Вы находитесь на странице: 1из 19

EXPLORING LEADERSHIP ROLE IN SCHOOL EFFECTIVENESS

AND THE VALIDATION OF MODELS OF PRINCIPALS EFFECTS


ON STUDENTS ACHIEVEMENT
Andreas Kythreotis, Ph.D, University of Cyprus, Nicosia, Cyprus
Petros Pashiardis, Associate Professor, University of Cyprus, Nicosia, Cyprus

For many years now, researchers in the area of educational leadership have attempted to
identify links between educational leadership and school effectiveness research. This
phenomenon is mainly due to the perception that educational leaders, especially school
principals, affect school effectiveness (Edmonds, 1979; Fuller, 1987; Rutter et al, 1979;
Mortimore, Sammons, Ecob & Stoll 1988; Lezotte, 1989; Levin & Lezotte 1990; Reynolds &
Cuttance, 1992; Cheng, 1994a; Pashiardis, 1995; Pashiardis, 1998; Pashiardis, 2004).
However, two main issues have arisen: Firstly, what positions or roles do leaders have in a
school organization? Secondly, under what conditions does school leadership affect students
achievement and to what extent?

In relation to the first issue, it is interesting that the vast majority of researchers have been
involved mainly in studies of principals leadership styles. However, recent research has also
been concerned with the leadership of persons who have other roles in a school organization
such as teachers, (Cheng, 1994b; Ogawa & Bossert 1995; Pounder, Ogawa & Adams, 1998;
Deal & Peterson, 1999, Harris & Muijs, 2003, Pashiardis, 2004). As a consequence, it is
necessary to investigate the parameters of many individuals leadership styles in order to give
a more complete overview of school leadership.
The second issue is more complex due to contradictory findings concerning effects of
leadership on students achievement. Some studies found no influence whereas others
identified some effects (Edmonds, 1979; Andrews & Soder, 1987; Heck 1992; Johnson
1993).

The meta-analyses, conducted by Hallinger & Heck (1996; 1998) and Witziers, Bosker &
Krger (2003), emphasized at least two important elements that differentiate the results
among many studies. Firstly, the different educational systems and cultures among the
various countries lead to different results (also in Pashiardis, Thody, Papanaoum &
Johansson, 2003). Secondly, the absence of intermediate variables between principals
leadership and students achievement tends to find no links between them (also in Teddlie &
Reynolds, 2000).

Based on the above, this study aims to examine the principals contribution to school
effectiveness among primary schools in Cyprus. This work has adopted specific theories with
regards to these three variables. More specifically, it is based on the Bolman & Deals (1984;
1991; 1997) theoretical model of leadership. Moreover, it adopted the Feitler & Gudgel
(1994) model of organizational culture of schools and the Midgley, Maehr, Hicks, Roeser,
Urdan, Anderman & Kaplans (2001) model of goal theory and school culture to investigate
the learning culture of classrooms. Finally, it used the integrated model of school
effectiveness, which is proposed by Hoy & Miskel (1996). The adoption of specific
theoretical models was based on two main common characteristics of leadership, culture and
effectiveness: At first, these three concepts are multidimensional, as each of them consists of
many dimensions. For example, integrated models of school effectiveness indicate many
criteria of inputs, transformations and outputs (Hoy & Miskel, 2001). Second, these three
2

concepts are presented at many levels in schools. For example, leaders can be identified at
school level (e.g. principal) or at classroom level (e.g. teachers) (Cheng, 1994b). Moreover,
there are many cultures in a school such as organizational culture, teachers culture, students
culture, classroom culture (Maehr & Midgley, 1996). Finally, multilevel models indicate
many levels of effectiveness such as the student level, the classroom level and the school
level (Creemers, 1994).

The first concept that we examined in this study was school leadership. The examination of
this concept involves many difficulties because of the large number of its definitions (Hoy &
Miskel, 2001). One theory that views the multidimensional nature of leadership, and
especially effective leadership, is the Bolman & Deals theory of leadership frames (1984;
1991; 1997). This theoretical framework is based on the assumption that four leadership
dimensions play important roles in effective leadership:
(1) The structural frame, which emphasizes goals, planning, and coordination;
(2) The human resource frame, which is sensitive to the human needs of others;
(3) The political frame, which recognizes the ways that people seek to advance their own
interests; and
(4) The symbolic frame, which focuses in the rituals, myths and ceremonies that give
meaning to organizational culture.

The evidence of several recent studies supports the main assumptions of the theory. At the
same time research identified new elements that complete the model. For example, Bolman &
Deal (1991; 1992a; 1992b) found that the leaders ability to use many frames is highly
correlated with their effectiveness.
A second concept that concerns this study is organizational culture. Again, the large number
of definitions contributes to the difficulty of studying this concept as well. Based mainly on
Schein definition (1992), Hoy & Miskel (2001) defined culture as the shared orientations
that hold the unit together and give it a distinctive identity (p.129). However, substantial
disagreement arises about what is shared (norms, values, philosophies, perspectives, beliefs,
expectations, attitudes, myths, or ceremonies). Another problem is determining the intensity
of shared orientations of organizational members.

Culture is very important for an organization as it affects significantly every aspect of it. such
as stability, cohesion, unity and ability for adjustments in an organization. Moreover, a
number of studies attempts to link school culture and effectiveness (Edmonds, 1979; Fyans &
Maehr; 1990; Cheng, 1993). As this study is interested in school culture, we examined
theoretical models of culture that had been adjusted for schools. The results of this review
identified only a small number of models. One approach that includes many dimensions, such
as teamwork and cooperation, communication, decision making, change and innovation,
responsibility and commitment, and vision and goals, of organizational culture at school level
is the model proposed by Feitler & Gudgel (1994).

The existence of many cultures at the school level has already been recognized (Maehr &
Midgley, 1996). However, there is a lack of theoretical models related to school learning
culture. An attempt to measure school learning culture has been made by Barnett,
McCormick & Conners in Australia (1999). This study used a model proposed by Midgley,
Maehr, Hicks, Rosser, Anderman & Kaplan (1996), who developed an instrument measuring
school learning culture and student motivation. There are many dimensions in their model
that were related to school learning culture such as academic emphasis, academic efficiency,
3

academic novelty, cheating behavior, disruptive behavior and success. An effort to validate
this model at classroom level could be an interesting undertaking.

The third concept that concerns this study is school effectiveness. Many difficulties exist
about the right definition of organizational effectiveness. According to Cameron (1984), the
theories of organizational effectiveness and the list of criteria are neither necessary nor
sufficient to evaluate this concept. In addition the war between the supporters of school
effectiveness and the supporters of school quality strengthen the confusion. A serious attempt
to fill the gap between effectiveness and quality has been made by Hoy & Miskel (2001) who
developed an integrated goal and resource system model of effectiveness. This model
addresses the importance of all aspects of a social system including effectiveness and quality
of inputs, transformation and outcomes. It could be argued here that the large numbers of
inputs, aspects of transformational processes and outcomes creates serious difficulties to have
an overall examination and validation of this model in one single study. As a result, Hoy &
Miskel proposed a small number of variables that could be included in an investigation.

The above theoretical frame attempts to indicate the need to investigate the principals effects
on student achievement because of contradictory findings among many studies. A serious
effort has been undertaken by Pitner (1988), who identified three main models, which
indicates the possible effects of principals leadership on student achievement:
1. Direct effects of leadership on students achievement (direct model);
2. Indirect effects of leadership effects on students achievement through intermediate
variables (indirect model)
3. Interactions between leadership, other variables and students achievement
(reciprocal model).
Literature review indicated some relationships between leadership and culture (Sashkin &
Sashkin, 1990; Sashkin & Walberg, 1993; Cheng, 1994) or between culture and effectiveness
(Levine & Lezotte, 1990; Cheng, 1993). In addition, this review showed the importance that
leadership, culture and effectiveness have for the existence, the everyday life and the
development of an organization such as a school. In consequence, the illumination of the
relationships among leadership, culture and effectiveness as well as the use of culture as the
intermediate variable between leadership and students achievement could be useful for both
theory and practice. In addition, the use of complex statistical analysis such as multilevel
analysis and structural equation modeling could contribute more effectively to this attempt
(Hallinger & Heck, 1996, 1998; Teddlie & Reynolds, 2000; Witziers, Bosker & Krger,
2003).

Research aims
Thus, the main purpose of the study was to investigate the principals contribution to school
effectiveness in primary schools in Cyprus and to validate three models of the principals
effects on the students achievement. More specifically, this study aims to:
1. identify the parameters, which determine principals leadership style, teachers
leadership style, schools organizational culture and classrooms learning cultures.
2. examine the effects of leadership style and culture at both classroom and school level
on schools effectiveness, so as to develop a multilevel model of school effectiveness
and to validate the main models of principals effects on students achievement in the
Greek language and Mathematics.
4

Methods
The main principles which were adopted in the methodology
The selection of specific methods used for the purpose of gathering data, which could answer
the questions of this study, depended on a series of basic principles of methodology. The
presentation of these principles is important, because it allows the evaluation of the quality of
the data of this study.
1. The longitudinal nature of this study.
2. The selection of the instruments for the measurement of the variables of the study.
3. The use of methods of evaluation of others (hetero-evaluation) for the measurement of the
explanatory variables.
4. The need for systematic control of generizibility of the results, which are selected from the
perceptions of group members.
6.The use of techniques of sampling which permits the maximum deviation at both classroom
and school level. The population in this study consisted of all primary schools in Cyprus,
which had at least two classes of sixth grade. In consequence, the selection of the schools was
based on a stratified sampling.

The variables of the study


Pupils final achievement scores in Greek Language and Mathematics were the two depended
variables for this study. Pupils prior achievement in the above two subjects, pupils
background factors, gender, leadership style and culture were the explanatory variables. As
two groups of data existed (one for Greek Language and one for Mathematics), two different
processes of multilevel analysis were conducted at three different levels: student level,
classroom level and school level. These variables are addressed in detail below.

Dependent Variables: Cognitive Outcome of Schooling


Students achievement scores in the two tests in Greek Language and Mathematics at the end
of the school year were considered as the two depended variables. These two tests were
developed in Cyprus (Kyriakides, 2005). The construction of the tests was subject to control
for reliability and validity. Structural equation modeling procedures were used to examine the
construct validity of each test and the fit statistics for both the Greek Language Test (2 =
93.1, d.f. = 32, p<.001, RMSEA = .034, CFI = .962) and Mathematics Test (2 = 491.7, d.f. =
165, p<.001, RMSEA = .027, CFI = .972) were acceptable. The reliability of the data of the
two tests was measured by calculating the relevant values of Cronbachs alpha for the scales.
These values were higher than .82 and this implies that we can be confident about the
reliability of the measures used to collect data on students knowledge in Greek language and
Mathematics (Cronbach, 1990).

Explanatory Variables
All the explanatory variables were categorized at three levels. First, prior knowledge and
background factors (socio-economic status and gender) were the three variables at student
level. Second, variables of the teachers leadership style and classroom learning culture were
included at classroom level. Third, variables of the principals leadership style, the principals
perceived effectiveness and school organizational culture were included at school level.

Explanatory Variables at Pupil level


Prior Knowledge in Greek Language and Mathematics. Students achievement on the two
tests in Greek language and Mathematics at the beginning of the last year of primary school
was considered as the prior knowledge.
5

Pupils Background Factors. In addition to students achievement, information was collected


on two further pupils background factors: pupils gender (0 = boys, 1 = girls) and pupils
socio-economic status (SES). Four variables were available: fathers and mothers
educational level (i.e., graduate of primary school, graduate of secondary school or graduate
of a college/university), the social status of fathers job and the social status of mothers job.
Following the classifications of occupations used by the Ministry of Finance of Cyprus, it
was possible to classify parents occupations into three groups which had relatively similar
sizes: occupations held by the working class, occupations held by the middle class and
occupations held by the upper-middle class. Representative parental occupations for the
working class are: farmers, truck drivers, machine operators in a factory; for the middle class:
police officers, teachers, bank officers; and for the upper-middle class: doctors, lawyers,
business executives. Relevant information for each child was obtained from school records.
Following the standardized values of the aforementioned four variables were calculated,
resulting in the SES indicator.

Explanatory Variables at Classroom Level


Teachers leadership style. Teachers leadership style was measured in accordance with the
first section of the Greek version of Leadership Orientations (Bolman & Deal, 1991,
1992a). The human resource frame subscale consisted of 7 items and the symbolic frame
subscale consisted of 9 items. The questionnaire was administered to all students asking them
to indicate their teachers leadership style (e.g. My teacher helps us to solve our problems).
The scales used had a range from 1 to 5 (1=never, and 5=always).
Classroom learning culture. Classroom learning culture was measured in accordance with the
Greek version of Patterns of Adaptive Learning Scales (Midgley et al., 1996; 1997; 2000),
which included nine factors. The questionnaire, which consisted of 40 items, was
administered to all teachers asking them to describe their school culture (e.g. The school
encourages ideas for improvement). The scales had a range from 1 to 5 (1= Never, and
5=Always). Nine interpretable factors (alpha coefficient ranging from .68 to .78) were
identified in accordance with the English version of the questionnaire and the relative
theoretical model.

Explanatory Variables at School Level


Principals leadership style. Principals leadership style was measured in accordance with the
Greek version of Leadership Orientations (Bolman & Deal, 1991) consisting of three
sections. The items in the first two sections asked teachers to describe their own principals
leadership style (structural, human resource, political and symbolic). The first section
included 19 items using Likert - type questions. The scales had a range from 1 to 5 (1= Never,
and 5=Always). The second section included six forced choice questions. Questions in both
sections asked teachers to describe their own principals style. Finally, the third section
included two items that asked teachers to rate their principals effectiveness as a manager and
as a leader in comparison with other principals having a similar level of experience and
responsibility (1= Never, and 5=Always).
School organizational culture. School organizational culture was measured by the Greek
version of the Organizational Culture Questionnaire, which included nine factors. The
questionnaire, which consisted of 40 items, was administered to all teachers asking them to
describe their school culture (e.g. The school encourages ideas for improvement). The
scales used had a range from 1 to 5 (1=Very Little Extent, and 5=Very Great Extent). Nine
interpretable factors (alpha coefficient ranging from .82 to .92) were identified in accordance
with the English version of the questionnaire and the relative theoretical model.
6

The process of development of the Greek version of each questionnaire


The development of the Greek version of each of the instruments measuring leadership style
and culture followed three phases. At the beginning, the translation of each questionnaire
from the English language to the Greek language was undertaken. Then, the examinations of
validity, generalizability and reliability were conducted. Specifically, the following
examinations were made: content, face and construct validity of the instruments,
generalizability of the results and reliability of the measurements of each scale of the
instruments. At the end, the Greek version of each questionnaire was finalized.

The process of data collection


Based on the main purpose and the three aims of the study as well as the five main principles
of the methodology, the process of data collection was divided into three main phases. With
regards to the first aim of the study, data were used for the examination of generalizability of
the results, construct validity and reliability of the results of the instruments (validation of the
questionnaires). Concerning the second aim of the study, data were used to examine the
effects of leadership style and culture on student achievement. Finally, with regards to the
third aim of the study, data were used to examine the relationship between leadership style
and culture. The first phase included the initial distribution of the two tests of achievement in
Greek Language and Mathematics to the sample of the study. The second phase was related
to the examinations of content and face validity of the four questionnaires of leadership style
and culture among a small sample. The third phase was connected to the final distribution of
the two tests and the four questionnaires to the sample of the study.

The population and the sample of the study


The sample of the study was composed of 25 primary schools, which included 61 classrooms
of sixth grade from a total population of 116 schools (from the 344 primary schools in Cyprus
during the school year 2002 2003). This sample represented 22% of the population of
primary schools in Cyprus that had at least two classrooms of sixth grade. 1224 students
responded to the test in Greek language, 1992 students responded to the test in Mathematics,
1043 students responded to teachers leadership questionnaire and 1153 students responded
to classroom culture questionnaire. Moreover, 180 teachers responded to both principals
leadership style questionnaire and school culture questionnaire.

The process of statistical analysis of the data of the study


The statistical analyses of the data were conducted in accordance with the aims of the study.
As a consequence, the analysis followed three phases, (a) the examination of validity and
reliability, (b) the multilevel analysis, and (c) the structural equation modeling.

Examination of validity and reliability


The first aim of the study was related to the parameters, which determine leadership style and
culture in schools. Due to this reason, statistical analyses for the examination of construct
validity, generalizability and reliability related to the four questionnaires were conducted.
The results of these examinations have already been presented above in the methodology
section.

Multilevel Analysis
The second aim of the study was related to the development of a multilevel model of school
effectiveness, which could be defined from the effects of leadership style and culture in
schools. Multilevel analysis is the appropriate method because of multistage sampling since
students are nested within classes and classes within schools. The statistical package
7

MlwiN was used for the multilevel analysis. This dependency has an important
consequence. If students achievement within a class or a school has a small range,
institutional factors at class or school level may have contributed to it (Snijders & Bosker,
1999). Multilevel analysis is a method of analysis, which cannot cope with missing data
(Bryk & Raudenbush, 1992). Students and teachers were asked to complete the
questionnaires when they were at school so that we had complete data from 22 schools, 53
classes and 1097 students.

Results
The influences of leadership style and culture on school effectiveness
There is substantial agreement that researchers should attempt to consider the multilevel
structure of the data collected in order to examine school effectiveness, and thus multilevel
modeling should be used as the method of analysis for educational effectiveness studies
(Godstein, 2003). Hence, MlwinN (Goldsein, Rasbsh, Plewis, Draper, Browne, Yang,
Woodhouse & Healy, 1998) was used to analyze our data because the observations are
interdependent and because of multistage sampling, since students are nested within schools.
Specifically, in order to examine the extent to which the variables of leadership style and
school culture show effects upon each of the two dependent variables, the analyses were
performed separately for each dependent variable. The first step in the analysis was to
determine the variance at student, classroom and school level without explanatory variables
(empty model). In subsequent steps, explanatory variables at different levels were added
starting at student level. Explanatory variables, except grouping variables, were centered as
Z-scores with a mean of 0 and a standard deviation of 1. This is a way of centering on a grand
mean and produces effects that are comparable. Grouping variables were entered as dummy
variables with one of the groups as baseline (boys=0, girls=1).
Effectiveness in Greek language
As mentioned above, the first step in the analysis was to determine the variance at student,
classroom and school level without explanatory variables (null or empty model). The
variance in the empty model is 375,83 (Table 3). Of the total variance, 332.95 (Standard
Error =14.57), 28.21 (S.E.=12.98) and 14.67 (S.E.=6.85) is accounted for the student, class
and school level respectively. Thus, 88.5% of the variance is at the student level, 7.5% at the
class level and 4.0% at the school level. Moreover, the variance at each level reaches
statistical significance and this implies that MlwiN can be used to identify the explanatory
variables, which are associated with student achievement in Greek Language.

In Model 1, the context variables (three variables: prior knowledge, socioeconomic status and
gender) at student level were added to the empty model (Table 1). The likelihood statistic
(X2) shows a 1309.22 points reduction between the empty model (Model 0) and Model 1
(from 9543.84 to 8234.62 points). In a chi-squared distribution with three degrees of freedom
the reduction was statistically significant (p< .001).

Table 1. Parameters estimates (and standard errors) for the


the first two models for the analysis of Greek Language achievement
Factors Model 0 Model 1
Fixed part (Intercept) 14.88 (1.38) 12.72 (1.43)

Student Level
Prior knowledge in Greek Language 0.83 (0.02)
SES 1.32 (0.36)
8

Sex 1.91 (0.62)

Variance components
School 4.0% 2.0%
Class 7.5% 4.3%
Student 88.5% 26.7%
Absolute 375.83 121.53
Explain 77%

Significant test
X2 9543.84 8234.62
Reduction 1309.22
Degrees of freedom 3
p-value .001

In Model 1/B, slopes of prior knowledge at level two (classroom) and level three (school),
were made random in order to examine whether schools are differentially effective in relation
to the prior knowledge of their students (Table 2). Snijders and Bosker (1999) argue that the
deviance tests are very convenient for testing parameters in the random part. Specifically, the
deviance produced by the residual maximum likelihood (REML) method can be used in
deviance tests when the two models compared have the same fixed parts and differ in their
random part. In this study the reduction of the likelihood statistic between Model 1 and
Model 1/B for Greek language is 24.83 points, which is statistically significant (X2=8208.79,
d.f.=4, p< .005 ). The following observations arise from the first column of Table 4. The
effects of all contextual factors are significant. Prior knowledge in Greek Language is the
strongest effect predicting the Greek Language achievement score. Moreover, the socio-
economic background (SES) has a strong effect on the Greek language achievement score.
Finally, gender affects achievement scores as girls achieve higher scores than boys in Greek
Language. This is in line with results of comparative studies (e.g. the PIRLS study), which
reveal that girls perform better than boys in language.

In Model 2, all the explanatory variables at classroom level were entered. The following
observations arise from the second column of Table 2. The reduction of the likelihood
statistic between Model 1/B and Model 2 for Greek language is 13 points, which is
statistically significant (X2=8195, d.f.=3, p< .01 ). None of the two variables of the teacher
leadership style had a statistically significant effect on the depended variable. As for the
variables concerning learning culture of classroom, three variables had a statistically
significant effect. Academic emphasis and academic efficiency had statistically, significant,
positive effects. On the contrary, disruptive behavior had a small but statistically significant
negative effect.

In Model 3, the explanatory variables at school level were entered. The following
observations arise from the third column of Table 2. First, the reduction of the likelihood
statistic between Model 2 and Model 3 for Greek language is 19.08 points, which is
statistically significant (X2=8176.81, d.f.=4, p< .005 ). Second, one variable of leadership
style, the principals, human resource frame, had a statistically significant positive effect,
whereas none of the variables concerning the principals

Table 2. Parameters (and standard errors) for the analysis of Greek Language achievement
(Students within classes, within schools)
Factors Model 1/ Model 2 Model 3 Model 4
9

Fixed part (Intercept) 12.42(2.27) 12.29(2.77) 12.63(1.97) 12.76(2.02)

Student Level
Prior knowledge in Greek 0.84(0.03) 0.84(0.03) 0.84(0.03) 0.84(0.03)
Language
SES 1.18(0.35) 1.18(0.35) 1.08(0.34) 1.00(0.34)
Gender 1.84(0.61) 1.84(0.61) 1.85(0.61) 1.84(0.61)

Classroom Level
Teachers leadership style
Human frame ** * N.S.S.
Symbolic frame N.S.S.
Classroom learning culture
Academic emphasis 3.02(0.79) 2.95(0.62) 2.40(0.55)
Academic efficacy 2.53(0.99) 2.42(0.82) 1.83(0.82)
Disruptive behavior -1.11(0.53) -0.90(0.45) N.S.S.
Master goal orientation N.S.S.
Performance approach goal N.S.S.
orientation
Classroom performance N.S.S.
Approach goal structure
Success N.S.S.
Academic self-handicapping N.S.S.
strategies
Cheating behavior N.S.S.

School Level
Principals leadership style
Structural frame* () N.S.S.
Human frame () N.S.S.
Political frame () N.S.S.
Symbolic frame () N.S.S.
Structural frame **() N.S.S.
Human frame () 0.85(0.16) 0.83(0.16)
Political frame () N.S.S.
Symbolic frame () N.S.S.
Effective management N.S.S.
Effective leadership N.S.S.
School organizational culture
Teamwork and cooperation N.S.S.
Decision making 0.51(0.17) 0.54(0.16)
Change and innovation 0.78(0.38) 0.80(0.38)
Trust and confidence N.S.S.
Communication N.S.S.
Responsibility and 0.50(0.19) 0.39(0.18)
commitment
General organization N.S.S.
practices
Vision and goals N.S.S.

Interactions
Principals symbolic frame 0.48(0.24)
Academic emphasis
Principals structural frame 0.44(0.19)
X Master goal orientation
Random part
Level 3 random effects
10

(between schools)
Intercept 36.761(16.607) 37.030(16.445) 21.347(6.412) 24.050(8.035)
Intercept/Baseline -0.415(0.147) -0.497(0.164) -0.462(0.197) -0.484(0.208)
Baseline 0.014(0.006 0.014(0.006) 0.015(0.006) 0.016(0.006)

Level 2 random effects


(between classes)
Intercept 102.771(42.069) 102.071(41.776) 81.828(35.229) 86.047(36.397)
Intercept/Baseline -1.050(0.520) -1.060(0.523) -0.853(0.457) -0.886(0.463)
Baseline 0.016(0.007) 0.016(0.007) 0.013(0.006) 0.012(0.006)

Level 1 variance
Intercept 93.329(4.184) 92.283(4.179) 93.244(4.178) 93.413(4.183)

Significance test
2 8208.79 8195.79 8176.71 8171.41
Reduction 13 19.08 5.30
Degrees of freedom 3 4 1
p-value .005 .01 .005 .025

*() = Principals leadership frame (Likert scale statements)


**() = Principals leadership frame (order scale statements)
*** N.S.S. = No statistically significant effect

effectiveness had any significant effect. Second, three variables of school organizational
culture had statistically significant effects. More specifically, change/innovation, decision-
making and responsibility/commitment had positive significant effects.
In Model 4, interactions among explanatory variables were entered. The following
observations arise from the fourth column of Table 2. First, the reduction of the likelihood
statistic between Model 2 and Model 3 for Greek language is 5.30 points, which is
statistically significant (X2=8171.41, d.f.=1, p<.025 ). Second, two interactions between the
principals leadership frames and variables of classroom learning culture had relatively small
but statistically significant positive effects. Both the interaction between the principals
symbolic frame and academic emphasis in a classroom and the interaction between the
principals structural frame and master goal orientation in a classroom affect students
achievement.

Differentiated classroom effectiveness or differentiated school effectiveness refers to the


capacity of the classroom or that of the school to be effective with different groups of pupils.
A random part of the multilevel model was developed in order to examine the differentiated
classroom and school effectiveness. The random part of Model 4 was used to investigate
differentiated school effectiveness with regards to prior knowledge in Greek Language (Table
2). Thus, the negative correlation between slope and intercept either at level 3 (-.781) or at
level 2 (-.858) means that at the end of the school year the students with the lower scores in
prior knowledge indicated more progress than students with higher scores in prior
knowledge. This was evident at both the school and the classroom level.

Effectiveness in Mathematics
The variance in the empty model is 185.84 (Table 3). Of the total variance, 165.43 (Standard
Error =7.31), 10.85 (S.E.=4.85) and 9.57 (S.E.=4,74) is accounted for the individual, class
and school level respectively.

Table 3. Parameters estimates (and standard errors) for the


the first two models for the analysis of Mathematics achievement
11

Factors Model 0 Model 1


Fixed part (Intercept) 18.90 (0.89) 16.45 (0.94)

Student Level
Prior knowledge in Greek Language 0.75 (0.02)
SES 1.17 (0.31)
Sex -1.14 (0.52)

Variance components
School 5.1% 3.5%
Class 6.0% 4.9%
Student 88.9% 37.5%
Absolute 84.98
Explain 54.4%

Significant test
X2 8588.39 7683.52
Reduction 904.87
Degrees of freedom 3
p-value .001

Therefore, 88.9% of the variance is at the student level, 6% at the classroom level and 5.1%
at the school level. Moreover, the variance at each level reaches statistical significance and
this implies that MlwiN can be used to identify the explanatory variables, which are
associated with student achievement in mathematics. In Model 1, the context variables at
student level were added to the empty model. The reduction of the likelihood statistic
between the empty model was 904.87 points (from 8588.39 to 7683.52 points). In a chi-
square distribution with three degrees of freedom the reduction was statistically significant
(p< .001).
In Model 1/B, slopes of prior knowledge at level two (classroom) and level three (school)
were made random in order to examine whether schools are differentially effective in relation
to the prior knowledge of their students (Table 3). The reduction of the likelihood statistic
between Model 1 and Model 1/B for Mathematics is 26.37 points (X2=7657.15, d.f.=4, p<
.001). The following observations arise from the first column of Table 6. The effects of all
contextual factors are significant. Prior knowledge in Mathematics is the strongest effect
predicting the Mathematics achievement score. Moreover, the socio-economic background
(SES) has a strong effect on the Mathematics achievement score. Finally, gender affects the
achievement score, as boys achieve a higher score than girls in Mathematics. This is in line
with findings of evaluative studies, which reveal that boys perform better on Mathematics
than girls (Kyriakides, 2005).

In Model 2, all the explanatory variables at classroom level were entered. The following
observations arise from the second column of Table 4. The reduction of the likelihood
statistic between Model 1/B and Model 2 for Mathematics is 13 points (X2=7650.01, d.f.=2,
p< .01 ). None of the two variables of the teacher s leadership style had a statistically
significant effect on the depended variable. As for the variables concerning the learning
culture of classroom, only one variable had a statistically significant effect. Academic self-
handicapping strategies had a statistically significant, negative effect.
12

Table 4. Parameters (and standard errors) for the analysis of


Mathematics achievement (Students within classes, within schools)
Factors Model 1/B Model 2 Model 3
Fixed part (Intercept) 16.84(1.52) 16.90(1.55) 17.18(1.31)

Student Level
Prior knowledge in Greek Language 0.74(0.02) 0.74(0.02) 0.74((0.04)
SES 1.22(0.30) 1.16(0.30) 1.23(0.29)
Gender -1.14(0.51) -1.17(0.60) -1.10((0.51)

Classroom Level
Teachers leadership style
Human frame *** N.S.S.
Symbolic frame N.S.S.
Classroom learning culture
Academic emphasis 1.25(0.59) N.S.S.
Academic efficacy N.S.S.
Disruptive behavior N.S.S.
Master goal orientation N.S.S.
Performance approach goal orientation N.S.S.
Classroom performance approach goal
structure
Success N.S.S.
Academic self-handicapping strategies -1.34(0.63) -1.36(0.53)
Cheating behavior N.S.S.

School Level
Principals leadership style
Structural frame* () N.S.S.
Human frame () N.S.S.
Political frame () N.S.S.
Symbolic frame () N.S.S.
Structural frame **() N.S.S.
Human frame () 0.51(0.14)
Political frame () N.S.S.
Symbolic frame () N.S.S.
Effective management N.S.S.
Effective leadership N.S.S.
School organizational culture
Teamwork and cooperation N.S.S.
Decision making N.S.S.
Change and innovation 1.09(0.29)
Trust and confidence N.S.S.
Communication N.S.S.
Responsibility and commitment N.S.S.
General organization practices N.S.S.
Vision and goals N.S.S.

Random part

Level 3 random effects


(between schools)
Intercept 19.381(5.890) 19.256(6.517) 6.734(2.499)
Intercept/Baseline -0.502(0.169) -0.533(0.186) -0.359(0.112)
Baseline 0.026(0.009) 0.021(0.009) 0.026(0.009)

Level 2 random effects


(between classes)
13

Intercept 41.675(16.675) 46.656(17.801) 44.313(16.482)


Intercept/Baseline -0.663(0.265) -0.745(0.279) -0.685(0.262)
Baseline 0.021(0.009) 0.021(0.009) 0.019(0.009)

Level 1 variance
Intercept 66.014(2.993) 65.883(2.985) 66.059(2.995)

Significance test
2 7657.15 7650.01 7645.09
Reduction 7.14 4.92
Degrees of freedom 2 1
p-value .001 .01 .05

*() = Principals leadership frame (Likert scale statements)


**() = Principals leadership frame (order scale statements)
*** N.S.S. = No statistically significant effect

In Model 3, the explanatory variables at school level were entered. The following
observations arise from the third column of Table 4. First, the reduction of the likelihood
statistic between Model 1/B and Model 2 for Mathematics is 4.92 points (X2=7645.09, d.f.=1,
p< .05). Second, one variable of leadership style, the principals human resource frame, had
a statistically significant positive effect, whereas none of the variables concerning the
principals effectiveness had any statistically significant effect. Second, one variable of
school organizational culture had statistically significant effects. More specifically,
change/innovation had a positive, statistically significant effect. Finally, no interactions
among explanatory variables had statistically significant effects on student achievement in
Mathematics.
As mentioned above, differentiated classroom or differentiated school effectiveness refers to
the capacity of the classroom or that of the school to be effective with different groups of
pupils. A random part of the multilevel model was developed in order to examine the
differentiated classroom and school effectiveness. The random part of Model 3 was used to
investigate differentiated school effectiveness with regards to prior knowledge in the Greek
Language (Table 4). Thus, the negative correlation between slope and intercept either at level
3 (-.859) or at level 2 (-.747) means that at the end of the school year the students with the
lower scores in prior knowledge indicated more progress than students with higher scores of
pupils in prior knowledge. This was evident at both the school and the classroom level.

Discussion, Implications, and Conclusions


In this section, we intend to discuss the main findings of the study. Firstly, this study found
direct effects of the principals leadership style on students achievement. This is in itself, a
significant finding. In consequence, the results of this study support the direct model (Pitner,
1988) of the principals effects on students achievement. Although the theoretical and the
methodological design of this work permitted the examination of both the indirect model and
the reciprocal model, the evidence supported direct effects. For this reason, it is important to
interpret this result.

One interpretation seems to be related to contextual factors such as the culture and the
educational system of Cyprus. As Hallinger & Heck (1996) supported, the culture and the
educational system of a country affect the principals leadership style. It appears that the
strong interpersonal relations in Cyprus is a main reason which explains the effect of the
14

human resource frame. This is the cultural explanation, which was also examined in Cyprus
by Pashiardis & Ribbins (2003).

Moreover, the size of primary schools in Cyprus, which is rather small and the nature of
primary schools compared to secondary schools, permit the development of both strong
interpersonal relations. As Witzers et al (2003) identified in their paper, effects of the
principals leadership were found in primary schools but no in secondary schools.

Finally, it seems that elements of the human leadership style, which are mostly characterized
as female, affect students achievement. Interpersonal relations constitute a more central
point of reference of the managerial style of women, something which entails a greater
sharing of power (Coleman, 1998). Blackmore & Kenway (1993) note that women show
greater care for their staff and pupils and focus mainly on teaching, learning and the pupils
needs (Fennell, 1997), are more task-founded and visionary, with a more democratic and
participative style, and are more flexible in the avoidance of clashes. It is interesting that the
role of the principals leadership human style in effectiveness is in agreement with
Pashiardis findings. In his study, the characteristics of effective primary school principals in
Cyprus were examined (1998). He found that that all of the effective principals (both male
and female) used their emotions, feelings and sense of egalitarianism in their management
style. Therefore, one would say that MBFE (Management By Feelings and Emotions) was
their philosophy (Pashiardis, 1998; 2004).

Secondly, the lack of any influences of the teachers leadership style on student achievement
offers an area of discussion about this issue. On the one hand, many studies identified a
significant impact of teachers at the classroom level of school effectiveness models both
abroad (Scheerens, 1990; Creemers, 1994; Scheerens & Bosker, 1997; Teddlie & Reynolds,
2000) and in Cyprus (Kyriakides, Campbell , Muljs & Robinson, 2005; Kyriakides &
Demetriou, in press). On the other hand, there is a lack of studies investigating the impact of
the teachers leadership style on student achievement. In addition, the validation of the
relevant questionnaire in this study indicates that it is difficult for students to evaluate their
teacher s leadership style.
Thirdly, the identification of effects of both school and classroom cultures on students
achievement, showed an agreement with studies in other countries (i.e. Little, 1982; Nias,
Southworh & Campbell, 1989). In addition, the findings related to the impact of classroom
culture offer some new perspectives. For example, the finding about the influence of
academic emphasis on students achievement is in agreement with the findings of many
studies in other countries (Kyriakides & Demetriou, in Press). Finally, the findings related to
the effects of academic efficacy and academic classroom handicapping strategies on students
achievement are interesting, as the review of literature indicated a lack of other relevant
studies. In general, the role of culture at both school and classroom level on students
achievement seems to be important.

Fourthly, the findings of the multilevel analysis are in line with the findings of other studies
conducted in Cyprus in order to examine the validity of multilevel models of school
effectiveness (Kyriakides, Cambell, Muljs & Robinson, 2005). For example, there is
agreement that variables at classroom level have more significant effects on student
achievement compared to variables at school level. In addition, the findings of this study are
in agreement with the findings of many other studies, which identified only small statistical
effects of the principals leadership on students achievement (Cheng, 1994; Hallinger &
Heck; 1998, Van de Grift & Houtven, 1999; Witziers et al, 2003). Finally, the findings of this
15

study are in agreement with the results of school effectiveness studies, which deal with
differentiated classroom and school effectiveness (Kyriakides, Campbell, Muljs & Robinson;
2005).

Implications
Those who are interested in school leadership, culture and effectiveness should pay attention
to the above results. Policy makers, school leaders and researchers should examine the role of
both leadership and culture in order to promote school effectiveness. At the national level,
policy makers either in Cyprus or abroad need to promote those changes that could reduce the
elements of the degree of centralization of the educational system and strengthen, self-
management schools, as research showed that this movement increases leadership effects on
students achievement. In addition the system of principals promotion needs to change in
order to offer opportunities and motives to those young teachers who can work as effective
principals. A third recommendation that could increase principals effectiveness is the
implementation of an appropriate program for principals preparation and further professional
development (Pashiardis & Orphanou; 1999). Finally, because of bureaucratic regulations
there is a main obstacle in being an effective school and creating a distinct school culture
(Pashiardis, 2004). The principals in Cyprus have no say on the appointment of personnel to
their schools. In consequence, the policy for decentralization and self-management of schools
should promote the delegation of more responsibilities to principals in order to establish a
more stable environment in schools.

Principals practices also need to promote those behaviours that might increase the human
resource frame, such as creating one to one relationships between the principal and each
teacher or between the principal and each student. For this reason, the principals presence in
every aspect of school life seems to be very important. Moreover, both principals and
teachers should emphasize a strong, positive culture where commitment and responsibility,
innovations and a shared decision-making process are the main characteristics. In addition,
teachers should stress a learning culture where their academic emphasis and academic
efficiency could be developed in classrooms. For this reason, teachers preparation and
further professional development and in-service training need to take place either at the
school level or at the local level.

Finally, more research is needed in the future in order to uncover more information about
some aspects of this study. Firstly, longitudinal studies need to take place. Because of the
frequent rotation of both principals and teachers among schools in Cyprus, changes in the
composition of both the leaders team and personnel lead to an unstable situation. For this
reason, the examination of principals succession patterns and its impact on school
effectiveness needs long time to take place. Moreover, the process of shaping or changing a
schools culture is time consuming. As a result, changes either of the principals leadership
style or in practicing a new personnel team and its distinct culture need time to affect
students achievement. To sum up, future search could offer more validation to the findings
to the present study.

Conclusions
The results of this study raise some interesting conclusions related to the nature of effective
leadership and its relation to culture and effectiveness in schools. Firstly, the direct impact of
the principals leadership style on students achievement indicates the validation of the direct
model of effects. Despite the criticism against this model more attention should be given to it.
16

The direct effects of other school or classroom variables such as culture on students
achievement do not provide obstacles to the direct effects of the principals leadership on
achievement. It seems that the one to one relationship between leaders and followers provides
a basis for this direct effect. This kind of relationship is related to the principals human
resource frame, which is often (wrongly) characterized as a female style (Pashiardis, 1998;
2004).

Secondly, contextual factors such as: (a) the degree of centralization of an educational system
in a country, (b) the frequent rotation of both principals and teachers from one school to
another, and (c) the lack of appropriate and sufficient in-service education, provide serious
obstacles in the efforts to create effective leadership and to shape a strong positive culture in
a school. In consequence, the effects of leadership and culture on school effectiveness could
become weak. Moreover, because of the frequent rotation of both principals and teachers in
the Cyprus educational system, it is difficult to apply longitudinal studies.

References

Andrews, R.L., & Bamberg, J.D. (1989). Teacher and supervisor assessment of principals
leadership and academic achievement. Washington: University of Washington.
Andrews, R., & Soder, R.(1987). Principal instructional leadership and school
achievement. Educational Leadership, 44(6), 9-11.
Barnett, K., McCormick, J., & Conners, R. (1999). A study of the leadership behavior of
school principals and school learning culture in selected New South Wales State
secondary schools. Sydney: Australian Association for Research in Educational Annual
Conference, Melbourne.
Barnett, K., McCormick, J., & Conners, R. (2000). Leadership behaviour of secondary
school principals, teacher outcomes and school culture. Sydney: Australian Association
for Research in Educational Annual Conference.
Bensimon, E.M. (1989).The meaning of Good Presidential Leadership: A Frame
Analysis. Review of Higher Education, 12,107-123.
Bolman, L.G., & Deal, T.E. (1984). Modern Approaches to Understanding and
Managing Organizations. San Francisco: Jossey-Bass.
Bolman, L.G., & Deal, T.E.(1991) Leadership Management Effectiveness: A Multi-
Frame, Multi-Sector Analysis. Human Resource Management, 30, 509-534.
Bolman, L.G., & Deal, T.E.(1992a). Leading and Managing: Effects of Context,
Culture, and Gender. Educational Administration Quarterly, 28, 314-329.
Bolman, L.G., & Deal, T.E.(1992b).Reframing Leadership: The Effects of Leaders
Images of Leadership. Clark, K.E., Clark, M.B., & Campell (eds.) Impacts of
Leadership. Greensboro, N.C.: Center for Creative Leadership.
Bolman, L.G., & Deal, T.E.(1997).Reframing Organizations: Artistry, Choice, And
Leadership (2nd Ed.).San Francisco: Jossey-Bass.
Bryk, A.S., & Raudenbush, S. (1992). Hierarchical Linear Models: Applications and
Data Analysis Methods. Newbury Park, CA: Sage.
Cameron, K.S.(1984). The Effectiveness of Ineffectiveness. Research In
Organizational Behavior, 6, 235-285.
Cheng, Y.C. (1993). Profiles of Organizational Culture and Effective Schools. School
Effectiveness and School Improvement, 4(2), 85-110.
Cheng, Y.C.(1994a). Principals leadership as a critical indicator of school
17

performance: Evidence from multi-levels of primary schools. School


Effectiveness and School Improvement, 5(3), 299-317.
Cheng, Y.C. (1994b). Teacher leadership style: a classroom-level study. Journal of
Educational Administration, 32(3), 54-71.
Cohen, L., Manion, L., & Morisson, K. (2000) (5th Ed). Research methods in
education. London: Routledge/Falmer.
Coleman, M. (1998). The Management Style of Female Headteachers. Educational
Management and Administration, 24(2), 163-164.
Creemers, B.P.M. (1994). The effective classroom. London: Cassell.
Cronbach, L.J. (3rd Ed) (1990). Essentials of Psychological Testing. New York: Harper &
Row.
Deal, T.E., & Peterson, K.D. (1998). How leaders influence the culture of schools.
Educational Leadership, 56, (1), 28-30.
Deal, T. E., & Peterson, K. D. (1999). Shaping School Culture: The Heart of
Leadership. San Francisco: Jossey-Bass Puplishers.
Edmonds, R. (1979). Effective schools for urban poor. Educational Leadership, 37(1), 15-24.
Fennel, H.A. (1999). Power in the principalship: four womens experiences. Journal
of Educational Administration, 37(1), 23-49.
Feitler F. & Gudgel R. (1994). Organizational Culture Questionnaire (School
Version). Unpublished paper.
Fyans, L.J.,& Maehr, M.L. (1990). School Culture, Student Ethnicity and Motivation.
Urbana, Illinois: The National Center for School Leadership.
Fuller, B. (1987). School effects in the Third World. Review of Educational Research,
57, 255-292.
Goldstein, H. (2003). Multilevel statistical model (3rd ed.). London: Edward Arnold.
Golsdtein, H., Rasbash, J., Plewis, I., Draper, D., Brown, W., Yang, M., Woodhouse,
G., & Healy, M. (1998). A users guide to MlwiN. London: Institute of Education.
Hallinger ,P. & Heck, R. (1996). The principals Role in School Effectiveness: An
Assessment of Methodological Progress, 1980-1995, Leithwood, K.,
Chapman, J., Corson, D., Hallinger, P. & Hart, A. International Handbook of
Educational Leadership and Administration, Part 2. Netherlands: Kluwer
Academic Puplishers.
Hallinger ,P. & Heck, R. (1998). Exploring the principals contribution to school
effectiveness: 1980 1995. School Effectiveness and School Improvement, 9(2),
157-191.
Harris, A., & Muijs, D. (2003). Teacher Leadership: Principal and Practice. Paper
presented at American Educational Research Association. USA:Chicago.
Heck, R., Larsen, T., & Marcoulides, G. (1990). Instructional Leadership and school
achievement: validation of a casual model. Educational Administration
Quarterly, 26(2), 94-125.
Hoy, W.K., & Miskel, C.G. (1991). Educational Administration: Theory, Research
and Practice (4th, Ed.). New York: Random House.
Hoy, W.K., & Miskel, C.G. (1996). Educational Administration: Theory, Research,
and Practice (5th Ed.). New York: McGraw-Hill.
Hoy, W.K., & Miskel, C.G. (2001). Educational Administration: Theory, Research,
and Practice (6th Ed.). New York: McGraw-Hill.
Kyriakides, L. (2005). Extending the Comprehensive Model of Educational
Effectiveness by an Empirical Investigation. School Effectiveness and School
Improvement,16(2), 103-152
Kyriakides, L., Cambell, R.J., Muljs, D, & Robinson, W. (2005). Differentiated
18

teacher and school effectiveness: Some implications for educational effectiveness


research. Oxford Review of Education.
Kyriakides, L., & Demetriou, D. (in press). Using international comparative studies for
establishing generic and differentiated models of educational effectiveness research: The
PISA study. Paper presented at ICSEI Conference, Barcelona, Spain.
Little, J. (1982). Norms of collegiality and experimentation. American Educational
Research Journal. 19(3), 325-340.
Levine, D.U., & Lezotte, L.W. (1990). Unusually effective schools. Madison, WI: the
National Center for Effective Schools Research and Development.
Lezotte, L.W. (1989). Selected Resources Complied for the 7th Annual Effective
School Conference. Rimrock: National School Conference Institute.
Maehr, M.L., & Midgley, C. (1996). Transforming School Cultures. Colorado: Westview
Press.
Midgley, C., Maehr, M.L., Hicks, L., Roeser, R., Urdan, t., Anderman, E., & Kaplan,
A. (1996). Manual: Patterns od Adaptive Learning Survey (PALS). Leadership
and Learning Laboratory, University of Michigan, Ann Arbor.
Midgley, C., Maehr, M.L., Hicks, L., Roeser, R., Urdan, .T, Anderman, E., & Kaplan,
A. (2001). Manual: Patterns od Adaptive Learning Survey (PALS). Leadership
and Learning Laboratory, University of Michigan, Ann Arbor.
Mortimore, P., Sammons, P., Ecob, R. & Stoll, L (1988). School Matters: the junior
years. Salisbury: Open Books
Nias, J., Southworth, G., & Cambell, P. (1989). Staff relationships in the primary
school. London: Casell.
Ogawa, R. T., & Bossert, S. T. (1995). Leadership as an organizational quality.
Educational Administration Quarterly, 31(2), 224-243.
Pashiardis, P. (1995). Cyprus Principals and the Universalities of Effective
Leadership. International Studies in Educational Administration, 23(1), 16-27.
Pashiardis, P. (1998). Researching the Characteristics of Effective Primary School
Principals in Cyprus: A Qualitative Approach. Educational Management and
Administration, 26(2), 117-130.
Pashiardis, P. (2004). Democracy and leadership in the educational system of Cyprus.
Journal of Educational Administration, 42(6), 656-668.
Pashiardis, P. & Orphanou, S. (1999). An insight into elementary principalship in
Cyprus. The International Journal of Educational Management, 13(5), 241-251.
Pashiardis, P., & Ribbins, P. (2003). On Cyprus: The Making of Secondary School
Principals. International Studies in Educational Administration, 31(2), 13-34.
Pashiardis, P., Thody, A., Papanaoum, Z., & Johansson (2003). European search for
consensus in diversity. Nicosia, Cyprus: Digiword Limited.
Pitner, N. (1988). The study of administration effects and effectiveness. In Boyan, N. (Ed.).
Handbook of Research in Educational Administration (pp. 99 122). New York:
Longman.
Pounder, D, Ogawa, R. T., & Adams, E. A. (1995). Leadership an Organization-Wide
Phenomena: Its Impact on School Performance. Educational Administration
Quarterly, 34(4), 564-588.
Reynolds, D., & Cuttance, P. (1992). School Effectiveness: research, policy and
practice. London: Cassell.
Rutter, M.Maugham, B., Mortimore, P., Ousten, J. & Smith, A. (1979). Fifteen Thousand
Hours: Secondary Schools and Their Effects on Children. London: Open Books.
Sashkin, M., & Sashkin, M.G. (1990). Leadership and culture building in schools:
quantitative and qualitative understandings. Paper presented at the Annual
19

Meeting of the American Educational research Association. Boston.


Sashkin, M, & Walberg, H. (1993). Educational Leadership and School Culture.
Berkely, California: NcCutchan.
Scheerens, J. (1990). School effectiveness and the development of process indication
of school functioning. School Effectiveness and School Improvement, 1(1), 61-80.
Scheerens, J. & Bosker, R. (1997). Effective Schooling: Research, Theory and
Practice. London: Cassell.
Schein, E.H. (1992). Organizational Culture and Leadership. San Francisco: Jossey-Bass.
Snijders, T. & Bosker, R. (1999). Multilevel Analysis: An Introduction to Basic and
Advanced Multilevel Modeling. London: Sage.
Teddlie, c. & Reynolds, D. (2000). The international handbook of school effectiveness
research. London: Falmer Press.
Van de Grift, W., & Houtven, A.A.M. (1999). Educational leadership and pupil
achievement in primary education. School Effectiveness and School Improvement,
10, 4, 373-389.
Willims, J.D.(1992). Monitoring School Performance: A Guide for Educators.
London: Falmer Press.
Witziers, B, Bosker, R.J., & Krger M.L. (2003) Educational Leadership and Student
Achievement: The Elusive Serach for an Association. Educational Administration
Quarterly, 39, 3, 398-425.

Вам также может понравиться