Вы находитесь на странице: 1из 41

Running head: EVALUATION OF NEW CURRICULUM 1

Results of a Quasi-Experimental Evaluation of a New Curriculum




Alaa Abd-El-Hafez and Donna M. D'Ambrosio
Long Island University





April 2014


Important!
This ANOVA project draft is provided as a personal and confidential form for DAP3.
It shall constitute serious academic dishonesty if you share this document or any portion of it
now or at any future time with any other person, including (but not limited to) current or
future students or faculty, in any form without prior written permission from Dr. Red Owl.
EVALUATION OF NEW CURRICULUM 2
Abstract
The purpose of the study was to conduct a quasi-experiment to determine the effectiveness of a
new curriculum compared to the existing curriculum as taught by 3 teachers in terms of students'
correct responses to a 40-item skills test. The data were obtained from a suburban public school.
These variables were curriculum and teachers. Students were randomly assigned to one of 6
experimental conditions representing all combinations of the curriculum factor and the teacher
factor (N = 24, each condition n = 4). A 2 x 3 analysis of variance (ANOVA) found statistically
significant main effects from curriculum (p = .053) where the old curriculum was superior to the
new, and also from teacher (p = .020), where one teacher was more effective than one other
teacher (p = .087). The study also found a statistically significant interaction effect (p = .049) for
combinations of curriculum and teacher. One teacher was more effective with the old
curriculum than another teacher using the new curriculum, and that same teacher was less
effective when teaching the new curriculum than teaching the old curriculum. The study finds
insufficient evidence to justify the expense and disruption of adopting the new curriculum.
These findings reflect the importance of professional development in training teachers in using
the new curriculum or leaving experienced teachers to develop their own curriculum. Additional
research is needed to better understand the impact of the new curriculum based on experience
level of teachers versus the hours of professional development.
Keywords: curriculum, factorial ANOVA, teacher effectiveness, suburban school, New York
EVALUATION OF NEW CURRICULUM 3
Results of a Quasi-Experimental Evaluation of a New Curriculum
The purpose of this study was to conduct a quasi-experiment to evaluate the effectiveness
of a new curriculum and to understand its potential use by different teachers in a suburban public
school in New York State. The following research questions guided the evaluation study:
RQ1: Is the new curriculum superior to the old curriculum in producing correct student
answers on the skills test in terms of main effects?
RQ2: Are there statistically significant main effects due to teachers in producing correct
student answers on the skills test?
RQ3: Is there a statistically significant interaction effect due to the interaction of
curriculum type and teacher that explains correct student answers on the skills test?
Methods
Sample and Unit of Analysis
The unit of analysis in this study is the curriculum, and the sample data were collected
and provided by the school district (Red Owl, 2014). The data include observations for 24
students, each of whom was taught by one of three teachers. The three teachers each taught four
students using the existing curriculum and four other students using the new curriculum.
Consistent with the design of the quasi-experiment, then, four students each were randomly
assigned to one of six conditions, as shown in Table 1.
[Insert Table 1 about here.]
Measures
The dependent variable in this analysis is the number of a students correct responses on
a skills test (correct), measured on a continuous scale with a possible range of 1-40.
EVALUATION OF NEW CURRICULUM 4
Two factor variables are included in the study. The first is a binomial factor (curric) that
indicates either the existing curriculum or the new curriculum. The second factor indicates the
students teacher during the quasi-experiment and is a polychotomous, categorical factor with
three levels, one for each of the three teachers in the study: Green, Jones, or Smith.
Data Analysis Methods
Descriptive statistics were calculated, factorial analysis of variance (ANOVA) was
conducted, and graphs were created using Stata/IC version 13.1. Microsoft Excel version 2007
was used in formatting the tabular data.
The ANOVA model specified as its outcome variable the number of correct answers on a
skills test (correct) and included both main and interaction effects for curriculum type (curric)
and teacher (teacher). The adequacy of the overall ANOVA model, the main effects, and the
interaction term were assessed based on the statistical significance of the respective F statistics
with an a priori acceptance criterion of p(F) .05.
To aid the description and visualization of the data, a series of horizontal box plots was
plotted for correct answers grouped by curriculum/teacher combinations. Interpretation of the
interaction effects observed in the ANOVA was supported by the development of a margins plot
with 95% CI bars.
Findings
Overall, the 24 students participating in the experiment earned a mean score of 25.50 (SD
= 6.70) correct answers on the skills test (correct). By curriculum, the mean score is 27.67 (SD =
5.57) for the existing curriculum and 23.33 (SD = 7.25) for the new curriculum. By teacher, the
mean scores are 30.13 (SD = 5.94) for Green, 23.00 (SD = 5.86) for Jones, and 23.38 (SD = 6.41)
EVALUATION OF NEW CURRICULUM 5
for Smith. Table 2 presents the mean correct scores and standard deviations within each of the
six experimental conditions, and Figure 1 displays horizontal box plots of these results.
[Insert Table 2 about here.]
[Insert Figure 1 about here.]
The overall factorial ANOVA model is statistically significant (F
(5, 18)
= 4.25, p = .010),
as shown in the results of the analysis presented in Table 3.
[Insert Table 3 about here.]
The main effect of teacher is statistically significant (F
(2, 18)
= 4.89, p = .020), and the
main effect of curriculum is weakly statistically significant (F
(1, 18)
= 4.29, p = .053). The post
hoc comparisons (after Bonferroni adjustment) of the main effects of teacher are shown in Table
4. These comparisons indicate that, after adjustment for multiple comparisons (k = 3), the only
statistically significant difference is M correct scores for Joness students < M correct scores for
Greens students, a difference which is weakly statistically significant (p = .087).
[Insert Table 4 about here.]
The interaction of curriculum and teacher is statistically significant (p = .049). Figure 2
displays an ANOVA margins plot with 95% CI bars, graphically reflecting the interaction effects
of curriculum and teacher. The post hoc comparisons (after Bonferroni adjustment) for the
various curriculum/teacher combinations are shown in Table 5. These comparisons indicate that,
after correction for multiple comparisons (k = 15), the only statistically significant contrasts
observed are for Smith Using the New Curriculum vs. Green Using the Old Curriculum (p =
.013) and for Smith Using the New Curriculum vs. Smith Using the Old Curriculum (p = .092).
[Insert Figure 2 about here.]
[Insert Table 5 about here.]
EVALUATION OF NEW CURRICULUM 6
Discussion
Each of the research questions that guided the current study is answered separately
below. The responses are based on the results of the ANOVA shown in Table 3, the ANOVA
margins plot displayed in Figure 2, and the post hoc comparisons shown in Tables 4 and 5.
Is the New Curriculum Superior to the Existing Curriculum, Independent
of the Teacher providing the Instruction?
When the teacher delivering the instruction is ignored, there is a weakly statistically
significant difference in students mean correct scores on the skills test based on the existing and
new curricula (F
(1, 18)
= 4.29, p = .053). The mean score of the students taught with the existing
curriculum (M = 27.67, SD = 5.57) is greater than the mean score of those exposed to the new
curriculum (M = 23.33, SD = 7.25) when the potential effects of their different teachers are not
taken into account.
Do Mean Correct Scores Differ by Teacher, Independent of the
Curriculum Type?
When the type of curriculum (i.e., existing vs. new) is ignored, there is a statistically
significant difference between the mean scores of groups of students based on the teachers to
whom they are assigned (F
(2, 18)
= 4.89, p = .020). Specifically, teacher Jones is less effective
than teacher Green (contrast = -7.13, p = .087), when the different curriculum types are not
considered. There is no statistically significant difference in the effectiveness of teacher Smith
vs. either teacher Green (p = .112) or teacher Jones (p = 1.000).
What Is the Effect of the New vs. Old Curricula After Accounting for
the Effects of the Teacher Delivering the Instruction?
EVALUATION OF NEW CURRICULUM 7
There is a statistically significant difference between the mean scores of groups of
students based on the combination of the specific curriculum with which they were taught
(existing vs. new) and the teachers to whose classes they were assigned (F
(2, 18)
= 3.60, p = .049).
When teacher Smith used the new curriculum, his students earned a mean score lower than that
achieved by teacher Greens students when she taught them with the old curriculum (M
Smith/New
=
17.75, SD
Smith/New
= 2.22, M
Green/Old
= 32.25, SD
Green/Old
= 4.57, contrast = -14.50, p = .013).
Further, teacher Smith was less effective when using the new curriculum than the old curriculum
(M
Smith/Old
= 29.00, SD
Smith/Old
= 2.58, M
Smith/New
= 17.75, SD
Smith/New
= 2.22, contrast = -11.25, p =
.092).
Conclusions and Implications
There is insufficient evidence to conclude that the new curriculum is more effective than
the old curriculum. Smith and Green are clearly more effective with the old curriculum than the
new curriculum. The clear conclusion and implication of the experiment is that more
professional development is required in training teachers with whichever curriculum is being
used. Another issue is that the existing curriculum is superior to the new curriculum. A third
issue that arises, is that if the old curriculum is effective in certain teachers, then forcing the new
curriculum upon them may result in undesired exam scores. If the results remain constant with
other schools and teachers, administrators should investigate the developing curriculum to teach
the subject at hand. Future research should focus on the factors that impact teachers' resistance
to the adoption of a new curriculum.

EVALUATION OF NEW CURRICULUM 8
References
Red Owl, R. H. (2014). Data set for quasi-experimental evaluation of a new curriculum [Stata
version 13.1 data set]. Retrieved from http://datalibrary.us/curriculum.dta

EVALUATION OF NEW CURRICULUM 9
Table 1

Assignment of Students to Curricula and Teachers in the Experiment


Teacher

Green Jones Smith Total
Old Curriculum 4 4 4 12
New Curriculum 4 4 4 12
Total 8 8 8 24

EVALUATION OF NEW CURRICULUM 10
Table 2

Mean Scores and SD by Experimental Condition

Teacher Old Curriculum New Curriculum Total
Green 32.25 28.00 30.13

4.57 7.02 5.94
Jones
21.75 24.25 23.00

2.99 8.18 5.85
Smith
29.00 17.75 23.38
2.58 2.22 6.41
Total
27.67 23.33 25.50
5.57 7.25 6.70

EVALUATION OF NEW CURRICULUM 11
Table 3

ANOVA Results

Source Partial SS df MS F p
Model 559.00 5 111.80 4.25 .010
Curriculum 112.67 1 112.67 4.29 .053
Teacher 257.25 2 128.63 4.89 .020
Curriculum x Teacher 189.08 2 94.54 3.60 .049
Residual 473.00 18 26.28


Total 1,032.00 23 44.87
Note. Boldface indicates effects that are statistically significant at the p .05 level.

EVALUATION OF NEW CURRICULUM 12
Table 4

Post Hoc Comparisons of Teacher Effects (After Bonferroni Adjustment)


Contrast t p 95% CI
Jones vs Green -7.13 -2.35 .087 [-15.03 , 0.78]
Smith vs Green -6.75 -2.22 .112 [-14.65 , 1.15]
Smith vs Jones 0.38 0.12 1.000 [-7.53 , 8.28]
Note. Boldface indicates contrasts that are statistically significant at p .05 level. SE = 3.02 for
all constructs.
EVALUATION OF NEW CURRICULUM 13
Table 5

Post Hoc Comparisons of Interaction Effects of Teacher and Curriculum (After Bonferroni Adjustment)


Contrast t p 95% CI
Green New vs Green Old -4.25 -1.17 1.000 [-16.50 , 8.00]
Jones Old vs Green Old -10.50 -2.90 .144 [-22.75 , 1.75]
Jones New vs Green Old -8.00 -2.21 .608 [-20.25 , 4.25]
Smith Old vs Green Old -3.25 -0.90 1.000 [-15.50 , 9.00]
Smith New vs Green Old -14.50 -4.00 .013 [-26.75 , -2.25]
Jones Old vs Green New -6.25 -1.72 1.000 [-18.50 , 6.00]
Jones New vs Green New -3.75 -1.03 1.000 [-16.00 , 8.50]
Smith Old vs Green New 1.00 0.28 1.000 [-11.25 , 13.25]
Smith New vs Green New -10.25 -2.83 .167 [-22.50 , 2.00]
Jones New vs Jones Old 2.50 0.69 1.000 [-9.75 , 14.75]
Smith Old vs Jones Old 7.25 2.00 .912 [-5.00 , 19.50]
Smith New vs Jones Old -4.00 -1.10 1.000 [-16.25 , 8.25]
Smith Old vs Jones New 4.75 1.31 1.000 [-7.50 , 17.00]
Smith New vs Jones New -6.50 -1.79 1.000 [-18.75 , 5.75]
Smith New vs Smith Old -11.25 -3.10 .092 [-23.50 , 1.00]
Note. Boldface indicates constructs that are statistically significant at p .05 level. SE = 3.62 for all contrasts.
EVALUATION OF NEW CURRICULUM 14

Figure 1. Horizontal boxplots of correct scores by three teachers using the old and new
curriculum. The whiskers represent the nearest adjacent observation to the upper or lower
boundaries.

15 20 25 30 35 40

Correct Answers
New Curriculum
Old Curriculum
Green Jones Smith
EVALUATION OF NEW CURRICULUM 15

Figure 2. ANOVA margins plot with 95% CI error bars reflecting the interaction effects of
curriculum and teachers. The post hoc comparisons for this plot are shown in Table 5.

10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
M
e
a
n

C
o
r
r
e
c
t

S
c
o
r
e
s

Green Jones Smith

Teacher
Old Curriculum New Curriculum

Error bars reflect 95% CIs.
EVALUATION OF NEW CURRICULUM 16
Appendix A
Stata Commands and Output

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
name: <unnamed>
log: C:\Users\Alaa\Dropbox\Alaa Donna\log-DAP3-20Mar14.smcl
log type: smcl
opened on: 20 Mar 2014, 22:25:22

. * STEP 1

.
. * Set up Stata session.

. scalar platform = "$S_OS"

.
. set more off

.
. if (platform == "Windows") set autotabgraphs on
EVALUATION OF NEW CURRICULUM 17

.
. clear

.
. graph drop _all

. * STEP 2

.
. * Load data from the web.

.
. use "http://datalibrary.us/curriculum.dta", clear

. * STEP 3

.
. * Describe the data set with a compact codebook.

.
. codebook, compact

EVALUATION OF NEW CURRICULUM 18
Variable Obs Unique Mean Min Max Label
-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
> ------------------
curric 24 2 1.5 1 2 Curriculum
teacher 24 3 2 1 3 Teacher
correct 24 20 25.5 14 38 Correct Answers
-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
> ------------------

. * STEP 4

.
. * Display cross-tab of curriculum and teacher to review

.
. * the assignment of subjects to experimental conditions.

.
. tab curric teacher

| Teacher
EVALUATION OF NEW CURRICULUM 19
Curriculum | Green Jones Smith | Total
---------------+---------------------------------+----------
Old Curriculum | 4 4 4 | 12
New Curriculum | 4 4 4 | 12
---------------+---------------------------------+----------
Total | 8 8 8 | 24


. * STEP 5

.
. * Create frequency tables of DV and 2 IVs.

.
. tab1 curric teacher correct

-> tabulation of curric

Curriculum | Freq. Percent Cum.
---------------+-----------------------------------
Old Curriculum | 12 50.00 50.00
New Curriculum | 12 50.00 100.00
---------------+-----------------------------------
EVALUATION OF NEW CURRICULUM 20
Total | 24 100.00

-> tabulation of teacher

Teacher | Freq. Percent Cum.
------------+-----------------------------------
Green | 8 33.33 33.33
Jones | 8 33.33 66.67
Smith | 8 33.33 100.00
------------+-----------------------------------
Total | 24 100.00

-> tabulation of correct

Correct |
Answers | Freq. Percent Cum.
------------+-----------------------------------
14 | 1 4.17 4.17
15 | 1 4.17 8.33
17 | 1 4.17 12.50
18 | 1 4.17 16.67
19 | 1 4.17 20.83
20 | 1 4.17 25.00
EVALUATION OF NEW CURRICULUM 21
21 | 2 8.33 33.33
23 | 2 8.33 41.67
24 | 1 4.17 45.83
25 | 2 8.33 54.17
26 | 1 4.17 58.33
27 | 1 4.17 62.50
28 | 1 4.17 66.67
30 | 1 4.17 70.83
31 | 1 4.17 75.00
32 | 1 4.17 79.17
33 | 2 8.33 87.50
34 | 1 4.17 91.67
35 | 1 4.17 95.83
38 | 1 4.17 100.00
------------+-----------------------------------
Total | 24 100.00

. * STEP 6

.
. * Calculate descriptive statistics for the dependent variable.

.
EVALUATION OF NEW CURRICULUM 22
. summarize correct

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 24 25.5 6.698475 14 38

. * STEP 7

.
. * Calculate mean correct scores with SD by factor levels of

.
. * curric, teacher, and interaction.

.
. bysort curric: summarize correct

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> curric = Old Curriculum

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
EVALUATION OF NEW CURRICULUM 23
correct | 12 27.66667 5.565042 18 38

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> curric = New Curriculum

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 12 23.33333 7.253004 14 35


.
. bysort teacher: summarize correct

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> teacher = Green

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 8 30.125 5.938675 21 38

EVALUATION OF NEW CURRICULUM 24
-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> teacher = Jones

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 8 23 5.8554 14 34

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> teacher = Smith

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 8 23.375 6.412878 15 32


.
. bysort curric teacher: summarize correct

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> curric = Old Curriculum, teacher = Green
EVALUATION OF NEW CURRICULUM 25

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 4 32.25 4.573474 27 38

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> curric = Old Curriculum, teacher = Jones

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 4 21.75 2.986079 18 25

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> curric = Old Curriculum, teacher = Smith

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 4 29 2.581989 26 32

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
EVALUATION OF NEW CURRICULUM 26
-> curric = New Curriculum, teacher = Green

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 4 28 7.023769 21 35

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> curric = New Curriculum, teacher = Jones

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 4 24.25 8.180261 14 34

-----------------------------------------------------------------------------------------------------------------------
-----------------------------------------------
-> curric = New Curriculum, teacher = Smith

Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
correct | 4 17.75 2.217356 15 20


EVALUATION OF NEW CURRICULUM 27
. * STEP 8

.
. * Create a table of M and SD correct responses

.
. * by experimental condition.

.
. table teacher curric, contents(mean correct sd correct) format(%6.2f)

------------------------------------------
| Curriculum
Teacher | Old Curriculum New Curriculum
----------+-------------------------------
Green | 32.25 28.00
| 4.57 7.02
|
Jones | 21.75 24.25
| 2.99 8.18
|
Smith | 29.00 17.75
| 2.58 2.22
EVALUATION OF NEW CURRICULUM 28
------------------------------------------

. * STEP 9

.
. * Create horizontal box plots of correct by teacher and curric.

.
. * The following block of code must be on a single line!

.
. graph hbox correct, over(teacher) over(curric) asyvars legend(rows(1)) scheme(s1color) ytitle(" " "Correct Answers" )
name(hboxplot, replace)

. graph save hboxplot "C:\Users\Alaa\Dropbox\Alaa Donna\hboxplot-AA-20march14.gph"
(file C:\Users\Alaa\Dropbox\Alaa Donna\hboxplot-AA-20march14.gph saved)
EVALUATION OF NEW CURRICULUM 29

. * STEP 10

.
. * Conduct factorial anova of correct on teacher and curric with

.
. * the interaction term.

.
. anova correct curric teacher curric#teacher

Number of obs = 24 R-squared = 0.5417
Root MSE = 5.12619 Adj R-squared = 0.4144

Source | Partial SS df MS F Prob > F
15 20 25 30 35 40

Correct Answers
New Curriculum
Old Curriculum
Green Jones Smith
EVALUATION OF NEW CURRICULUM 30
---------------+----------------------------------------------------
Model | 559 5 111.8 4.25 0.0099
|
curric | 112.666667 1 112.666667 4.29 0.0530
teacher | 257.25 2 128.625 4.89 0.0201
curric#teacher | 189.083333 2 94.5416667 3.60 0.0485
|
Residual | 473 18 26.2777778
---------------+----------------------------------------------------
Total | 1032 23 44.8695652

. * STEP 11

.
. * Estimate marginal effects of the interaction term.

.
. set cformat %6.2f

.
. margins teacher#curric

Adjusted predictions Number of obs = 24
EVALUATION OF NEW CURRICULUM 31

Expression : Linear prediction, predict()

---------------------------------------------------------------------------------------
| Delta-method
| Margin Std. Err. t P>|t| [95% Conf. Interval]
----------------------+----------------------------------------------------------------
teacher#curric |
Green#Old Curriculum | 32.25 2.56 12.58 0.000 26.87 37.63
Green#New Curriculum | 28.00 2.56 10.92 0.000 22.62 33.38
Jones#Old Curriculum | 21.75 2.56 8.49 0.000 16.37 27.13
Jones#New Curriculum | 24.25 2.56 9.46 0.000 18.87 29.63
Smith#Old Curriculum | 29.00 2.56 11.31 0.000 23.62 34.38
Smith#New Curriculum | 17.75 2.56 6.93 0.000 12.37 23.13
---------------------------------------------------------------------------------------

. * STEP 12

.
. * Produce a margins plot of the interaction effects

.
. * with error bars.
EVALUATION OF NEW CURRICULUM 32

.
. * The following block of code must be on a single line!

.
. marginsplot, legend(rows(1)) plotopts(msymbol(Oh) msize(vlarge)) plot2opts(lcolor(brown) mcolor(brown) msymbol(Dh)
msize(large)) ci2opts(lcolor(brown)) ylabel(10(2)
> 40, angle(horizontal) labsize(vsmall)) xlab(1(1)3, labsize(vsmall)) ytitle("Mean Correct Scores" " ") xtitle(" "
"Teacher") title("") note(" " "Error bars reflect 9
> 5% CIs.") scheme(s1color) name(marginsplot, replace)

Variables that uniquely identify margins: teacher curric

. graph save marginsplot "C:\Users\Alaa\Dropbox\Alaa Donna\marginsplot-AA-20march14.gph"
(file C:\Users\Alaa\Dropbox\Alaa Donna\marginsplot-AA-20march14.gph saved)
EVALUATION OF NEW CURRICULUM 33

. * STEP 13

.
. * Perform post-hoc comparisons on teacher.

.
. * Because teacher has more than two levels,

.
. * we need Bonferroni adjustment.

.
. pwmean correct, over(teacher) effects cimeans bonferroni

10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
M
e
a
n

C
o
r
r
e
c
t

S
c
o
r
e
s

Green Jones Smith

Teacher
Old Curriculum New Curriculum

Error bars reflect 95% CIs.
EVALUATION OF NEW CURRICULUM 34
Pairwise comparisons of means with equal variances

over : teacher

--------------------------------------------------------------
| Unadjusted
correct | Mean Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
teacher |
Green | 30.13 2.15 25.66 34.59
Jones | 23.00 2.15 18.53 27.47
Smith | 23.38 2.15 18.91 27.84
--------------------------------------------------------------

---------------------------
| Number of
| Comparisons
-------------+-------------
teacher | 3
---------------------------

---------------------------------------------------------------------------------
| Bonferroni Bonferroni
EVALUATION OF NEW CURRICULUM 35
correct | Contrast Std. Err. t P>|t| [95% Conf. Interval]
----------------+----------------------------------------------------------------
teacher |
Jones vs Green | -7.13 3.04 -2.35 0.087 -15.03 0.78
Smith vs Green | -6.75 3.04 -2.22 0.112 -14.65 1.15
Smith vs Jones | 0.38 3.04 0.12 1.000 -7.53 8.28
---------------------------------------------------------------------------------

. * STEP 14

.
. * Perform post-hoc comparisons on teacher#curric interaction term.

.
. * Because teacher#curric has more than two levels, we need

.
. * multiple comparisons with Bonferroni adjustment.

.
. * The following block of code must be on a single line!

.
EVALUATION OF NEW CURRICULUM 36
. pwmean correct, over(teacher curric) effects cimeans mcompare(bonferroni)

Pairwise comparisons of means with equal variances

over : teacher curric

-----------------------------------------------------------------------
| Unadjusted
correct | Mean Std. Err. [95% Conf. Interval]
----------------------+------------------------------------------------
teacher#curric |
Green#Old Curriculum | 32.25 2.56 26.87 37.63
Green#New Curriculum | 28.00 2.56 22.62 33.38
Jones#Old Curriculum | 21.75 2.56 16.37 27.13
Jones#New Curriculum | 24.25 2.56 18.87 29.63
Smith#Old Curriculum | 29.00 2.56 23.62 34.38
Smith#New Curriculum | 17.75 2.56 12.37 23.13
-----------------------------------------------------------------------

-----------------------------
| Number of
| Comparisons
---------------+-------------
EVALUATION OF NEW CURRICULUM 37
teacher#curric | 15
-----------------------------

-------------------------------------------------------------------------------------------------------------------
| Bonferroni Bonferroni
correct | Contrast Std. Err. t P>|t| [95% Conf. Interval]
--------------------------------------------------+----------------------------------------------------------------
teacher#curric |
(Green#New Curriculum) vs (Green#Old Curriculum) | -4.25 3.62 -1.17 1.000 -16.50 8.00
(Jones#Old Curriculum) vs (Green#Old Curriculum) | -10.50 3.62 -2.90 0.144 -22.75 1.75
(Jones#New Curriculum) vs (Green#Old Curriculum) | -8.00 3.62 -2.21 0.608 -20.25 4.25
(Smith#Old Curriculum) vs (Green#Old Curriculum) | -3.25 3.62 -0.90 1.000 -15.50 9.00
(Smith#New Curriculum) vs (Green#Old Curriculum) | -14.50 3.62 -4.00 0.013 -26.75 -2.25
(Jones#Old Curriculum) vs (Green#New Curriculum) | -6.25 3.62 -1.72 1.000 -18.50 6.00
(Jones#New Curriculum) vs (Green#New Curriculum) | -3.75 3.62 -1.03 1.000 -16.00 8.50
(Smith#Old Curriculum) vs (Green#New Curriculum) | 1.00 3.62 0.28 1.000 -11.25 13.25
(Smith#New Curriculum) vs (Green#New Curriculum) | -10.25 3.62 -2.83 0.167 -22.50 2.00
(Jones#New Curriculum) vs (Jones#Old Curriculum) | 2.50 3.62 0.69 1.000 -9.75 14.75
(Smith#Old Curriculum) vs (Jones#Old Curriculum) | 7.25 3.62 2.00 0.912 -5.00 19.50
(Smith#New Curriculum) vs (Jones#Old Curriculum) | -4.00 3.62 -1.10 1.000 -16.25 8.25
(Smith#Old Curriculum) vs (Jones#New Curriculum) | 4.75 3.62 1.31 1.000 -7.50 17.00
(Smith#New Curriculum) vs (Jones#New Curriculum) | -6.50 3.62 -1.79 1.000 -18.75 5.75
(Smith#New Curriculum) vs (Smith#Old Curriculum) | -11.25 3.62 -3.10 0.092 -23.50 1.00
EVALUATION OF NEW CURRICULUM 38
-------------------------------------------------------------------------------------------------------------------

. log close
name: <unnamed>
log: C:\Users\Alaa\Dropbox\Alaa Donna\log-DAP3-20Mar14.smcl
log type: smcl
closed on: 20 Mar 2014, 22:35:26

EVALUATION OF NEW CURRICULUM 39
Articles Used as Models of Reporting Analysis of Variance
Galguera, T. (1998). Students' attitudes toward teachers' ethnicity, bilinguality, and gender.
Hispanic Journal of Behavioral Sciences, 20, 411-428. doi:10.1177/07399863980204001
Gammage, K. L., Carron, A. V., & Estabrooks, P. A. (2001). Team cohesion and individual
productivity: The influence of the norm for productivity and the identifiability of
individual effort. Small Group Research, 32, 3-18. doi:10.1177/104649640103200101
Geringer, J. M. (2010). Musicians' preferences for tempo and pitch levels in recorded orchestral
music. Journal of Research in Music Education, 58, 294-308.
doi:10.1177/0022429410380464
Kaniuka, T. S. (2010). Reading achievement, attitude toward reading, and reading self-esteem of
historically low achieving students. Journal of Instructional Psychology, 37, 184-188.
Retrieved from
http://cwplib.proxy.liu.edu/login?url=http://search.proquest.com.cwplib.proxy.liu.edu/do
cview/613407214?accountid=12142
Lupini, W. H., & Zirkel, P. A. (2003). An outcomes analysis of education litigation. Educational
Policy, 17, 257-279. doi:10.1177/0895904803017002004
Murrell, L. E., & Hughey, A. W. (2003). The effects of job attributes, institutional mission
emphasis, and institution type on perceptions of the provost position. Educational
Administration Quarterly, 39, 533-565. doi:10.1177/0013161X03255562
Newton, R. M. (2006). Does recruitment message content normalize the superintendency as
male? Educational Administration Quarterly, 42, 551-577.
doi:10.1177/0013161X06291413
EVALUATION OF NEW CURRICULUM 40
Shi, L. (2004). Textual borrowing in second-language writing. Written Communication, 21, 171-
200. doi:10.1177/0741088303262846
Tucker, L. W., & Parks, J. B. (2001). Effects of gender and sport type on intercollegiate athletes'
perceptions of the legitimacy of aggressive behaviors in sport. Sociology of Sport
Journal, 18, 403-413. Retrieved from Human Kinetics database.
Williams, L. R. (2005). Effect of music training and musical complexity on focus of attention to
melody or harmony. Journal of Research in Music Education, 53, 210-221.
doi:10.1177/002242940505300303

EVALUATION OF NEW CURRICULUM 41
Researchers Contributions
Alaa K. Abd-El-Hafez
Abd-El-Hafez read and reflected on Galguera (1998), Geringer (2010), Lupini and Zirkel
(2003), Shi (2004), and Williams (2005) of analysis of variance. She met with the team on
March 23 (3.5 hours online), on March 23 (1.5 hours online with Red Owl), on March 29 (1.5
hours online), on March 30 (1.0 hours reviewing corrections), and on April 1 (.75 hours
finalizing the project). She ran all of the analysis of variance code herself even though the other
member of the team did the same. Abd-El-Hafez participated in every aspect of the project. She
also assisted in reviewing the draft and final versions with Red Owl's APA tips and the APA
Manual to make sure the documents were fully compliant with APA guidelines.
The total time devoted to this project by Abd-El-Hafez was 8 hours and 15 minutes.

Donna M. D'Ambrosio
D'Ambrosio read and reflected on Gammage, Carron, and Estabrooks (2001), Kaniuka
(2010), Murrell and Hughey (2003), Newton (2006), and Tucker and Parks (2001) as models of
analysis of variance. She met with the team on March 23 (3.5 hours online), on March 23 (1.5
hours online with Red Owl), on March 29 (1.5 hours online), on March 29 (1.25 hours online
with R. Wottawa), on March 30 (1.0 hours reviewing corrections), and on April 1 (.75 hours
finalizing the project. She ran all of the analysis of variance code herself even though the other
member of the team did the same. D'Ambrosio participated in every aspect of the project. She
also assisted in reviewing the draft and final versions with Red Owl's APA tips and the APA
Manual to make sure the documents were fully compliant with APA guidelines.
The total time devoted to this project by D'Ambrosio was 9 hours and 30 minutes.

Вам также может понравиться