The Impact of Transformational Leadership Style of the School Principal
on School Learning Environments and Selected Teacher Outcomes: A Preliminary Report
Alan M. Barnett
Self-concept Enhancement and Learning Facilitation Research Centre
University of Western Sydney, Australia
Paper presented at NZARE AARE, Auckland, New Zealand, November 2003
BAR03777

---

The Impact of Transformational Leadership Style of the School Principal
on School Learning Environments and Selected Teacher Outcomes: A Preliminary Report
Alan M. Barnett
Self-concept Enhancement and Learning Facilitation Research Centre
University of Western Sydney, Australia
The purpose of this paper is to report on an investigation of the relationships between the
transformational and transactional leadership behaviours of school principals in New South
Wales State secondary schools and some selected teacher outcomes and school learning
environment constructs.
A survey was carried out in 52 randomly selected schools involving 458 teachers from across
New South Wales. The Multifactor Leadership Questionnaire Form 5X (Short) developed by
Bass and Avolio (1997) was used to measure leadership behaviour, while, the School
Learning Environment Questionnaire developed by Fraser (1986) was used to assess school
learning environment. Factor analysis was used to determine the validity of the leadership
model developed by Bass and Avolio (1997) and the school learning environment model
developed by Fraser (1986) in the Australian school context. A factor analysis of leadership
items suggested that one transformational factor (vision), one transformational / transactional
hybrid factor (individualised consideration) and one non-leadership factor (laissez-faire)
factor were evident. An analysis of school learning environment items identified seven
factors. Four outcome factors were incorporated; overall satisfaction with leadership,
perceptions of teacher influence, perceptions of teacher effectiveness, and perceptions of
teacher control.
Multilevel modelling analysis was used to explore the relationship between leadership
behaviours, school learning environment factors and teacher outcomes. Contrary to what
might be expected, results from the analysis of the leadership behaviours factors with teacher
outcomes suggested that teacher outcomes like overall satisfaction with leadership is more
closely and highly correlated with individualised consideration rather than with vision.
Further, the leadership behaviour factors demonstrated differential correlations with each of
the school learning environment factors, indicating that principals may target their leadership
behaviour to have maximum impact in any effort at modifying school learning environment.
Introduction
Much effective school research over the past two decades has concentrated on examining the
relationship between the leadership behaviour of school principals and the enhancement of
organisational performance (Shum and Cheng, 1996). Of particular interest have been studies
that have highlighted the mediating role principal's serve between teachers and learners
(Silins and Murray-Harvey, 1999). Interestingly, results from these studies have suggested
that principals have the ability to indirectly effect student achievement by improving the tone
or learning environment of a school (Johnson, Livingston, Schwartz and Slate, 2000).
However, while the concepts of school leadership and school learning environment seem to
be intuitively linked, there have been few studies that have related these concepts together
(Griffith, 1999).

---

Further, recent paradigm shifts in conceptualising leadership have also encouraged
educational researchers to consider these relationships from the perspective of new leadership
models. Prominent among them is the transformational and transactional leadership model
(Burns, 1978, Bass, 1985, Bass and Avolio, 1994; Leithwood and Jantzi, 1990), which
suggests that follower performance can be lifted to beyond what is normally considered to be
acceptable (Bass, 1985). Further, transformational leaders are able to manipulate and alter
their environmental constraints in order to achieve their performance goals (Kirby, King and
Paradise, 1992).
Theoretical Framework
The theoretical framework presented in this paper is based on a mediated-effects model of
effective schools as discussed by Hallinger and Heck (1998, p.162). This mediated-effects
model provides a more complex representation of administrator effects within schools than
does a simple direct effects or a moderated effects approach. It is acknowledged that
antecedent variables can have an important causal influence that effect desired outcomes such
as student achievement. However, the focus of this study is to examine the relationship
between the leadership practices of the principal and school and classroom variables, namely
school learning environment and teacher satisfaction.
In this model, the variable of the principal's role is assumed to be both a dependent and an
independent factor (Hallinger and Heck, 1998). As a dependent variable, the principal is
subject to the influence of external antecedent factors such as socioeconomic status, or
prevailing external environment conditions such as technological change. As an independent
variable, the principal is considered to be the agent of change, influencing directly the actions
of teachers, the learning conditions within the school, and the attainment of outcomes such as
teacher job satisfaction and indirectly, student learning outcomes.
Figure 1. Theoretical framework guiding research on leadership, school learning environment
and selected teacher outcomes (adapted from Hallinger and Heck, 1998, p.732).
Antecedent
Principal
Intervening School
Student
Variables
Leadership
and Classroom
Achievement
Variables
Transformational and Transactional Leadership
Transformational leadership is hypothesised to occur when leaders and followers unite in
pursuit of higher order common goals, when "one or more persons engage with others in such
a way that leaders and followers raise one another to higher levels of motivation and
morality" (Burns, 1978, p. 20). This implies that the leader-follower relationship is one in
which the purposes of both become fused, creating unity and collective purpose (Barker,
1990). The leader motivates followers to "work for transcendental goals instead of immediate
self-interest, for achievement and self-actualisation rather than safety and security" (Murray
and Feitler, 1989, p. 3), and creates within followers a capacity to develop higher levels of
commitment to organisational goals (Leithwood and Jantzi, 2000).

---

Transactional leadership is hypothesised to occur when there is a simple exchange of one
thing for another. Burns (1978, p. 19) argued that transactional leadership occurs "when one
person takes the initiative in making contact with others for the purpose of exchange of
valued things". In this relationship the leader and the led exchange needs and services in
order to accomplish independent objectives (Barker, 1990; Kirby, Paradise and King, 1992).
Bass et al. (1997) conceptualised a third type of leadership, laissez -faire leadership, which
was hypothesised to occur when there is an absence or avoidance of leadership. In this case
decisions are delayed, and reward for involvement absent. No attempt is made to motivate
followers, or to recognise and satisfy their needs (Bass and Avolio, 1997).
Transformational leadership models emphasize that "transformational leaders are able to alter
their environments" to meet their desired outcomes (Kirby, King and Paradise, 1992, p. 303).
Transformational school leaders do this by promoting educational restructuring and
innovation, focusing on building vision, encouraging collaborative participation and raising
the role of followers to that of leader (Silins, 1994).
School Learning Environment
School learning environment refers to that set of factors that can be regarded as influencing
the feel or personality that a school exudes. School learning environment can also be defined
as that set of internal characteristics that distinguishes one school from another and influences
the behaviour of its members, both staff and students (Hoy and Miskel, 1987).
Fraser (1986) argued that school learning environment factors can operate at both classroom
and at school levels. Although both are related, school-level environments are more global. A
classroom climate may be limited to involve relationships between student and student and
between student and teacher within the boundaries of an individual room.
While not all inclusive, this study seeks to examine those school learning environment
constructs identified by Fraser (1986) which operate at the school level rather than the
classroom level, namely student supportiveness, affiliation, professional interest,
centralisation, formalisation, innovation, resource adequacy and achievement orientation.
The Purpose Of The Study
Recent reviews of research in school leadership have indicated that few studies have
examined the effects of transformational and transactional leadership on teacher outcomes
such as perceived satisfaction with leadership and perceptions of teacher effectiveness. Even
fewer studies have addressed the issue of the effects of leadership behaviour on aspects of a
school's learning environment (Griffith, 1999).
Specifically, this paper focuses on examining two leadership behaviours, namely, the
dissemination of vision (transformational behaviour) and individualised consideration (a
hybrid transformational / transactional behaviour) and their role in influencing teacher
perceptions of school learning environment and aspects of teacher job outcomes.

---

Method
Government schools in New South Wales have been grouped into forty administrative
districts, each with its own District Office, District Superintendent and support staff. Schools
were randomly ordered within their administrative districts. The first, second and third
schools within each district were contacted regarding possible involvement in this study. Of
the 117 schools contacted, 52 secondary schools from 31 regions agreed to participate in this
study. After data cleaning, the total sample size consisted of 458 staff and 49 principals.
Teacher Demographics
The staff sample size (n=458) consisted of 200 males (43.7%) and 235 (51.3%) females
(missing = 23 or 5.0%), the majority of whom (60.0%) were full time teachers. A smaller
number (n=132) came from promotions positions held within their schools (28.9%). The
majority of respondents (n=340; 74.2%) had more than 11 years teaching experience, and had
been in their current school for more than 6 years (49.1%). 291 staff (63.5%) reported
teaching within their current principal for more than 2 years.
Instruments
The Multifactor Leadership Questionnaire (5X - Short) (Bass and Avolio, 1997) was used to
examine transformational and transactional leadership constructs. Although this instrument
reportedly measures 5 transformational and 4 transactional constructs, data analysis,
particularly confirmatory factor analysis could only identify one transformational factor,
vision and one hybrid transformational / transactional factor, individualised consideration. A
non-leadership factor, laissez-faire leadership, was also extracted from the data set.
The School Learning Environment Questionnaire (SLEQ) developed by Fraser (1986) was
used to examine eight learning environment constructs. Again, confirmatory factor analysis
could only find support for seven of Fraser's (1986) eight original constructs, and included
Student Supportiveness, Affiliation, Professional Interest, Centralisation, Innovation,
Resource Adequacy and Achievement Orientation.
Four scales were developed to examine selected teacher outcomes. These included
satisfaction with leadership (taken from MLQ-5X (Short)), perceptions of teacher
effectiveness and perception of teacher influence (taken from Patterns of Adaptive Learning
Questionnaire - (Maehr, Midgley, Hicks, Roeser, Urdan, Anderman and Kaplan, 1996)), and
perceptions of teacher control (taken from School Learning Environment Questionnaire).
Data Analysis
Data was collected using the MLQ-5X (Short) and SLEQ instruments along with a number of
demographic and teacher outcome based questions. Initially the data were screened and
examined for outliers and missing data using PRELIS 2.30 (Joreskög and Sorböm, 2001;
Rowe, 2000) before fitting explanatory multilevel models. Cases with more than 20%
missing data were discarded. The EM method (Joreskög and Sorböm, 2001) of estimation of
missing data was employed to impute missing values for the remaining cases, which in total
accounted for less than 5% of the data set.
Factor scale scores for each of the variables used in the study were calculated. Confirmatory
factor analysis was then used to examine each of the constructs developed in the
questionnaires used.

---

Results And Discussion
The scope of this paper does not permit a complete discussion of all the results of this
investigation. Therefore, consideration will be made of the major findings of this study - the
development of the leadership, teacher outcomes and school learning environment constructs,
and the results of fitting the variance components and multilevel models for teacher outcomes
and some selected school learning environment factors.
The Development of Leadership, Teacher Outcomes and School Learning Environment
Constructs
Leadership
Confirmatory factor analysis of the data gathered from the MLQ-5X (Short) instrument
provided support for three leadership constructs. This was a disappointing result, as the
MLQ-5X (Short) claims to be based on ten constructs: five transformational constructs
(idealised influence - attributes; idealised influence - behaviours; inspirational motivation;
intellectual stimulation; individualised consideration), three transactional leadership
constructs (contingent reward; management by exception - active; management by exception
- passive), and one non leadership factor (laissez-faire) (Bass and Avolio, 1997, p. 34).
However, the data did support one transformational construct, "vision"; one hybrid
transformational / transactional construct, "individualised consideration"; and one non
leadership construct, "laissez-faire".
The construct "vision" consisted of items that emphasised the leader’s concern in setting in
place a desirable future state for their school. Key item phrases included values and beliefs,
optimistic about the future, enthusiasms about what needs to be accomplished, having a
strong sense of purpose and articulating a compelling vision for the future. The Cronbach
alpha for this construct was α= .7991.
The construct "individualised consideration" consisted of items that demonstrate the leader's
concern in dealing individually with each follower. Items included treats me as an individual,
considers me as having different needs, abilities and aspirations from others and provides me
with assistance in exchange for my efforts. The Cronbach alpha for this construct was α=
.8228.
Lastly, the construct "laissez-faire" consisted of items that demonstrated the leader's non-
interest in their leadership function. Items included in this factor highlighted the leaders
avoidance of decision making, for example, is absent when needed, avoids making decisions
and avoids getting involved when important issues arise. The Cronbach alpha for this
construct was α= .8280.
Factor correlations between each of the three leadership variables are shown below in Table
1.

---

Table 1 : Correlation Matrix of the Three-Factor Leadership Model
VI
IC
LF
Vision (VI)
1.00
.
Individualised
0.69
1.00
Consideration (IC)
0.03
.
Laissez-Faire (LF)
-0.55
-0.63
1.00
0.04
0.04
.
(significance levels shown in italics)
a. Teacher Outcomes
Four teacher outcome constructs were developed that examined differing aspects of teacher
perceptions of their role and function in their schools. They included teacher perceptions of
overall satisfaction with their leader (Cronbach alpha α= .8960); perceptions of influence
(Cronbach alpha α= .7254); perceptions of teacher effectiveness (Cronbach alpha α= .6405)
and perceptions of focus of control (Cronbach alpha α= .5938).
Factor correlations between the four teacher outcome variables are given below in Table 2.
Table 2 Four factor teacher outcome correlations
SAT
INFLU
TEFF
TCON
Global satisfaction with
1.00
leadership (SAT)
.
Perceptions of influence
.23
1.00
(INFLU)
.06
.
Perceptions of effectiveness
.47
.13
1.00
(TEFF)
.05
.07
Perceptions of control (TCON)
.31
.19
.85
1.00
.06
.08
.05
.
(significance levels shown in italics)
b. School Learning Environment
Confirmatory factor analysis techniques were applied to the data gathered using the SLEQ
(Fraser, 1986), which was designed to measure eight constructs of school learning
environment. Confirmatory factor analysis found support for seven of the original eight
variables, which included student supportiveness, affiliation, professional interest,
centralisation, innovation, resource adequacy and achievement orientation. An analysis of the
data could not support the inclusion of the eighth factor, formalisation, in this study.
The first variable, student supportiveness, referred to teacher's perceptions regarding the
amount of support students received in their respective schools. Items included "most
students are pleasant and friendly to teachers" and "students get along well with teachers".
The Cronbach alpha for this construct was α= .7873.
The construct, affiliation referred to teacher's perceptions of collegiality among the staff they
worked with, and included items such as "I feel accepted by other teachers" and "I feel I cold
rely on my colleagues for assistance if I should need it". The Cronbach alpha for affiliation
was α= .7266.

---

The third construct, professional interest, measured teacher perceptions of the degree to
which they talked about matters relating to their teaching positions in their respective schools.
Items included "teachers frequently discuss teaching methods and strategies with each other",
and "teachers show considerable interest in the professional activities of their colleagues".
The Cronbach alpha for professional interest was α= .6667.
The next construct, centralisation, referred to teacher's perceptions of the decision making
processes within their schools, and their part in it. Items included "teachers are frequently
asked to participate in decision concerning administrative policies and procedures", and "I
have very little say in the running of the school". The Cronbach alpha for this construct was
α= .7100.
The fifth construct, resource adequacy, measured teacher perceptions of the availability of
teaching resources for them to adequately perform their teaching function. Items emphasised
the availability of videos, tape recorders and photocopying facilities. The Cronbach alpha for
this construct was α= .7143.
The construct innovation measured the degree to which teachers perceived they could
implement new ideas in their teaching function. Key item phrases included that teachers were
encouraged to be innovative, and that experimentation was encouraged in the school. The
Cronbach alpha for this construct was α= .6107.
The last construct measured was achievement orientation, or teacher's perception of pressure
to be engaged in their work. Items included that "teachers have to work long hours" and "It is
hard to keep up with your work load". The Cronbach alpha for this construct was α= .6862.
Congeneric models were fitted to each of the constructs used in this study. An analysis of fit
statistics indicated that each was at appropriate levels (Rowe, 2000; Goldstein, 1995). Finally,
a correlation matrix (shown in Table 3) was calculated for each of these factors, with the
majority of correlations being at levels of .450 and under.

---

Table 3: Seven factor school learning environment model correlations
)
t
)
S
)
N
O
e
res
s
(S
I
N
n
(CEN)
n
(AFF)
Int
I)
i
on (
A)
i
on (A
at
t
udent
i
v
e
n
es
(P
(R
S
s
i
onal
vat
t
ralisatio
Achievement
r
i
e
nt
Affiliatio
Inno
e
s
ource Adequacy
O
upport
r
ofes
R
S
P
Cen
Student Supportiveness (SS)
1.00
.
Affiliation (AFF)
0.44
1.00
0.05
.
Professional Interest (PI)
0.20
0.60
1.00
0.06
0.05
.
Centralisation (CEN)
0.27
0.48
0.49
1.00
0.05
0.05
0.05
.
Innovation (INN)
0.22
0.47
0.77
0.81
1.00
0.06
0.05
0.04
0.04
.
Resource Adequacy (RA)
0.42
0.44
0.27
0.41
0.41
1.00
0.05
0.05
0.06
0.05
0.06
.
Achievement Orientation (AO)
0.02
0.01
-0.01
-0.08
-0.03
-0.07
1.00
0.06
0.06
0.06
0.06
0.06
0.06
.
(significance levels shown in italics)
Multilevel Modelling
Introduction
Multilevel modelling provides a useful analytical tool to examine relationship between the
explanatory variables (vision, individualised consideration and laissez-faire leadership) and
the school learning environment response variables (student supportiveness, affiliation,
professional interest, centralisation, innovation, resource adequacy and achievement
orientation) and selected teacher outcomes (global satisfaction with leadership, perceptions of
influence and effectiveness, and teacher control).
While multilevel modelling does not infer causality between the variables examined, it is
particularly suited to this application, as the data has a nested data structure. This allows
relationships to be examined both at a teacher level and a school level.
b. Variance Component Models
Initially, a two level variance component model was fitted to the data in order to determine
the proportion of variance in both the response and explanatory variables. The results are
shown in Table 4. Fixed residual estimate values (school intercepts) are close to zero,
reflecting that they have been standardised and then normalised. Standardisation transforms
all values regardless of their distributions and original unit of measurement to a distribution
with a mean of zero and a standard deviation of 1, which in turn allows comparison across
variables. Normalisation further adjusts standardised scores so that cross comparisons
between variables can be made using a common metric. The ratio of parameter estimates to
their associated standard errors have yielded significant t-values (> 1.96, critical t-value), and

---

indicate that at the school level (σu02) residual variance is statistically significant at the p<
0.05 level for the three leadership response variables, global teacher satisfaction with
leadership and teacher perceptions of effectiveness, and the school learning environment
variables of student supportiveness and innovation. At the teacher level (σ 2
e ), all response
variables are indicated as being statistically significant at the p< 0.05 level.
Table 4 also indicates that 20.8% of the variation in the leadership factor vision is due to
differences between schools, and while 79.2% is due to differences within teachers. Further,
this variation is statistically significant at both levels to the p< 0.05 level. Between school
differences in individualised consideration (19.7%) and laissez faire leadership (25.1%) were
also found, while within teacher differences accounted for 80.3% (individualised
consideration) and 74.9% (laissez faire leadership) respectively. Clearly, most of the variation
between the three leadership factors under consideration is at the teacher level, which is not
an unreasonable finding. This simply reflects that each school has its own principal in a
position of leadership, who brings a unique leadership style to their position.
A further finding is that most of the variation in leadership (vision, 79.2%; individualised
consideration 80.3% and laissez faire 74.9%) can be found at the teacher level. This suggests
that leaders form relationships with individual followers rather than with followers as a total
group (Barnett, McCormick and Conners, 2000).
As with the variation in leadership factors, the majority of the variation in the school learning
environment occurred at the teacher level (σ 2
e ), which were statistically significant to the p<
0.05 level. Variation at this level was found to be high - student supportiveness (73.2%),
affiliation (98.5%), professional interest (95.1%), achievement orientation (93.4%),
centralisation (93.7%), innovation (83.5%) and resource adequacy (94.9%). At the school
level (σu02), only two variables, student supportiveness (26.8%)and innovation (16.5%) were
found to be statistically significant at the p< 0.05 level. This results indicates that while some
variation in school learning environment occurs at the between school levels, they are
confounded by the variation that occurs at the teacher level.
Statistically significant variations within teachers were also found for the teacher outcome
variables examined in this study. For the variable global satisfaction with leadership, 77.1%
of the variation was accounted for at the within teacher level, while perceptions of teacher
influence (96.5%), perception of teacher effectiveness (76.8%) and perceptions of teacher
control (95.1%) all accounted for statistically significant, high variations. Only global
satisfaction with leadership (22.9%) and perceptions of teacher effectiveness (23.2%) were
found to be significant at the between school level.

---

Document Outline

• Introduction
• The Purpose Of The Study
• Results And Discussion
  • • Table 1 : Correlation Matrix of the Three-Factor Leadership Model
      • Table 2 Four factor teacher outcome correlations
• Multilevel Modelling
  • Teacher perceptions of global satisfaction with leadership
• Multilevel
• Fixed:
• Teacher Level
• School Level
• Random:
  • Multilevel
• Fixed:
• Teacher Level
• School Level
• Random:
  • • Summary and Conclusions
    • References

---