned earlier in Chapter Two as part of the discussion of

Vygotsky's zone of proximal development. Vygotsky believed that the sphere of

a child's cognitive activity could reach beyond those acts which a child could

perform individually to include acts which were successfully accomplished under

the guidance of a teacher or more capable peer. Jenny seems to understand

KidSim better or simply may be the more competent artist (units 264-266). Anne

may be more willing to just "get things close", rather than be perfect (unit 271).

Jenny is seeking to share the role of Creating a New Piece with Anne by giving

drawing suggestions to her.

In the following example, Mark and Cary are writing the first of many rules

which will control how their Balloon falls after the Jumbo Jet hits it. Mark has

control of the mouse.

118
824 C: Now we got to make the balloon go down. We
just need one little spotlights. Push, pull down.
825 C: Okay, now move the balloon down.
826 M: Which side? This side?
827 C: Left, yeah.
828 M: Left's on this side.
829 C: The other one.
830 M: Okay, "X" it.
831 C: You only went down, you got to go back to
APPEARANCE.
832 C: Start the clock.
833 M: It won't pop because it's too high. So we need to
move it down.
834 C: Now we got to make it keep on going
down. Go to NEW RULE.
835 M: Ah, thank you.
836 C: Stop the clock. Do we want it to get skinny or do
you just want it to fall?
837 M: Just to fall.
838 C: Go to NEW RULE. Keep on pulling it down. Just
move over that way.
839 C: And pull the balloon down.
840 M: "X" it.

This talk occurred as Mark and Carys' method of Role Sharing was

silently worked out. Mark, the less proficient rule writer of this pair, took control

of the mouse while Cary gave the commands as to what to do (units 824-831,

832). Cary queries for Mark's input (unit 836) while continuing to drive the

collaboration forward.

Role Sharing , a very democratic way to include each person in the

collaboration process, succeeds as long as each member chooses to participate.

The technique loses most of its power if the less skilled party (i.e., Mark in this

case) elects to "just watch". Mark and Cary continued their system of sharing

responsibilities until Mark tires of Rule Writing (unit 1042).


1038 C: So we want it to go into the trees.
1039 M: That's too many rules to learn.
1040 JD: You're not having any to learn them, except
when you need them.
1041 C: NEW RULE.

119
1042 M: I don't like setting up rules Cary, you can do it.
We want it to go down.
1043 C: Ha, Ha, Ha, that's what I was saying. Now we
got to move it down. Wait, (to JD) "I Don't Care
Square", right?
1044 JD: Ahhh, right below it, I guess. It's up to you.
1045 M: We don't care about nothing down there.
1046 C: I said so too. Now that's it!.

Cary took control of the mouse at this point (unit 1043) and did not return it to

Mark until the next day as a new piece was created. Mark's choice to

discontinue Role Sharing may have been a primary contributing factor in the

boredom he exhibited throughout much of the last day's interaction. Mark's many

Disengaged Solo Talk remarks referred to earlier, may be related to the lack of

"input" he felt as Cary maintained control of the mouse while rules were written.

Our third pair of collaborators, Amber and Marie, perhaps the most

"cooperative" pair of students in this study, also employed Role Sharing during

their collaboration. Marie is using the mouse in the following example, while

Amber "coaches" her as they draw "Bob", their rock.

615 M: Let me see if it will hold color right now, oops. I


just, I don't want it this color.
616 A: You're going to have to erase it. Good.
617 M: All right.
618 A: Now...
619 M: You want big rocks?
620 A: I want...
621 M: Oopsy, that's not right.
622 A: Make it gray. If it's going to be teeny tiny, I'm
going to scream.
623 M: I'll make a bigger one and connect it to it.
624 A: Okay. I think that's too dark gray.

Marie continues to draw, listening to Amber's wishes (unit 622), by

"connecting" two small rocks together to create a larger one (unit 623). There

was a concern by the girls at this point in their interaction that the pieces they

were creating were not coming out big enough on the game board.

120
625 M: There's this color...
626 A: Ehh.
627 M: Or there's this one...
628 A: It's getting better. Try one more higher up. Yeah.
And we could put a dark animal on it.
629 M: Animal on it? Oh, yeah...
630 A: Umm hmm.
631 M: ...on it, I thought you meant on the same drawing.
Make it a huge one, oops... take the mouse...

Marie has completed the rock and passes control of the mouse to Amber (unit

631).

All three pairs of students employed Role Sharing at some point during

their microworld construction in order to allow each partner to stay engaged.

Ultimately, when people are collaborating together at the computer, only one

person at a time may control the keyboard or mouse. The other serves as an

"advisor or interested audience" member. Through Role Sharing, each person

in the advisor role is given a way to have "input" and therefore derive a sense of

ownership of their joint creation. Through the latter "audience member" role,

opportunities are presented which allow for the expression of praise about a

current action or result. This expression of praise is the Functional Activity

called Applause.

Applause

Everyone needs encouragement and confirmation from time to time.

Applause is defined here as verbal reassurance that the current outcome is a

positive one. Applause is affirmation for the person currently controlling the

computer that "all is well" or "something extra nice just happened." Applause

tends to occur in short, one or two line Talk Units.

121
The following excerpts contain a number of examples of Applause from

each of the six students that participated in this study.

67 Amber: That looks awesome!!


136 Amber: I think it's a cool giraffe. Okay, I think you
should (put) one more right there.
432 Amber: That's kinda pretty, I think...

271 Marie: Oh, that looks neat. It looks like a little island.
372 Marie: Oh, this is cool.
427 Marie: I like it like that.

231 Anne: Beautiful tail. Make its tail come around.


382 Anne: That looks like a cutie elephant.[childish voice]
833 Anne: Yes, that's good. And that looks good.
837 Anne: Yeah, that looks good. Just leave it like that.
I like that. That looks dingy. Sorry. Oh cool!
Yeah, I like that.
628 Jenny: I like the head. I like the ears.
629 Anne: Are you just saying that?
630 Jenny: No, I like it. Looks good.

184 Mark: That looks good Cary.


963 Mark: That was sweet.
1902 Mark: Wow...

915 Cary: Yeah, it's lookin' okay.


1513 Cary: That's a pretty good hang glider, ya I think.

Applause is the icing on the collaboration cake. Applause is not required

for successful collaborative communication between computer users to occur,

but it can make the experience more rewarding when one's efforts are

recognized and acknowledged. Applause from a fellow collaborator goes

beyond the typical "OK, that's finished, on to the next task," type of remark.

Applause says something extra special and "kind of great" just occurred.

Applause is the best kind of peer communication.

122
Object Negotiation

Each individual experiences the world in a unique way. We all have

different histories. We all have seen something, been somewhere, or have

accomplished something distinctly different from our nearest neighbor or our

closest friend. Each of us has built a personal life history of memories about the

common and uncommon objects and events in our lives. We alone have this

unique view of the world.

Normally, in day-to-day living, our view of the world goes unchallenged.

Whether by choice or by accident, we are not often called upon to defend how

the world looks or feels to us. The six children who participated in this study did

not have this luxury. They each were asked to reveal how some part of the world

appeared to them. In doing so they created new worlds of their own and for a

few moments saw each other's world through the other's eyes.

We have already seen one example of Object Negotiation, Anne and

Jenny's rather heated discussion about what grass looks like. Grass is a very

common object to the small-town girls. Each of them feels certain about its

appearance. Common objects are ubiquitous across a group of people. Every

person recognizes these "everyday" things (Norman, 1990). What they look like

is not often open to debate.

Anne and Jenny discovered that something as common as grass can be

represented in different ways, as evidenced by the following post-hoc interview

excerpt.
471 JD: What didn't you like about Anne's, rather, Jenny's
grass?
472 A: It looked too plain. It had no squigglies. It was
just plain green.
551 JD: What, what should grass look like?

123
552 J: Grass is like that and it's like all colored in, and
she just had it going all over the place.

Drawing proficiency on the computer may also have been a contributing

factor for each collaborator as they tried to represent their ideas visually. The

dispute may have occurred as a result of one or both of them having the inability

to "picture" grass with KidSim the way that they saw it in their own minds.

Eventually, Anne and Jenny were able to negotiate their way to some

compromise grass. Their final object was mostly "colored in" and at the same

time it did have some "squigglies" .

Through Object Negotiation Anne and Jenny were able to use a visual

demonstration to illustrate just what "they had in mind" to the other person. A

visual demonstration is the performance of one's ideas through visual examples.

Drawing in KidSim's Appearance Window, they were able to interactively explore

and animate each other's creations. The computer allowed them to merge their

ideas into one. Creating objects for a microworld does not really require building

them from scratch, instead objects are synthesized from what one already

knows.

All six students employed Object Negotiation while they were creating

their microworlds. Mark and Cary negotiated what their plane should look like.

169 C: There's its tail. I can't get it to make one...yeah I


can.
170 M: Looks like a boat to me.
171 C: Cause, I'm not done with it.
172 M: It does look like that would be a seat.
173 M: We could have a pond, and then the airplane
crashing into the boat.
174 C: You wanna do the airplane?

124
Cary is not amused (unit 174). Mark is offering his opinion that what Cary

is drawing does not look like an airplane. Mark goes as far to say that Cary's

drawing looks more like a boat to him. Mark attempts to negotiate with Cary, by

suggesting that they do not need to have this drawing be an airplane at all. It can

really be a boat (unit 173). Cary continues to draw, adjusting the "tail" more to

Mark's liking.
177 C: Does that look a little bit more like a plane?
178 M: Yes.
179 C: Good, now I'll just erase that top part and draw
the color. Uh oh. OOPS.
180 M: What is it? Is it a circle?
181 C: It's supposed to be. Now, I'll color it in. There.
It's something, ain't it funny?
182 M: Yeow it is.

Mark agrees with his friend Cary (unit 182). But we are not sure whether

Mark agrees with Cary that the drawing is "something" or is Mark just agreeing

that it is "funny looking". One unit later Mark tells us how he really feels.

184 M: That looks good Cary.


185 C: Cool, thanks. Go to the paint bucket.
186 C: Do we want a black plane?
187 M: Yes.
188 C: What color should that... lets make that gray.

Cary has now drawn a plane that Mark thinks "looks" good. After coloring

it in black and making the plane's "cockpit" gray, Cary gives his opinion of the

drawing (unit 203).

203 C: There, that looks like a little bit like an airplane.


204 M: Now you have to fix that in there, right?
205 C: What do you mean, fix it?
206 M: Like ah (whispers).

Mark points out a problem he sees (a dot of white on the plane's body that

was not filled in with black) and offers Cary advice on what to do. Throughout

125
the plane's creation Cary has listened to Mark and has adjusted his drawing

accordingly. Cary checks with Mark one last time to get his final approval.

213 C: There we go, that looks like a plane a little bit,


don't it?
214 M: Uhm hmm.
215 C: Okay. Go to "4", "X" it.

Together they have created a drawing that they both acknowledge is an

airplane. Had this creation been the work of one party alone (e.g., Cary), no

doubt a very different looking object would have resulted. The final appearance

that the plane took was synthesized from "ideas of planes" that both of the boys

possessed. The plane was negotiated into being.

A third example of Object Negotiation comes from Amber and Marie.

Marie suggests that "Animal Life" could use some birds.

328 M: Make some birds.


329 A: Okay.
330 M: I wonder if you can mold with that?
331 A: Oh, (laughs)
332 M: Why? Okay, now, clear the whole thing. Okay.
Should I just make "M" birds?
333 A: "M" birds?
334 M: No, that was bad.
335 A: That's my bird!

Amber opens the Appearance Window (unit 329) and begins to draw a

bird. Marie makes a comment about the generic "blob of clay" that comes up

with every new piece (unit 330). Marie sees Amber's efforts and volunteers to

make "M" birds instead (unit 332). Amber has never heard of "M" birds and

objects to Marie erasing her effort (unit 335). Marie offers Amber an explanation.

336 M: "M" birds, they look like M's from a distance, and
then put a little head on them.
337 A: That's a perfect size. Because, we don't want
them real big. That's an "M" bird?
338 M: That looks awful, oh, you know...

126
339 A: I didn't say it looked awful...
340 M: I said it looks awful.
341 A: I've never seen an "M" bird.
342 M: Well, they just look from a distance... and they're
easier to draw, looks awful. See, I can't, I wish I could
do straight lines.

Marie introduces Amber to the wonderful world of "M" birds. Amber is still

somewhat skeptical (unit 337), in her experience she has "never seen an 'M'

bird." However, it is Marie who is playing the role of the critic here (unit 338).

Marie has shown Amber a new way of seeing birds, at least "from a distance".

"M" birds are a good example of the kind of "object knowledge" differences

one may find across different people. Marie has one conception of how birds at

a distance may be pictured, Amber did not share this view (at least not at first).

By demonstrating just what an "M" bird was, Marie was able to negotiate with

Amber and convince her that this new way of picturing birds was acceptable.

Each student pair had to negotiate the "look and feel" of objects while

creating pieces for KidSim. Every student came to this interaction with pre-

conceived notions of what things look like and how things behave. Object

Negotiation is collaborative communication about a game piece's appearance or

behavior. These objects were negotiated into being.

Summary of Functional Activities

Collaborative communication between computer users was found to

function in three capacities during the learning of KidSim. Each of these newly

discovered functions is embedded across Computer Activities found throughout

the collaboration. All three functions contribute in an interesting way to the

student interactions observed for this research.

127
As people are collaborating together at the computer, only one person at a

time may control the keyboard or mouse. The other serves as an "advisor or

interested audience" member. Through Role Sharing, each person in the

advisor role is given a way to have "input" and therefore derive a sense of

ownership of their joint creation.

Applause is affirmation for the person currently controlling the computer

that "all is well" or "something extra nice just happened". Applause is not

required for successful collaborative communication between computer users to

occur, but it can make the experience more enjoyable and rewarding. Applause

is the icing on the collaboration cake.

Collaborative communication about a game piece's appearance or

behavior took place through Object Negotiation . Every student came to this

interaction with pre-conceived notions of what things look like and how things

behave. Each student pair had to negotiate the "look and feel" of objects while

creating pieces for KidSim.

Role Sharing, Applause, and Object Negotiation give collaborative

microworld construction its "cooks", "icing" and "cake"; all things necessary for

"baking" a tasty KidSim experience.

This concludes the presentation of the results of the study. The following

chapter provides a summary and discussion of these results as well as their

possible implications for the present and future research.

128
Chapter 4

Summary and Discussion

Introduction

Six fifth grade students were recently given a unique opportunity. They

were given the freedom to design a world of their own. No adult told them what

to create. No teacher gave them an outline to follow. They simply were told to

let their ideas take them far as some software and a brief stretch of time would

allow. Not every child accepted this offer equally, nor should we expect them to.

In the end, they built some pretty amazing places, places which had been hidden

from view in those microworlds we call children's imaginations.

Along the path to these creations they spoke about their plans,

discoveries, likes and dislikes. Talking between themselves or to a supportive

observer, they coordinated the construction of characters, acts, and interactions.

They talked throughout, revealing something of what was within them.

Summary of the Results

A summary of the results will now be presented which offers answers to

the two research questions posed by this study: R1) What types of talk occur

during the collaborative learning of computer software? and R2) What functional

role does communication play during a collaborative microworld construction

experience?

Talk between the collaborators was discovered to follow a pattern which

initially was very similar to categories reported by the SLANT research group

(Mercer, 1994). SLANT originally reported finding the categories of

129
Disputational, Cumulative and Exploratory Talk. This dissertation confirmed the

existence of these three and added the categories of: Help Talk, Query Talk,

and Solo Talk. Each of the six Talk Activity types expresses a different variety of

verbal communication.

(1) Researcher and Student Initiated Help Talk, are the primary interaction

activities between the students and the researcher/facilitator. Researcher

Initiated Help Talk occurred primarily as a form of supplemental training or as a

reminder of KidSim details which had already been discussed. Student Initiated

Help Talk verified the progress of student learning and confirmed the recently

acquired attributes of the software.

(2) Cumulative Talk is the interaction activity in which each speaker

indicated to the other that s/he is "mentally present" and aware of the exchanges.

(3) Query Talk is the interaction activity which indicated that more

information is needed before the current operation may proceed.

(4) Solo Talk is the interaction activity which indicated that a collaborator is

"going it alone" for the moment and may serve as a subtle plea for more attention

from the other partner. Solo Talk can be either engaged or disengaged.

Engaged Solo Talk is talk about the current computer activity. Disengaged Solo

Talk is talk which is "off topic" or not pertinent to the current computer activity.

(5) Disputational Talk is overtly negative in tone and is the interaction

activity which most clearly says: "I don't agree with you".

And finally, (6) Exploratory Talk is the interaction activity which signals that

one partner is willing to share knowledge with the other.

The six Talk Activities also have parallels with Teasley and Roschelle's

(1993, p. 236) categories of discourse events. Two of their four categories, Turn-

130
Taking and Narrations, complement the current findings well. Turn-Taking was

defined as series of smooth turn transitions which followed, from each

participants point of view, a predictable pattern (p.237). Although not confirmed

from the participant's viewpoint, Cumulative Talk, may also be characterized as a

predictable pattern of smooth turn transitions. Narrations, otherwise, appear to

be somewhat similar to Engaged Solo or Exploratory Talk, as they both involve

"announcing of one's intentions" which correspond to physical actions (Teasley &

Roschelle, 1993, p. 238).

The four primary Computer Activities (e.g., Creating a New Piece, Rule

Writing, Modifying A Piece's Appearance and Running the Clock Forward )

were supported and driven by a flowing mixture of the six talk types. Higher

quantities of Cumulative Talk were a good indicator that the current Computer

Activity was well understood by both partners. Creating a New Piece in KidSim

is much like drawing objects on paper. It stands to reason that the more familiar

or comfortable each partner is with the current microworld construction task, the

more likely they are to have a conversation which is largely "give and take", or

cumulative in nature.

Help Talk was prevalent whenever novel developments occurred or more

information about KidSim was required. Rule Writing and Modifying A Piece's

Appearance each created situations which were new to the participants in this

study. The students needed more "help" via Help Talk to understand and

reinforce the complexities of "graphical rewrite rules" and "piece appearance

changes". Cumulative Talk lessened and Help Talk increased as circumstances

became less familiar.

131
Query Talk occurred spontaneously throughout the interaction as

individuals requested information from each other. These questions sometimes

came as the "non-mousing" partner sought to understand what the person

currently controlling the computer was trying to accomplish. At other times, a

partner's query allowed one partner to encourage the other to selectively

examine their work in a positive way without falling into the "disputational trap" of

lingering disagreement.

Disputational Talk was generated whenever a partner refused to give a

reason for their dislike of a current action. The "familiarity factor" discovered

during Creating a New Piece could also be at work here. Perhaps, the more

familiar we are with an object or action, the more likely it is that we will have

strong opinions about it. The expressing of these opinions may appear without

justification at first as one may simply be too upset to truly articulate his or her

complete thoughts. Sequences involving Object Negotiation became

disputational whenever the students' "ideas" of worldly objects clashed with their

perception of how those objects were represented on the screen. Ultimately,

Disputational Talk proved to be counter productive. Those students who tried it

soon realized that they had to switch to another talk type (often Query or Help

Talk) if their objections were truly to be heard.

Engaged Solo Talk was used by Role Sharing pairs to give non-mousing

members a way to stay involved in the interaction as the silent partner performed

on-screen actions. Disengaged Solo Talk was especially evident with one

student in particular when he lost interest in the current KidSim activity. Solo

Talk gave participants a means of expressing their thoughts out loud, particularly

during long, involved Creating a New Piece or Modifying A Piece's

132
Appearance interaction sequences. The possibility of Engaged Solo Talk

provided the means for these collaborators to have a completely "open ended"

conversation. They could leave the "lines of communication open" without

worrying that their "connection" to their partner would be broken by the lack of a

verbal response.

Exploratory Talk, while low in actual percentage count, became a highly

effective way for one partner to pull the other into their "zone of understanding".

Complex portions of the KidSim construction process, such as the "Select, Copy,

Paste, then Paint" steps involved in Modifying A Piece's Appearance, were a

fertile ground for Vygotskys "more capable peer" to take the opportunity to teach.

One collaborating partner would grasp a possible solution to a current problem

and the other became the pupil, listening and watching as new insights were

revealed. Exploratory Talk provided the peers a way to educate each other.

Talk between the collaborators functioned through Role Sharing,

Applause, and Object Negotiation, to support, cheer, reflect, and guide their

interactions with each other as they learned to program in KidSim. Microworld

construction via "programming by demonstration" allowed these pairs of best

friends to become involved in each other's actions. It is believed they learned

something from each other. After all, they built a world together.

Discussion of the Results

This dissertation delineated the process of collaboratively building a

microworld in KidSim through a merging of discourse analytic and grounded

theory approaches. Discourse analysis served as a means to distill, separate,

and reinforce distinct categories of Talk and Computer Activities. Grounded

133
theory provided the mechanism to isolate the three varieties of Functional

Activities.

This study sought to create an environment where students could

collaborate within the protective circumference of Vygotskys zone of proximal

development. Each child was given the opportunity to grow a little "outside

themselves". They had a chance to find out from each other what they could not

experience alone. By being creative together, these children discovered

something beyond the development of microworlds, they also saw a small part of

how the other views the world.

KidSim microworld construction was found to have a natural flow to it

which follows the primary Computer Activities presented in these Results. The

microworld is planned, either a piece at a time, or as a semi-whole. Pieces

are created for which interaction rules are written. The game clock is run

and the appearance of pieces are modified. Worlds may be saved

permanently or changed at will. Each activity is connected to the one before it

and acts as a bridge to the one that follows. The other eleven Computer

Activities act as "mortar" between these five "bricks" of microworld construction.

Through six varieties of Talk Activities these KidSim collaborators found

appropriate avenues to express their feelings and ideas. Cumulative Talk gave

them a road on which to proceed when communication was most open and

agreement for how to plan their course of action ran high. Query Talk allowed

them to slow down for a moment, back up, or to make a turn. Disputation Talk

actually halted forward progress. The extra weight of ill feelings had to be

unloaded before further movement could occur. Help Talk "paved the way" and

tried to lift these "vehicles" over potholes as unobtrusively as possible. Solo Talk

134
came both in the form of an one-lane bridge over which only a-driver-at-a-time

was heard moving and as an exit ramp which led away from the conversation

altogether. At few very special moments, Exploratory Talk became the map

which pointed the way in new directions. Collaborative talk allowed the KidSim

explorers to navigate a world of their own making. A world of many twists and

turns, but one which only they were able to truly inhabit. We were mere

observers of their excursions.

Implications

This study isolated six different varieties of Talk Activities which help us

better understand the flow of communication during collaborative computer use.

The prevalence of one talk type over another points to the relative progress of

the human-computer interaction. Positive human-computer interaction progress

is occurring when Cumulative Talk is in great abundance. Moments of Query

Talk indicate a slowing down of that progress and sequences of Exploratory Talk

a speeding up. Help Talk can either be a catalyst for growth in new directions or

serve as a safety net when the situation goes awry. Solo Talk by an engaged

partner indicates that they wish to have verbal input into the interaction whether

or not they are currently controlling the mouse. Disengaged Solo Talk is a sign

that boredom is setting in or that the current Computer Activity is beyond their

immediate comprehension. Periods of Disputational Talk are a sure sign that a

disagreement about some aspect of the microconstruction has occurred.

The sixteen highlighted Computer Activities give us some insight into how

artificial worlds created on the computer "fit" into the real world inhabited by

children. The "familiarity'" of a Computer Activity, its "worldliness", relates directly

135
to what kind of talk is produced to support it. The more an on-screen action is

like a real-world action (i.e. the more familiar it seems to the "user"), the greater

the likelihood of sustained "give and take" or Cumulative Talk between partners.

Likewise, the more foreign a human-computer interaction "feels", the more likely

that talk will be produced which is halting in nature (e.g., Query or Help Talk).

The three Functional Activities illustrated have important implications for

the study of the whole of human communication, beyond that which occurs

during collaborative computer use. Each of these Functional Activities is a

"micro-example" of how real world interactions occur. They act as "condensed

examples" of interactions which take place when two persons are working

together on a project. They force into being situations which in "real life" may

take much longer to occur. Role Sharing, Applause, and Object Negotiation

could have interesting implications for all manner of collaborative effort. First, the

notion of Object Negotiation points to computer simulation as a catalyst for

team workers to truly "get inside" each other's perspectives. Secondly, long

collaborative work periods without appropriate Applause may feel unrewarding.

Third, the overall input of each member of a collaborative team might be better

balanced if Role Sharing was encouraged through a shared computer interface.

The overall patterns of communication found in this study (i.e., lots of Help

Talk during Rule Writing, the rare examples of Exploratory Talk, etc.), are directly

related to how much "talk energy" a Computer Activity requires. The "smoother"

an individual's current human-computer interaction, the less need there is for

great quantities of talk and the more likely that pauses will accumulate. In other

words, the "easier" a Computer Activity is for an individual, the less need they

may have to talk about it. Note that some individuals may like to talk more when

136
they feel they are accomplishing their goals, but the point here is that there

seems to be less of an "external need" for them to do so. Help Talk, for example,

consumed a lot of everyone's "talk energy" as the very flow of communication

increased from a two-way to a three-way interaction while microworld building

progress slowed. Increased amounts of Engaged Solo Talk, in particular, were a

good sign that at least one member of the collaborative pair was deeply engaged

with the current computer action, as they said nothing, or very little, to their "solo-

speaking" partner.

A final implication of this study speaks directly to those involved in the

design of computer software for children. The powerful technique of

"programming by demonstration" is only as useful as the software tools used for

its production. These 5th grade KidSim users, while rushed for time, did manage

to build functioning microworlds with large amounts of Help Talk. Yet, in the end,

this researcher believes, these students only partially understood exactly how

KidSim's various "pieces" fit together. These were all good, if not "all A",

students. Yet, only one of them, Marie, truly assimilated KidSim's basic functions

in the time allowed. This is a sign that the interface or its various parts still need

"work" to easily '"fit" into the worlds and minds of these younger children. This

researcher believes that younger children have a tough time with any tool or

feature that is sometimes hidden, sometimes not. Constant presence, or a lack

of "modality" for all of the interface's features would help to reinforce "where and

what" things are to a child. After all, all a child's crayons do not move around or

change color inside the box just because the child is not currently using them.

The Rule Window, with its collapsing boxes, seems a particular problem area.

137
Limitations

Chief among the limitations found in this study was the abbreviated

interaction period allowed for each student pair. Realizing that the subjects of

this study were actively enrolled elementary school children, there was a limit as

to how long each child could be pulled from their regular classes for this project.

However, with this limitation in place, adequate training time was still a concern.

Three one hour sessions were not enough time to thoroughly cover the finer

points of the KidSim architecture. Limited interaction time also restricted each

day's post-interaction interview to a five minute meeting. This daily time alone

with a participant was only adequate to assess each individuals reaction to some

of the proceeding events. More interview time is needed to provide an

opportunity for in-depth probing.

The sample size of six, while manageable in terms of transcription volume,

was a limitation for many reasons. Females out numbered males two to one,

possibly influencing the results. This sample, qualitative by nature, can be

considered representative only to the extent that these students are the same

age as their fifth grade peers. The sample was highly selective in several

respects. These were not just random pairs of students, but best friends. Also,

the students were "singled-out" out by their teachers as "computer literate" and

children who were thought to handle new challenges well. One or both of these

attributes may not be typical of an average eleven-year-old. Also, the age of the

children itself may be considered a limitation.

Scientists at Apple Computer, producers of the KidSim software,

mentioned to this researcher that they had only brief experience with KidSim by

users who were under twelve-years-old. The young age of these participants, as

138
noted earlier, may have contributed to the difficulty most of them had mastering

KidSim's rule writing technique.

Another limitation is the great quantity of Help Talk present throughout

many steps of the microworld construction process. Help Talk took on the

pragmatic role of "extended training support" instead of that of "intermittent

assistant" that was originally envisioned. Several factors contributed here.

One factor mentioned earlier was an inadequate pre-interaction training time

which pushed the learning of KidSim software over into the actual interaction

days. Secondly was the existence of "non-verbal eye contact cues" from the

children that clearly "asked" the researcher for input without using spoken

language. A third factor that contributed to the high quantity of Help Talk was the

nature of this study which "in and of itself" was a bit "out of the ordinary". These

children were meeting together, outside of their normal classroom, with an

unfamiliar adult. The researcher found himself "talking to them a bit more", than

he might have otherwise, to help them feel extra comfortable in this unusual set

of circumstances.

Solo Talk, although divided here into Engaged and Disengaged

components could have benefited from a further breakdown into Solo Talk "while

using or not using the mouse". Such additional analysis would have proved

useful in determining if Engaged Solo Talk was either categorized as "self-talk

from a mousing partner" (i.e., as originally defined) or would be more

appropriately labeled "talk with the intention to give verbal input" in the case of an

"Engaged non-mousing partner". Unfortunately, such a determination would be

difficult to achieve as the inclusion of non-verbal artifacts, such as the identity of

the current mouse user, was not recorded as a part of the original transcripts.

139
Another limitation concerns the appropriateness of a publicly accessible

environment for such a concentrated interactive experience. One student in

particular, was very distracted whenever an outside adult came into the "after

hours" library environment that was the location for this study.

KidSim, however wonderful in its potential, was at the time of the study still

"beta" software and therefore not yet finished. This was a limitation as some

features, such as the editing of rules, had not yet been incorporated into the

program.

Overall, one of the key limitations of the study was not so much concerned

with the research environment but with the analysis technology itself. Discourse

analysis using the qualitative analysis software package, NUDIST (version

3.06), proved to be a much more time consuming enterprise for this researcher

than initially hoped. Instead of allowing existing sub-categories to be easily

divided into new groups, each new sub-category required another "pass" through

all transcribed speech to assign every talk unit appropriately. The most recent

edition of this program (as of this writing, August 1998) is version 4.0. Only with

this most recent version was the author able to easily "split" current categories

into sub-groupings.

Future Research

Although not specifically designed to gauge the influence of gender during

interaction with KidSim, this study noted an interesting tendency. The one pair of

boys involved in this study (Mark and Cary) chose to write rules for their pieces

much earlier than either pair of girls. Only thirty minutes into the first day's

interaction the boys had drawn two characters and had begun to give them rules.

140
Both pairs of girls on the other hand, elected to create the majority of their pieces

first and then assign rules to them.

The boy's incorporated much more "action" into their creation, building

roughly drawn pieces with lots of movement. The girl's worlds were much more

detailed by comparison, with a limited amount of movement. The boys were

more interested in animating a current piece on the game board, than in outlining

the entire world in advance. This finding may simply be another example of the

stereotype that "boys are more action oriented and girls more detail oriented".

The grounded theory approach utilized here may only hint at these and other

broader questions. Answers must await a different approach that seeks more

generalizable solutions.

This study provides more evidence for a theory of computer use that

Dailey first suggested in 1995. Computer users may be categorized along a

continuum from explorers to producers. Explorers, are people who wish to

exploit the features of a computer program, discovering all they can about its

possibilities. Producers, however, concern themselves more with the outcome of

computer use. Producers, focus on the "output" that the computer provides. Of

these six peer collaborators, Jenny, Amber and Marie seem to be more oriented

toward software exploration, while Anne, Mark, and Cary, tended to be more

interested in the creation of computer output (i.e., creating or moving pieces on

KidSim's gameboard).

Once they had gotten into the "flow" of working with KidSim, these children

completely adjusted to the situation and became absorbed into the interaction

with the computer. Although they seemed to forget all about being observed or

tape recorded, they knew help was not far away if they needed it. The computer

141
and KidSim itself, proved to be so engaging that they let "down their guards" and

became uninhibited for a few hours. The SLANT researchers noted that

computers capture the attention of children in a way that is not approached in its

effectiveness by other kinds of classroom collaborative encounters (Mercer,

Phillips, & Somekh, 1991). Perhaps by introducing these kids to an engaging

computer program like KidSim, we gave them the freedom to "relax" and simply

enjoy this new "game", without being constantly aware that they were being

recorded.

As mentioned above, any future study of KidSim or its successors should

provide for a longer interaction. An extended training period is in order which

would allow the finer points of the program to be explored more deeply. Younger

children could be compared with older ones on an activity by activity basis.

Would a second or third grader understand KidSim's rules or rule grouping?

Perhaps children could be placed into pairs of differing KidSim experience,

contrasting novices with more accomplished users. The goal might be changed

from building new worlds to the teaching of KidSim to a peer, so that Vygotskys

zone of proximal development might be stretched even further.

A final area for future research might be increased attention to the non-

verbal elements of collaborative communication during human-computer

interaction. Tying a video taped sequence of the participants in action to a

scrolling display of their transcribed words would give another dimension to the

analysis. Coded comments could be highlighted and "extracted" with the

relevant video frame number noted. These two "streams of data", verbal plus

non-verbal, would give the researcher a more complete "vision" of the interaction

experience. This material could then be revisited at some future point. It would

142
serve as a beginning for future investigations that relied less upon notes or

memory alone to help recreate a "field day's" true interaction.

Final Thoughts

Qualitative communication research can become very personal, touching

actual lives and changing who we are as investigators. Because of the intimate

nature of this study, this researcher has gained new respect for the infinite

imaginations of children, their desires and their plans. Given a wonderful tool like

KidSim they can more fully show us what they are dreaming and thinking.

Computers and microworld simulation software give our children a

different way to explore their world. "KidSim kids" are free to explore without

restraints, deciding what their creations will look like, they build the pieces and

create the animation without the limits often imposed on them by adults. No

"grownup" has given them a limited set of characters or actions. In the words of

Mark: "...you're the boss of yourself". Perhaps self-reliance is the best lesson

KidSim teaches users.

In this research, three pairs of best friends collaborated with each other to

create their own "universes." Best friends were chosen for this study, as they are

thought to be cognitively similar in nature (Infante, Rancer, & Womack, 1990).

However, these three pairs of 5th graders proved to be somewhat different from

each other, at least as exhibited by their actions. Each pair had a physically

dominant partner (Jenny, Cary, & Marie) who controlled the mouse the majority

of the time, particularly during rule writing. Two of the pairs had a partner who

was more verbal than the other (Anne & Mark). Finally, all three pairs seemed to

143
have a partner who preferred drawing to rule construction (Anne, Mark, &

Amber). Since they were best friends, they did seem to respect the other with

politeness and kindness. These collaborating best friends did try to work

together, only occasionally getting on each other's nerves.

This researcher would welcome the opportunity to work with best friends

again. Much in the way of "peer friction" was eliminated by having students work

together who already know each other well. All of the '"getting to know

you" kind of interaction was avoided and serious effort could begin sooner.

Although restricted in scope and limited in duration, this study sought to

explore the essence of student collaboration during human computer interaction,

their talk. Through grounded theory, several Computer and Functional Activities

were isolated and illustrated. The close inspection of computer users actually

interacting with software, in the user's own environment, proved to be a very

good way to predict how well a given piece of software is designed for a

particular set of users.

The author was particularly gratified to witness how "easy to learn" parts of

a computer program can be made for young users. KidSim's draw module,

overall, seems to be a good re-creation of a child's "paint set" or "Crayon box".

On the other hand, some parts of KidSim were frustrating to these 5th graders.

Rule Writing does present a degree of difficulty which only an increased level of

concentration appeared to conquer. Grouping rules, in particular, was a

challenge to these children. More time devoted in training to Rule Writing is

certainly required. However, throughout the investigation it was the children who

consistently wished for more overall time with the software to create places of

their own.

144
Apple Computer should be commended for developing a "programming by

demonstration" application for children. KidSim gives children a way to create a

new kind of "playground" where they are the masters. Giving power to children is

a good way to improve their self confidence. We as adults could also benefit

from this kind of "self-programming" power. One day, perhaps, all computer

programming will be done through "brain windows". Complex computer

languages might become like Morse Code, something for the history books.

Whatever the future might hold, one outcome of this research is a

certainty: these students enjoyed themselves. Overheard on the last day of the

study was one child's remark to another: "The other kids think we're making a

video game." What better fun for an eleven-year-old?

145
ENDNOTES

1
Originally, four pairs of 5th graders were contacted to participate in the study. However,
recordings of one pairs talk proved untranscribable, due to long silences mixed with large
amounts of whispering.
2
This device, a ProPresenter Mac/PC Multi-Sync, allows not only the videotaping of the IIsi
output, but also has the capability to zoom-in on a given part of the screen- allowing for the
close study of a small area of the image.

Po Pe
3 pi =
1 Pe Where pi = intercoder reliability, Po = observed percent agreement, and
Pe = percent agreement to be expected on the basis of chance alone.
4
Note: The total number of Talk Units associated with Computer Activities is higher than the
actual number of analyzed Talk Units. This occured as a result of dual coding of some activities
(i.e., Planning the Microworld, did not have a distinct Computer Activity of its own, which, as
noted on p. 104, often took place while Running the Clock Forward, etc.)
5
First the original appearance must be opened and selected entirely via the SELECT ALL
command (under the EDIT menu in Appendix A, Figure A2). Second, this selection must be
COPIED. Third, this copy must be PASTED into a NEW APPEARANCE. Once pasted, the NEW
APPEARANCE may then be modified in some way to look different from the old.
6
The editing of previously created rules was not possible in this version of KidSim.

146
REFERENCES

Adams, D., Carlson, H., & Hamm, M. (1990). Cooperative learning &
educational media. Englewood Cliffs, NJ: Educational Technology
Publications.

Ahern, T. C. (1991). The effect of graphic interface on participation, interaction,


and student achievement in a computer-mediated small group discussion.
Dissertation, Pennsylvania State Univ. Dissertation Abstracts
International. 1991 Oct Vol 52 (4-A) 1296.

Anderson, R. J. (1994). Representations and requirements: The value of


ethnography in system design. Human Computer Interaction, 9(2),
151-182.

Anderson, R. J., Heath, C. C., Luff, P., & Moran, T. P. (1993). The social and
the cognitive in human-computer interaction. International Journal of
Man Machine Studies, 38(6), 999-1016.

Anderson, W. T. (1990). Reality isnt what it used to be. San Francisco:


Harper & Row.

Apple Computer, I. (1991). Apple classrooms of tomorrow: Philosophy and


structure [and] Whats happening where. Report ACOT-R-14. Apple
Computer.

Bench-Capon, T. J., & McEnery, A. M. (1989). People interact through


computers not with them. Interacting with computers, 1(1), 31-38.

Biocca, F. (1993). Communication research in the design of


communication interfaces and systems. Journal of Communication
43(4), 59-68.

Biocca, F., & Levy, M. R. (Eds.). (1995). Communication in the age of virtual
reality. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Blaye, A., Light, P. H., & Rubtsov, V. V. (1992). Collaborative learning at the
computer: How social processes interface with human-computer
interaction. European Journal of Psychology of Education, 7(4), 257-267.

Brand, S. (1987). The media lab: Inventing the future at M.I.T. New York:
Penguin Books.

Bruner, J. (1985). Vygotsky: a historical and conceptual perspective. In J. V.


Wertsch (Ed.), Culture, communication, and cognition: Vygotskian
perspectives. Cambridge, England: Cambridge University Press.

147
Clements, D. H., Nastasi, B. K., & Swaminathan, S. (1993). Young
children and computers: Crossroads and directions from research.
Young Children, 48(2), 56-64.

Cypher, A. (Ed.). (1993). Watch what I do: Programming by demonstration.


Cambridge, MA: The MIT Press.

Cypher, A., & Smith, D. C. (1995). KidSim: End user programming of


simulations. In Association for Computing Machinery, Special
Interest Group in Computer-Human Interaction, Annual Conference,
(pp. 27-34). Denver, Colorado: Association for Computing Machinery.

Dailey, J. C. (1995). Learning the screen: An ethnography of childrens


communication during human-computer interaction. Unpublished
manuscript.

deSouza, C. S. (1993). The semiotic engineering of user interface languages.


International Journal of Man-Machine Studies, 39, 753-773.

DeVillar, R. A., & Faltis, C. J. (1991). Computers and cultural diversity.


Albany, NY: State University of New York Press.

Dillenbourg, P., & Self, J. A. (1992). A computational approach to socially


distributed cognition. European Journal of Psychology of Education,
7(4), 353-372.

Elliot, J. (1983). A curriculum for the study of human affairs: the contribution of
Lawrence Stenhouse. Journal of Curriculum Studies, 15(2), 105-123.

Erkens, G., & Andriessen, J. E. (1994). Cooperation in problem solving and


educational computer programs. Computers in Human Behavior, 10(1),
107-125.

Fisher, E. (1993). Characteristics of childrens talk at the computer and its


relationship to the software. Language and Education, 7(2), 97-114.

Gergen, J. (1986). Correspondence versus autonomy in the language of


understanding human action. In D. W. Fiske & R. A. Shweder (Eds.),
Metatheory in social science (pp. 136-162). Chicago: University of
Chicago.

Glaser, B. G., & Strauss, A. L. (1967). The Discovery of grounded theory.


Chicago: Aldine Publishing Company.

148
Hale, D. P., Hurd, J. E., & Kasper, G. M. (1991). A knowledge exchange
architecture for collaborative human-computer communication. IEEE
Transactions on Systems, Man, and Cybernetics, 21(3), 555-564.

Harel, I. (1990). Children as software designers: A construction approach


for learning mathematics. Journal of Mathematical Behavior, 9(1), 3-93.

Hoyles, C., Healy, L., & Pozzi, S. (1994). Groupwork with Computers: An
overview of findings. Journal of Computer Assisted Learning, 10(4),
202-215.

Infante, D., Rancer, A. S., & Womack, D. F. (1990). Building communication


theory. Prospect Heights, IL: Waveland.

Jacob, E. (1990). Alternative approaches for studying naturally occurring


human behavior and thought in special education. Journal of Special
Education, 24(2), 195-211.

Jones, A., & Mercer, N. (1993). Theories of learning and information


technology. In P. Scrimshaw (Ed.), Language, classrooms, and
computers. New York: Routledge.

Killiam, H. W. (1991). Rogerian psychology and human-computer interaction.


Interacting with computers, 3(1), 119-128.

Krippendorff, K. (1980). Content analysis: An Introduction to its methodology.


Beverly Hills: Sage

Lajoie, S. P., & Derry, S. J. (Eds.). (1993). Computers as cognitive tools.


Hillsdale, NJ: Lawrence Erlbaum Associates.

Laurel, B. (Ed.). (1990). The art of human-computer interface design


Reading, MA: Addison-Wesley.

Light, P. H., & Mevarech, Z. R. (1992). Cooperative learning with computers:


An introduction. Learning and Instruction, 2, 155-159.

Lindlof, T. R. (1995). Qualitative communication research methods .


Thousand Oaks, CA: Sage Publications.

Littleton, K., Light, P. H., Joiner, R., & Messer, D. (1992). Pairing and
gender effects on childrens computer-based learning. European
Journal of Psychology of Education, 7(4), 311-324.

149
May, L. (1991). Sociolinguistic research on human-computer interaction: A
perspective from anthropology. Social Science Computer Review, 9(4),
529-540.

Mercer, N. (1994). The quality of talk in childrens joint activity at the


computer. Journal of Computer Assisted Learning, 10(1), 24-32.

Mercer, N., Phillips, T., & Somekh, B. (1991). Research note: Spoken
language and new technology (SLANT). Journal of Computer
Assisted Learning, 7, 195-202.

Mevarech, Z. R., & Light, P. H. (1992a). Cooperative learning with computers.


Learning and Instruction, 2(3), 155-285.

Mevarech, Z. R., & Light, P. H. (1992b). Peer-based interaction at the


computer: Looking backward, looking forward. Learning and
Instruction, 2(3), 275-280.

Newman, D., Griffin, P., & Cole, M. (1989). The Construction zone:
Working for cognitive change in school. Cambridge, England:
Cambridge University Press.

Norman, D. A. (1990). The design of everyday things. New York: Doubleday.

Norman, D. A., & Draper, S. W. (Eds.). (1986). User centered system


design. Hillsdale, NJ: Lawrence Erlbaum Associates.

Norman, M. A., & Thomas, P. J. (1991). Informing HCI designs through


conversation analysis. International Journal of Man Machine
Studies, 35(2), 235-250.

OMalley, C. (1992). Designing computer systems to support peer learning.


European Journal of Psychology of Education, 7(4), 339-252.

OMalley, C. (Ed.). (1995). Computer Supported Collaborative Learning


Berlin: Springer-Verlag.

Papert, S. (1980). Mindstorms: Children, computers, and powerful ideas.


New York: Basic Books, Inc.

Pea, R. D. (1993). Seeing what we build together: Distributed multimedia


learning environments for transformative communications. The
Journal of the Learning Sciences. 3(3)(285-299).

Reilly, B. (1992). The negotiations of group authorship among second


graders using multimedia composing software. Report ACOT-R-27.
Apple Computer.

150
Resnick, L. B., Levine, J. M., & Teasley, S. D. (Eds.). (1991). Perspectives on
socially shared cognition Washington, D.C.: American Psychological
Association.

Rieber, R. W., & Carton, A. S. (Eds.). (1987). The collected works of


L. S. Vygotsky New York: Plenum Press.

Roschelle, J. (1994). Collaborative inquiry: Reflections on Dewey and


learning technology. The Computing Teacher, 21(8), 8-9.

Roschelle, J., & Teasley, S. D. (1995). The construction of shared knowledge in


collaborative problem solving. In C. OMalley (Ed.), Computer Supported
Collaborative Learning (pp. 69-97). Berlin: Springer-Verlag.

Scrimshaw, P. (Ed.). (1993). Language, classrooms, and computers.


New York: Routledge.

Scott, W. A. (1955). Reliability of content analysis: The case for


nominal scale coding. Public Opinion Quarterly, 19, 321-325.

Skinner, B. F. (1938). The behavior of organisms: an experimental


analysis. New York: Appleton-Century-Crofts.

Smith, D. C., Cypher, A., & Spoher, J. (1994). KidSim: Programming agents
without a programming language. Communications of the ACM ,
37(7), 54-67.

Staggers, N., & Norcio, A. F. (1993). Mental models: Concepts for human-
computer interaction research. International Journal of Man-Machine
Studies, 38(4), 587-605.

Stoneman, Z., Brody, G. H., & MacKinnon, C. (1984). Naturalistic observations


of childrens activities and roles while playing with their siblings and
friends. Child Development, 55, 617-627.

Suchman, L. (1987). Plans and situated actions: The problem of human-


machine communication. Cambridge, England: Cambridge University
Press.

Suppe, F. (1977). The structure of scientific theories. Urbana, Illinois:


University of Illinois Press.

151
Teasley, S. D., & Roschelle, J. (1993). Constructing a joint problem space: The
computer as a tool for sharing knowledge. In S. P. Lajoie & S. J. Derry
(Eds.), Computers as cognitive tools. Hillsdale, New Jersey: Lawrence
Erlbaum Associates.

Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard University


Press.

Wade, A., Abrami, P. C., Poulsen, C., & Chambers, B. (1995). Current
resources in cooperative learning. Lanhan, Maryland: University Press of
America.

Wadsworth, B. J. (1971). Piagets theory of cognitive development. New York:


Longman, Inc.

Waern, Y. (1993). Varieties of learning to use computer tools. Computers in


Human Behavior, 9(2-3), 323-339.

Wenger, M. J. (1991). On the rhetorical contract in human-computer


interaction. Computers in Human Behavior, 7(4), 245-262.

Wertsch, J. V. (1985). Vygotsky and the social formation of the mind.


Cambridge, MA: Harvard University Press.

Wertsch, J. V. (1991). A sociocultural approach to socially shared cognition. In


L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.), Perspectives on
socially shared cognition (pp. 85-100). Washington, D.C.: American
Psychological Association.

Williams, N., & Hartley, P. (Eds.). (1990). Technology in human


communication. London: Pinter Publishers.

152

Вам также может понравиться