Mapping a strong school culture and linking it to sustainable school improvement

Abstract

This article illuminates the key elements of a strong school culture that have been linked with sustainable school improvement. Policy literature and conversations highlight the importance of school culture as the softer strategy in school improvement. Within this context, this article reviews existing research literature to theorize the key elements of a “strong school culture.” Based on this, the article attempts to measure the key elements of a strong school culture and explores how those cultural elements are associated with sustainable school improvement, drawing from large survey data in the U.S. Implications for policy and research are discussed in depth.

Full text

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
1
Mapping a Strong School Culture and Linking It to Sustainable School Improvement
1
Abstract
This article illuminates the key elements of a strong school culture that have been linked with
sustainable school improvement. Policy literature and conversations highlight the importance of
school culture as the softer strategy in school improvement. Within this context, this article
reviews existing research literature to theorize the key elements of a “strong school culture.”
Based on this, the article attempts to measure the key elements of a strong school culture and
explores how those cultural elements are associated with sustainable school improvement,
drawing from large survey data in the U.S. Implications for policy and research are discussed in
depth.
Keywords: Strong school culture; sustainable school improvement; professional learning
community; academic press; student support; trust & respect; negativity
1
An earlier version of this paper was presented at XXXX. (Omitted for Blind Review).

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
2
This article aims to illuminate the key elements of a strong school culture that have been
linked with sustainable school improvement. Initial policy work related to rapid school
improvement focused on solutions that involve pressure for accountability along with an
emphasis on hiring and firing to bring staff skills into line quickly (Darling-Hammond & Berry,
1999). The policy conversations in many countries still emphasize “the problem” as one of
qualifications and competence of staff, or of introducing new programs (Duke, 2004). At the
same time, however, there is an increasing emphasis in the policy and practice “turnaround
literature” on the importance of creating supportive school environments for teachers as well as
students (Thapa, Cohen, Guffey & Higgens-D’Alessandro, 2013).
The policy literature rarely uses the phrase “school culture” which is probably due, in
part, to the variability in the way it is conceptualized and used in research. Organizational
culture is a slippery concept with attendant measurement issues, particularly when quantitative
approaches are used (Jung, et al., 2009; Zamutto & Krakower, 1991). Malinowki (2009), in a
recent review of instruments developed to examine school culture, found that few were fully
developed and even fewer validated in multiple country contexts. However, other related terms,
such as “collaborative environment” and “team leadership” are increasingly prominent in policy
discussions, suggesting a willingness to examine the role of school culture as a softer strategy in
sustainable school improvement.
Research that underpins this shift has often emphasized looking, one-at-a-time, for
characteristics of school culture that were associated with student learning. A few, using the
work of Hoy and his colleagues (Hoy, Tartar & Hoy, 2006; Wu, Hoy & Tartar, 2013) have
developed composite models that have proven useful. In linking school culture to school

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
3
improvement, while there is a growing consensus that school culture is the integral part of school
improvement (cf. Gruenert & Whitaker, 2015; Louis & Lee, 2016), an emerging body of
literature on school culture has paid special attention to the facilitating role of school leaders
(mostly principals) in bridging school culture and school improvement (e.g., Copland, 2003;
Hollingworth et al., 2018; Leithwood & Jantzi, 1990). Research on the specific paths by which
leaders help to sustain school improvement over time is less dense, but a few recent studies
(Hollingworth et al., 2018; Palmer & Louis, 2017) report how principals forge and foster a
positive school culture to support school improvement that endures, while Leithwood (2018)
documents the challenges of sustaining improvements in school leaders’ capacities to facilitate
change.
Aside from the critical role of school leadership in lasting school improvement, however,
what elements of a strong school culture are created and shared collectively by teachers and how
a strong school culture plays out for sustaining improvement initiatives are empirically under-
researched. Although the literature of teacher leadership touches on both school culture and
school improvement, in a broader term, its foci have been put more on understanding the
mechanism of teacher collaboration as a form of leadership (e.g., Harris, 2003; York-Barr &
Duke, 2004), the role of teacher leadership in professional learning and development (e.g., Harris,
2003; Lin et al., in-press; Poekert, 2012), the relationship between teacher leadership and teacher
agency (e.g., Charteris & Smardon, 2015; Priestley et al., 2015), and the impact of teacher
leadership on student learning and/or engagement (e.g., Dove & Honigsfeld, 2010; Leithwood &
Jantzi, 2000).

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
4
Reflecting the gaps we noted in leadership studies, it is appropriate to examine a larger
group of competing measures of a stronger school culture to determine which may be the most
critical as levers for change. Within this context, this article centers on two research questions:
• What are the key elements of a strong school culture in conjunction with school
improvement identified in existing studies? This question extends the theorizing work that
we have done in previous studies (e.g., Louis, 2006; Louis & Lee, 2016) by situating it in
a compact view of the distinctive nature of schools as work settings.
• What is the relationship between a strong school culture and sustainable school
improvement as measured by student learning outcomes? This question presents a test of
the significance of culture for one of the critical objectives for schools as professional
organizations: increasing student learning.
Using survey data gathered from 3,983 teachers in the U.S., this study aims to shed light
on the conceptual and empirical links between a strong school culture and sustainable school
improvement by answering the research questions. To this end, the article consists of four parts.
In the next section, we detail our theorizing a strong school culture. We delineate the
characteristics of the elements of a strong school culture closely related to school improvement
that have been identified by previous research. Based on that, we propose possible relationships
between a strong school culture and sustainable school improvement. Second, we describe the
data and analytical approaches used in this study. Third, we present key results drawing from our
analysis: 1) psychometric properties of measures for a strong school culture and 2) relationship
between a strong school culture and sustainable school improvement. Finally, we offer

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
5
discussions and implications of the results for research and policy in terms of sustainable school
improvement.
Theoretical Framework
A persistent challenge facing schools globally is not the absence of school improvement
initiatives, but their sustainability given that more often than not school improvement initiatives
end in episodic, relatively short-lived or inconsistent (cf. Cheng, 2010; Masters, 2010) or fail to
incorporate “reculturing” that entails deep change (Fullan, 2001). How can educators ensure the
sustainability of school improvement? There is a relatively long line of research that connects
school culture to the sustainability of improvement, beginning with Rosenholtz’s (1991)
distinction between schools that were “stuck” and those that were “improving.” The theme of
sustainability has been somewhat limited by the prevalence of cross-sectional study designs, but
Hargreaves and Goodson (2006) emphasized, in their longitudinal study of declining schools,
that culture was a key explanatory variable, while Harris (2006) noted that when schools focus
on improvement with a weak school culture, they are unlikely to prosper. Many studies that look
at culture and sustained improvement are qualitative studies of a few schools (e.g., Bellei, et al.,
2016; Hipp, et al., 2008; Stringer, 2008) but a notable exception is Sleegers, et al.’s (2014)
longitudinal investigation of a large sample of Dutch elementary schools over four years. This
work suggested that teachers’ individual capacities had the greatest impact on changes in their
classrooms but that organizational conditions were particularly important in enhancing teacher’s
motivations and learning.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
6
By experience we also know that heroic teachers alone, as depicted in Hollywood films,
cannot ensure sustainable school improvement (Toole & Louis, 2002), while Leithwood (2018)
points to the need for persistence and collaboration both within the school and external partners
in creating sustained change. It is reasonable to assume that schools with a strong positive culture
where teachers collaboratively work together with collegiality, trust, and shared responsibility,
school improvement would be more likely to continue as the outgrowth of a strong school
culture. Indeed, a few previous studies have documented the linkage between teachers’
professional interactions/relationships and sustainable school improvement (e.g., Coburn, et al.,
2010; Orlina, 2010). This suggests that school improvement driven by teachers’ professional
interactions and networks would last longer than top-down and/or externally imposed
innovations that often create teacher frustration (cf. Lee, 2018; Lin & Lee, 2018; Serbing & Bryk,
2000). Leithwood (2018) emphasizes that school cultures are critical, although long-term
sustainability also requires attention to local community cultures as well as larger policy cultures.
Notably, the concept of professional learning communities (PLCs) within schools has
been influential in explicating the role of teachers’ collaborations and interactions in school
improvement (cf. McLaughlin & Talbert, 2010; Stoll & Louis, 2007; Voelkel & Crispeels, 2017;
Watson, 2014). Research on PLCs views professional interactions as an organizational
mechanism that provides new ideas, critical feedback, and multiple perspectives for teachers
through specific practices such as reflective dialogue (Ancess, 2003; Hord, 1997; Little, 2002;
Louis et al., 2010; Watson, 2014). In this paper we situate PLCs as the “building blocks for
school culture” (McLaughlin and Talbert, 2010, p. 35), and we explore the linkage between a
strong school culture and sustainable school improvement. Specifically, we choose to explore

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
7
two underlying components (organizational learning and professional community) that have
recently been combined into programs and initiatives to introduce PLCs. Notably, they have
emerged from different research trajectories.
The first is the development of a culture of organizational learning, by which we mean
the habits of searching for new information, processing that information with others,
incorporating and evaluating new ideas, and of generating ideas within the school as well as
importing them from outside (Brown & Duguid, 1991; Honig, 2008; Leithwood & Louis, 1998;
Schechter, 2008; Schechter & Qadach, 2011). Organizational learning is important because it has
been shown to be associated with student learning as well as effectiveness in other sectors
(Marks, Louis, & Printy, 2002).
In recent years, research on organizational learning has been accompanied by an
increasingly robust literature that focuses on teachers’ professional cultures, often referred to as
professional community (Louis, Marks & Kruse, 1996; McLaughlin & Talbert, 2010). In this
study we conceptualize three dimensions of professional community. First, we note that while
teachers need to feel responsible for the learning that occurs in their own classrooms, in a strong
professional community they have a collective sense of contributing to the learning opportunities
and outcomes of all students – beyond the current year and their currently assigned students. We
conceptualize this line of practices as “shared responsibility.” Previous studies suggest that this
construct is critical in predicting student learning (Lee, Louis, & Anderson, 2012). A second
dimension is “deprivatization of practice” which involves the opening of classroom doors and
the open sharing of classroom teaching practices through observation and subsequent discussion
(Louis et al., 2010). This practice has been incorporated in the OECD’s Teaching and Learning

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
8
International Survey, which suggests that it is a key component in a school-based PLC but occurs
less often that other PLC dimensions in most countries (Lee & Kim, 2016). The third dimension
is “reflective dialogue” that requires deeper discussions about “what works and what needs to
change in classrooms in order to improve student learning” (Louis & Lee, 2016, p. 539), that go
beyond simply sharing information over time with multiple colleagues (cf. Lin & Lee, 2018).
OECD’s TALIS study also includes this dimension as a key part of school-based PLCs (Lee &
Kim, 2016). We wish to highlight that these three dimensions of professional community,
coupled with organizational learning, have been shown to be associated with various positive
improvement outcomes both in the U.S. and other countries (Bolam et al., 2005; Lee, Walker, &
Bryant, in press; Lomos, Hofman, & Bosker, 2011; Louis & Marks, 1998).
Aside from the current knowledge of the elements of PLCs, the empirical basis for
determining what school cultures may be most effective is limited, with many competing
frameworks in studies that typically examine only one feature of a school’s culture (or climate,
depending on the author).
2
Several characteristics that have been shown to be related to student
achievement in multiple studies are the subject of this inquiry.
First, we examine “academic press,” referring to the degree to which a school clearly
gives priority to academic standards and creates a sense of importance among both staff and
students around academic achievement (Kahne, Sporte, Torre, & Easton, 2008; Lee & Smith,
1999). Examinations of schools over long time periods suggest that this may be an important
2
This paper will sidestep the long-running argument about the distinction between climate and culture (Hoy, 1990)
by asserting that culture typically refers to a quality of values, norms and behaviors in a school that is stable over
time (Schein, 1992), while climate more typically refers to the tone or feeling of a setting that is more mutable
(Gruenert & Whitaker, 2015). We realize that there will be disagreements about this more fundamental
conceptualization but will focus on the implications of aspects of school settings that seem to make a difference and
that can be fostered by leaders.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
9
feature of “improving” schools (Mosenthal, Lipson, Torncello, Russ, & Mekkelsen, 2004).
Second, the school improvement literature has emphasized the need for both pressure and
support over a long period and support for students has been shown to be reflected both in
engagement and achievement. Coherent instructional programs work because they provide
strong supports for students (Hallinger, Walker, & Lee, 2010; Newmann, Smith, Allensworth, &
Bryk, 2001), and are related to achievement (Heck & Hallinger, 2009). Note that research
demonstrating the link between student support and sustainable school improvement is rare.
3
Third, a deep line of research has emerged about the importance of “trust and respect”
among adults in schools in determining both improvement in school practices and student
learning (Bryk & Schneider, 2002; Hoy, Tartar, & Witkoskie, 1992; Tschannen-Moran & Hoy,
1998). Recent studies further illuminate how trust and respect among school staff function as a
moral or emotional resource for school improvement. Demerath’s (2018) case study shows that
trust and respect serve as an important feedback loop, forging and mediating positive emotions
that, in turn, shape school improvement culture. A recent Chilean study (Weinstein, Raczunski
& Peña, 2018) elaborates the processes by which trust is developed and maintained among
school professionals. At the same time, however, longitudinal research with a focus on the link
between trust/respect and sustainability of school improvement is still thin.
Finally, where teachers feel racial or cultural tensions, where they have a negative
attitude toward students, and where absenteeism is high, student achievement tends to be low.
We call this “negativity.” This measure might be thought of as the opposite of academic
3
Note that the construct of “student support” in this study measures how and to what extend students are supported
in school (see Appendix 1). In this regard, the construct is conceptually different from student voice or student
participation that highlights the importance of student agency in (sustainable) school change and improvement (cf.
Bergmark & Kostenius, 2009; Mitra, 2018).

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
10
optimism, which has also been shown to be related to student learning (Hoy, Tarter, & Hoy,
2006). There are three reasons why we include this negative aspect of school culture in our
investigation. First, such negative features of school culture are more often than not identified in
typical school settings. Second research suggests that such negative features of organizational
culture may have disproportionate impacts on organizational outcomes (e.g., Louis & Lee, 2016),
compared to positive aspects of organizational culture (e.g.,Venkataramani, 2013). Third, the
relationship between negativity and sustainability of school improvement is still under-
researched.
Methods
Data Sources
This paper is a secondary analysis of data focusing on defining and measuring a strong
school culture that is linked to sustainable school improvement as measured by changes in
school-level student learning outcomes over three consecutive years. The larger study design
involved a random sample of nine U.S. states, and a subsequent stratified (by district size,
poverty, and diversity) random sample of four districts within those states. Within each of the
sampled districts, a secondary school and an elementary school were selected to collect survey
data from teachers in 2008. In total, teachers and principals in 182 schools were sampled. Within
the sampled buildings, teachers were asked to fill out paper and pencil surveys. Scales to
measure the variables described below were developed from a teacher survey questionnaire to
measure their perceptions of the eight factors noted below. Teacher surveys were administered

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
11
during a faculty meeting.
4
Data about the school’s achievement levels were obtained from public
data sources (i.e., states’ test scores for measuring Adequate Yearly Progress). The final analysis
for this paper included 3,983 teachers from 133 schools; 49 out of the 182 schools where
teachers did not sufficiently respond to the survey items used as key variables of the final
analysis were excluded. As Table 1 shows, the characteristics of the sampled schools show
significant variation in terms of size, levels of family poverty among the student body, and
racial/ethnic diversity.
>>Insert Table 1 here<<
Measures
The study included two broad categories of variables: school-culture related variables and
school-level achievement measure. We define these below (see also Appendix 1 for detailed
information about survey items).
School-culture related variables. School-culture related variables were comprised of 35
indicator variables representing different aspects of a school culture, which resulted in generating
eight latent constructs through confirmatory factor analysis:
• Organizational Learning: This construct was built using 3 items (alpha = .87) such as
“How many teachers in this school seek out and read current findings in education?”
High values (i.e., on a 6-point scale where 6 is “strongly agree”) indicate a high level of
organizational learning.
4
More details about the sampling and data collection procedures may be found elsewhere (Louis, et al., 2010, pp
301-18). The survey development is described in the same source. Surveys are available from the second author.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
12
• Shared Responsibility: This construct was based on individual teacher responses on 4
items (alpha = .85) such as “How many teachers in this school take responsibility for
improving the school outside their own class?” High values (i.e., on a 6-point scale where
6 is “all”) indicate a high level of shared responsibility.
• Reflective Dialogue: This included 4 items (alpha = .83) such as “How often in this
school year have you had conversations with colleagues about what helps students learn
best?” High values (i.e., on a 5-point scale where 5 is “10 times or more”) indicate that
teachers interact with colleagues more frequently through dialogues and conversations
related to student learning.
• Deprivatized Practice: This construct was measured by 4 items (alpha = .80) such as
“How often in this school year have you visited other teachers' classrooms to observe
instruction?” High values (i.e., on a 5-point scale where 5 is “10 times or more”) indicate
that teachers have deprivatized practices more frequently such as visiting to colleagues’
classrooms or inviting colleagues to their classrooms for enhancing their teaching
practices.
• Academic Press: This latent construct was measured by 6 items (alpha = .89) such as
“We have well defined learning expectations for all students.” High values (i.e., on a 6-
point scale where 6 is “strongly agree”) indicate a high level of academic press.
• Student Support: This construct was based on 5 items (alpha = .86) such as “Resources
are allocated to support students who have greater needs.” High values (i.e., on a 6-point
scale where 6 is “strongly agree”) indicate a high level of student support.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
13
• Trust and Respect: This construct was built using 6 items (alpha = .93) such as “Most of
my colleagues can be relied upon to do as they say they will do.” High values (i.e., on a
6-point scale where 6 is “strongly agree”) indicate that a high level of trust and respect is
embedded in teachers’ work life.
• Negativity: This measure was based on 3 items (alpha = .68) such as “Teachers at this
school are absent habitually” and “There are race or cultural tensions at this school.”
High values (on a 6-point scale where 6 is “strongly agree”) indicate that those negative
incidents such as absenteeism and tensions among members are salient.
School-level achievement. To link school culture and student learning, we used school-
level student achievement because individual student achievement data were not available from
all states. This variable for group comparisons of school culture was based on aggregate student
achievement of language arts proficiency, measured at two different time points (i.e., in 2005/6
and 2006/7).
5
We used school-level achievement test data to create three proportional school
groups: Low, Medium, and High in order to investigate how the elements of a strong school
culture are associated with the levels of school achievement (See the notes below Table 1 for
details).
6
More importantly, to further explore the relationship between strong school culture and
“sustainable” school improvement, we purposively subsampled schools that showed either
continuous improvement or continuous decline in student learning outcomes over three
consecutive years from 2005/6 to 2007/8. This supplementary analysis included 34 of the 133
5
The correlation coefficient of language arts proficiency between 2005/6 and 2006/7 was .940 (p<.01).
6
We acknowledge that standardized achievement tests are a limited measure of student learning and school
performance. However, when considering sustainability, there are few data sources that allow an investigation of
other important school outcomes.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
14
schools. For a broader implication, we used both language arts proficiency and mathematics as
school-level achievement.
Analytical Strategy
To investigate the first research question (i.e., the link between a strong school culture
and school-level achievement), it was critical to develop a psychometrically valid measure of a
strong school culture. We thus introduce our approach and measurement results in detail. Data
were analyzed using confirmatory factor analysis (CFA) and a follow-up investigation
employing latent mean analysis, a form of structural equation modeling. Through CFA, we
sought to test our proposed model measuring a strong school culture. Specifically, we examined
the psychometric properties and construct validity of our proposed model. To this end, we
investigated convergent validity, discriminant validity, and the overall model fit, together with
reliability analysis. Based on this measurement model, we further investigated how latent
constructs representing different aspects of a strong school culture are associated with school-
level achievement by employing latent mean analysis in order to examine the first research
question. Specifically, we compared the latent means of each construct by school achievement
level (i.e., low, mid, and high). To this end, tests of configural, metric, scalar, and factor variance
were first conducted.
Several key indices were used to assess model fit. These included chi-square test statistic,
standardized root mean square residual (SRMR), root-mean-square-error of approximation
(RMSEA), comparative fit index (CFI) and the Tucker-Lewis index (TLI). In particular, we

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
15
relied more on standard cutoff recommendations of SRMR, RMSEA, CFI, and TLI (Hu &
Bentler, 1999) rather than chi-square statistic, which is sensitive to sample size (Bentler, 1990).
Before these main analyses, our initial analyses using descriptive statistics indicated that
missing data ranged from 1.2% to 3.8% for each variable representing different aspects of school
culture. Missing data were imputed using the expectation-maximization (EM) algorithm on the
total sample. Our preliminary analyses also revealed that the normality assumption of the data
was not met, as illustrated in Table 2 (see Curran, West, & Finch, 1996). To compensate, we
utilized a bias-corrected approach to bootstrapping, which involved resampling and replacing the
original dataset 1,000 times prior to CFA and latent mean analysis. Based on the bias-corrected
bootstrap procedure (cf. Efron, 1987), we adjusted the parameter estimates, standard errors, and
effect sizes.
To explore the second research question about the link between a strong school culture
and sustainable school improvement, we conducted an additional analysis with the 34 schools
that showed either continuous improvement or continuous decrease in student achievement over
three consecutive years. We employed a MANOVA test instead of latent mean analysis because
the invariance test for these subsamples, a pre-requisite for conducting latent mean analysis, was
not met.
Results
Descriptive Statistics
Table 2 presents descriptive statistics of each survey item such as means, standard
deviations, skewness, and kurtosis. This suggests that, on average, the sampled teachers

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
16
perceived relatively positive organizational cultures in terms of organizational learning, shared
responsibility, reflective dialogue, academic press, student support, and trust/respect. Based on
the descriptive statistics, an investigation on the normality of the dataset using critical ratios
(C.R.) indicated that the data did not meet the normality assumption. As such, we used a bias-
corrected bootstrapping to address this issue.
>>Insert Table 2 here<<
CFA Measurement Model: Defining and Measuring a Strong School Culture
A confirmatory factor analysis measurement model, consisting of the eight constructs,
was conducted. As illustrated in Figure 1, the construct of professional learning community,
which was proposed as a second-order factor consisted of four sub-factors including
organizational learning, shared responsibility, deprivatized practice, and reflective dialogue.
This confirmed current empirical research (Walker et al., 2014) and theoretical considerations
(Louis, 2006). To further validate our four-factor conceptual model, we tested a competing
model, consisting of one second-order factor of professional community (i.e., shared
responsibility, deprivatized practice, reflective dialogue) and one first-order factor of
organizational learning. The alternative was based on research suggesting that organizational
learning emerges as a consequence of professional community (Lin & Lee, 2018). Results

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
17
indicated that there was no significant difference between the proposed model and the alternative
models.
7
As such, the proposed model was used for subsequent analyses.
>>Insert Figure 1 here<<
The CFA model indicated an acceptable overall model fit (see Hu & Bentler, 1999): CFI
= .921, TLI = .914, RMSEA = .055, SRMR = .063, and χ2= 7072.4, df = 546. Based on this, we
further investigated the psychometric properties and construct validity. We first checked the
factor loadings of all the indicator variables. A majority of the indicator variables showed
excellent factor loadings (i.e., higher than .70) or good (i.e., higher than .50) factor loadings
(Tabachnick & Fidell, 2007). Although the factor loadings from the second-order factor of
professional learning community to reflective dialogue (.425) and deprivatized practice (.352)
were relatively lower than others, their coefficients for factor loadings were also statistically
significant (p<.001).
8
Along with inspecting the factor loadings and their statistical significance, we further
investigated the average variance extracted (AVE) of each construct to confirm convergent
validity (Campbell & Fiske, 1959). Higher AVE values suggest that indicator variables are more
representative of each construct.
9
Three constructs (i.e., student support, negativity, and
deprivatized practice) show relatively lower AVE (i.e., lower than .05) while the other constructs
7
The proposed model indicated the following model fit: CFI = .921, TLI = .914, RMSEA = .055, SRMR = .063.
The model fit of the alternative model was: CFI = .922, TLI = .914, RMSEA = .055, SRMR = .063. For more
information about the model-testing work, contact the first author.
8
These results are not tabled but are available from the first author.
9
AVE is computed as follow: AVE = (∑square standardized loadings)/[(∑square standardized loadings) +
(∑measurement error)].

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
18
exceed 0.50, the conventional threshold value (Hair, Black, Babin, & Anderson, 2009). We
conclude that convergent validity with respect to student support (.40) and deprivatized practice
(.41) was partially obtained, but for negativity (.25), a concern remained.
We were left with a measurement question that has theoretical implications, namely
whether the measured dimensions of school culture are genuinely distinctive. Thus, we
investigated discriminant validity. Due to the presence of a few pairs of constructs having high
correlations in our measurement model, we scrutinized whether those constructs having high
correlations can be distinguishable. In relation to discriminant validity, our concern was mainly
on one particular pair of latent constructs—i.e., academic press and student support, showing the
highest correlation (.753) in our measurement model (see Table 3). Although the highest
correlation coefficient was lower than a conventional threshold of .85, which signals poor
discriminant validity or collinearity (Kenny, 2011), we wished to ensure the discriminant validity
of the two constructs. Since there is no explicit way to check collinearity in CFA, we used
several different approaches.
>>Insert Table 3 here<<
We first examined whether the AVE values of academic press and student support are
greater than the square of their correlation (Netemeyer, Johnston, & Burton, 1990). The result
indicated that each of their AVE values (i.e., .528 for academic press and .401 for student
support) were not greater than the square of their correlation (.567). Consequently, we cross-
checked this result with three other investigations.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
19
First, we examined model fit by comparing a competing model which constrained the
correlation of the two constructs to one with the proposed model (Kenny, 2011). The chi-square
test indicated that the two models were significantly different (Δχ2 = 132, df = 1). That is, the
model comparison indicated that discriminant validity exists between the two constructs. Given
these mixed results, we compared the original model another competing model, which collapsed
the two constructs and combine them into one construct (Kenny, 2011). The result indicated that
the original model maintains better model fit CFI = .921, TLI = .914, RMSEA = .055, SRMR
= .063) than the competing model (CFI = .890, TLI = .881, RMSEA = .067, SRMR = .065). This
supported the presence of discriminant validity and psychometric distinctiveness of Academic
Press and Student Support. Finally, we employed an additional complementary assessment using
the correlation coefficient (.753) and standard error (.024) between the two constructs, using the
comparative approach suggested by Anderson and Gerbing (1988). This final investigation
suggested that discriminant validity exists between the two constructs. Considering the sum of
these investigations, we concluded that there is discriminant validity between academic press and
student support to warrant treating them as both theoretically and psychometrically distinct.
In summary, our investigation indicated that the proposed model achieved acceptable
model fit. Also, the constructs in the model showed solid reliability and appeared to obtain
discriminant validity and convergent validity whereas some constructs exhibited partial
convergent validity.
Latent Mean Analysis: Linking a Strong School Culture to School-Level Achievement

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
20
To further link a strong school culture and school-level achievement, measured by 2005/6
and 2006/7, we conducted latent mean analysis. Specifically, we compared teachers from low-
performing schools (1,204 teachers) with their counterparts in mid- (1,242 teachers) and high-
performing (1,274 teachers) schools.
10
A series of invariance tests of the CFA model (see Figure
1) with bootstrapping was conducted.
11
Based on the invariance tests, the latent mean model with
bootstrapping indicated an acceptable model fit: χ2(1758) = 8627.9, CFI = .910, TLI = .908,
RMSEA = .032 and SRMR = .069.
The latent mean model showed that teachers in low-performing schools (i.e., the
reference group) perceived weaker school cultures than their counterparts in mid- and high-
performing schools, as presented in Tables 4 and 5. Specifically, compared to teachers in low-
performing schools, their counterparts in mid-performing schools turned out to perceive
significantly stronger school cultures in terms of professional learning community (.197***),
academic press (.318***), student support (.345***), and trust/respect (.286***). Conversely,
teachers in mid-performing schools perceived significantly lower levels of school cultures in
terms of negativity (-.468***). The effect sizes (Cohen’s d) further reinforced this conclusion of
significantly stronger cultures among teachers in mid-performing schools than teachers in low-
performing schools: professional community (.305), academic press (.376), student achievement
(.375), trust/respect (.305), and negativity (-.332).
The same pattern was identified in the comparison between low- and high-performing
schools. Compared to teachers in low-performing schools (reference group), their counterparts in
10
Due to missing values of academic achievement of a few schools, the number of teachers in the latent mean
analysis was slightly reduced (i.e., from 3,983 to 3,720).
11
We compared the values of ΔTLI, ΔRMSEA, ΔSRMR, and ΔCFI for invariance tests rather than using Δχ2 given
its statistical stringency regardless of the nature of data such as sample sizes.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
21
high-performing schools perceived significantly stronger school cultures. At the same time, the
effect sizes further informed us that the differences in teachers’ perceptions of school cultures
between low- and high-performing schools were wider than the discrepancies between low- and
mid-performing schools (see Table 5).
All these results indicate that there were significantly positive associations between the
latent means of school culture constructs and the levels of school performance.
>>Insert Table 4 here<<
>>Insert Table 5 here<<
MANOVA: Linking a Strong School Culture to Sustainable School Improvement
We categorized the 34 continuously declining/continuously improving schools into two
additional groups using an additional categorization (high-performing vs. low-performing
schools) –i.e., four groups in total (see Table 6) based on school-level achievement for the three
consecutive years from 2005/6 to 2007/8. As an example, School 3 is a high performing school
with a continuous increase in school-level achievement (High Performing School with
Sustainable Improvement), while School 8 is a high performing school but shows a continuous
decline in achievement (High Performing School without Improvement).
>>Insert Table 6 here<<

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
22
Next, we conducted a MANOVA test with the 34 schools instead of latent mean analysis,
as the invariance test was not met.
12
Before reporting the MANOVA results, we briefly
summarize the descriptive statistics here (see also Appendix 2 for details). On average, high-
performing schools with sustainable improvement showed the highest levels in most dimensions
of strong school culture (e.g., PLC, academic press, trust & press) than the other types of schools.
Also, high-performing schools with sustainable improvement showed the lowest level in the
negative school culture (i.e., Negativity) than the other schools. Conversely, on average, lowest-
performing schools without improvement showed the lowest levels in all the cultural components.
At the same time, however, the cases of low performing schools with sustainable improvement
showed similarly positive features of those school cultural dimensions to those of high-
performing schools. The MANOVA results in Table 7 confirmed that there were significant
group differences in the cultural elements by the type of schools.
>>Insert Table 7 here<<
Post-hoc tests further confirmed the patterns of the descriptive statistics described
above.
13
At the same time, the post-hoc test also suggests that there was no statistical difference
between low-performing schools with sustainable improvement and the other two groups of
high-performing schools in the domain of PLC, Academic Press and Student Support.
Statistically, low-performing schools with sustainable improvement lagged only behind the high-
12
We tested the same CFA measurement model with the subsamples. The model fit was good (e.g., CFI = 910;
RMSEA =.057) with Chi-square = 2382.98, df = 546. This means that the latent constructs of strong school culture
were also clearly identified with teachers in the subsampled schools.
13
These results are not tabled but are available from the first author.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
23
performing schools in Trust & Respect and Negativity. A visual depiction of the results, shown
in Appendix 2, is, perhaps, the most compelling reminder that strong cultures can propel
improvement even in schools that start with low achievement, thus linking school culture with
early sustainability.
Limitations
Before our discussion on the results, we acknowledge that there are several
methodological and theoretical limitations in this study. First, although our proposed model
showed solid reliability, all significant factor loadings, and discriminant validity, the partial
convergent validity of some constructs (Student Support, Negativity and Deprivatized Practice)
suggests that further investigations are needed. Second, further qualitative and quantitative
investigations of the school culture constructs are necessary to understand the dynamics and
interactions of culture, improvement, and sustainability. Third, a longitudinal analysis that
includes school culture variables as well as student outcomes would be desirable to capture
which aspects of school culture most influence school performance, as would studies that include
individual student outcomes.
Discussion
What are the key elements of a strong school culture in conjunction with school improvement,
which are identified in existing studies? In this study, we identified academic press, student
support, trust & respect, low negativity (or optimism as the opposite of negativity), professional
learning community (consisting of shared responsibility, reflective dialogue, deprivatized
practice, and organizational learning). More importantly, results suggest that schools strongly

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
24
equipped with those cultural elements showed higher levels of school performance than that of
their counterparts; there were significantly positive associations between school culture
constructs and the levels of school performance. While the results are understandable intuitively,
given the positive organizational features embedded in those cultural elements, the results can be
further understood by looking closely at emerging research literature.
First, our work affirms the earlier efforts of Hoy and his colleagues to develop measures of
school culture that are not uni-dimensional. We have, we believe, added significantly to this
effort by incorporating additional measures of school culture that have frequently been
associated with student achievement, particularly those that are associated with the robust
scholarship on Professional Community/Professional Learning Communities. Specifically, in the
literature of PLC, educators are characterized as knowledge workers who pursue collective
learning (Organizational Learning), which is promoted when they clearly give priority to student
learning and academic standards (Academic Press). To this end, teachers in the context of PLC
are also expected to share their responsibility for student learning – i.e., Shared Responsibility –
to which their opening classroom activities (Deprivatized Practice) and day-to-day dialogues
reflecting their instruction (Reflective Dialogue) are instrumental. It is reasonable to expect that
this kind of collaborative culture may reduce difficult and/or uncertain work as emotional labor
(Low Negativity), which is sometimes inevitable from teachers’ professional lives, and can
gradually shape mutual respect and trust among colleagues (Trust & Respect). All this
professional context can form a norm that students as vulnerable long-term members should be
supported by adult members in school (Student Support). Theorizing school culture is a daunting
task. However, we believe that the results from our analysis combined with the emerging

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
25
literature can be a platform to add momentum to theorizing school culture as a multi-dimensional
construct.
Our second research question is: What is the relationship between a strong school culture
and sustainable school improvement as measured by student learning outcomes? There were
significantly positive associations between the cultural elements of school and the levels of
school performance. More importantly, our analysis further suggests that there was a clear
linkage between schools with a strong culture and their continuous improvement in school-level
achievement. That is, the cultural elements are critical to sustainable school improvement,
measured by academic achievement. Even low-performing schools appeared to be able to
sustain the improvement of academic achievement, especially when they were strongly equipped
with those cultural elements. This suggests that the effect of school culture on school
performance is not short-lived. It can be an enduring effect that counters organizational inertia.
This finding reflects Rosenholz’s (1969) and Timperly and Robinson’s (2001) investigations,
which suggest that teachers’ cultural assumptions either propel a school toward continuous
improvement or hold it back.
Although our research questions did not focus on measurement, we cannot ignore the
significance of our investigation for the measurement of school culture. While the psychometric
properties of our constructs and the model are relatively strong, a few important issues remain.
First, we wish to note that the average score of teachers’ perceptions of deprivatized practices in
their schools was relatively low which likely accounts for the insignificant group differences in
deprivatized practices between our improving and stuck schools. Further efforts to improve our
understanding and measurement of deprivatized practice seem advisable. The weak convergent

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
26
validity of the construct of negativity suggests a need to pay more attention to how best to
conceptualize and measure underlying or more passive indicators of a non-nurturing culture that
result in absenteeism and racial/ethnic or other tensions.
We wish to reassert that studies conducted in different countries and regions, which have
also employed the same constructs of a strong school culture, have been validated. Specifically,
the cultural elements such as PLC measured by the same or similar survey question items have
been validated in several Asian countries (Ho, Lee, & Tang, 2016; Lee & Louis, 2018; Lee,
Walker & Bryant, in -press; Lin, Lee, & Riordan, in-press) as well as the Netherlands (Lomos,
Hofman & Bosker, 2011). In this regard, this study can pave a way to measure a strong culture
that is positively linked to sustainable school improvement not only in the U.S. but elsewhere.
Our findings have significance beyond measurement and the advancement of a consistent
theoretical framework for capturing school culture. In particular, school culture as it is
experienced by adult professionals is largely ignored in today’s policy conversations in spite of
the accumulating evidence from multiple studies and countries that it is associated with student
learning. There have been a lot of policy interventions or programs with a focus on teachers,
school leaders, and/or “comprehensive/core” curriculum to sustain school improvement, and
increasing attention given to improvement science, evidence-based practice, and data use. But
policy has not addressed the underlying conditions that create improvement in some schools and
decline or stasis in others. Arguably, implementing innovative ideas for teachers, leaders, and
curriculum into practice is a far more complex process than we expect, because it is not a simple
one-off event of restructuring schools but a complex, day-to-day process of reculturing schools
(cf. Fullan, 2001). In other words, while such a focus on teachers, leaders, and curriculum is, in-

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
27
and-of-itself, legitimate, a missing-link in the emphasis is a strong school culture. We note, of
course, that the variables that we include do not fully capture all of the elements that may be
critical elements of a school’s culture and point, for example, to collective efficacy (Goddard,
Goddard, Sook Kim & Miller, 2015; Schechter & Qadach, 2011) and collective ethical beliefs
(Starrett, 2005, Voelkel & Crispeels, 2017).
We argue, in addition, that there is a need to shift the policy conversations toward
understanding that a “strong culture” can be critical for sustainable school improvement. We
believe that the proven linkages between the cultural elements of school and the continuous
improvement of school performance found by this study can serve as a good starting point for the
shift of the policy conversations on sustainable school change. Of course, we still do not have a
fuller picture about how and why school culture works. As examples, we need to deepen our
understanding about 1) why the key elements of a strong culture identified in the study vary
across schools, 2) why and how certain cultural elements work differently in different schools,
and 3) whether similar dimensions of a strong school culture exist in geographically various
jurisdictions. In this regard, future studies can benefit from our approaches (e.g., identified
elements of a strong school culture, validated survey instrument with large scale data).
Aside from the possible benefit for future studies on school culture, at the same time, we
wish to note that educators and policy makers should be cautious as to whether it is worth
incorporating quantified metrics of school culture into the current accountability system used for
evaluating school performance and improvement. In particular, there are increasing concerns that
formalized PLCs may undermine teacher professionalism where they are disconnected from
teacher initiated learning (Louis et al., 2010; Philpott & Oates, 2017). Although using

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
28
measurement of school culture as an input into the accountability system might be regarded as a
logical future step given the array of increasing accountability metrics facing schools and
teachers, we speculate that adding school culture into accountability metrics would be likely to
end up trivializing its complexity. In conclusion, we believe that there is a need to shift the
policy conversations toward understanding how a strong culture plays out for sustainable school
improvement but also the need for intellectual discussions and empirical explorations on
problems, paradoxes, and possibilities of school culture before moving into systemic policy
conversations.
References
Ancess, J. (2003). Beating the odds: High schools as communities of commitment. New York:
Teachers College Press.
Anderson, J., & Gerbing, D. (1988). Structural equation modeling in practice: A review and
recommended two-step approach. Psychological Bulletin, 103, 411-423.
Bellei, C., Vanni, X., Valenzuela, J. P., & Contreras, D. (2016). School improvement
trajectories: An empirical typology. School Effectiveness and School Improvement, 27(3),
275-292.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107,
238–246.
Bergmark, U., & Kostenius, C. (2009). ‘Listen to me when I have something to say’: Students’
participation in research for sustainable school improvement. Improving Schools, 12(3),
249–260.
Bolam, R., McMahon, A., Stoll, L., Thomas, S., & Wallace, M. (2005). Creating and sustaining
effective professional learning communities. University of Bristol. Retrieved on March 5,
2016 from http://dera.ioe.ac.uk/5622/1/RR637.pdf
Brown, J.S. & Duguid, P. (1991). Organizational learning and communities-of-practice: Toward
a unified view of working, learning, and innovation. Organization Science. 2(1), 40-57.
Bryk, A.S. & Schneider, B. (2002). Trust in schools: A core resource for improvement.
New York: Russell Sage Foundation
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the
multitrait-multimethod matrix. Psychological Bulletin, 56, 81-105.
Charteris, J., & Smardon, D. (2014). Dialogic peer coaching as teacher leadership for
professional inquiry. International Journal of Mentoring and Coaching in Education,
3(2), 108-124.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
29
Cheng, Y.C. (2010). Paradigm shifts in leadership for learning. Hong Kong: Hong Kong
Institute of Education.
Coburn, C. E., Russell, J., Kaufman, J. H., & Stein, M. K. (2012). Supporting sustainability:
Teachers’ advice networks and ambitious instructional reform. American Journal of
Education, 119, 137–182.
Copland, M. (2003). Leadership of inquiry: Building and sustaining capacity for school
improvement. Educational Evaluation and Policy Analysis, 25(4), 375-395.
Cotton, K. (2001). New small learning communities: Findings from recent literature? Retrieved
January 20, 2011 from http://www.nwrel.org/scpd/sirs/nslc.pdf
Curran, P.J., West, S.G., & Finch, J. (1996). The robustness of test statistics to non-normality
and specification error in confirmatory factor analysis. Psychological Methods, 1, 19-6-
29.
Darling-Hammond, L., & Berry, B. (1999). Recruiting, preparing, and retaining qualified
teachers to educate all of America's children in the 21st century. The Journal of Negro
Education, 68(3), 254-279.
Demerath, P. (2018). The emotional ecology of school improvement culture: Charged meanings
and common moral purpose. Journal of Educational Administration, 56(5), 488-503.
Dove, A., & Honigsfeld, A. (2010). ESL coteaching and collaboration: Opportunities to develop
teacher leadership and enhance student learning. TESOL Journal, 1(1), 3-22.
Duke, D. (2004). The challenges of educational change. New York: Pearson.
Efron, B. (1987). Better bootstrap confidence intervals. Journal of the American Statistical
Association, 82(397), 171-185.
Fullan, M. (2001). The new meaning of educational change. New York: Teachers College Press.
Goddard, R., Goddard, Y., Sook Kim, E., & Miller, R. (2015). A theoretical and empirical
analysis of the roles of instructional leadership, teacher collaboration, and collective
efficacy beliefs in support of student learning. American Journal of Education, 121(4),
501-530.
Gruenert, S., & Whitaker, T. (2015). School culture rewired: How to define, assess, and
transform it. Alexandria, Virginia USA: ASCD.
Hair, J.F., Black, W.C., Babin, B.J., & Anderson, R.E. (1998). Multivariate data analysis.
Upper Saddle River, NJ: Prentice-Hall.
Hargreaves, A., & Goodson, I. (2006). Educational change over time? The sustainability and
nonsustainability of three decades of secondary school change and continuity.
Educational Administration Quarterly, 42(1), 3-41.
Hallinger, P., Walker, A. & Lee, M. (2010). A Study of Successful Practices in the IB Program
Continuum. Hong Kong: Asia Pacific Center for Leadership and Change, Hong Kong
Institute of Education.
Harris, A. (2003). Teacher leadership as distributed leadership: Heresy, fantasy or possibility?.
School Leadership & Management, 23, 313–324.
Harris, A. (2006). Leading change in schools in difficulty. Journal of Educational Change, 7(1-
2), 9-18.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
30
Heck, R. H., & Hallinger, P. (2009). Assessing the contribution of distributed leadership to
school improvement and growth in math achievement. American Educational Research
Journal, 46(3), 659-689.
Hipp, K. K., Huffman, J. B., Pankake, A. M., & Olivier, D. F. (2008). Sustaining professional
learning communities: Case studies. Journal of Educational Change, 9(2), 173-195.
Ho, D., Lee, M. & Teng, Y. (2016). Exploring the relationship between school-Level teacher
qualifications and teachers' perceptions of school-based professional learning community
practices. Teaching and Teacher Education, 54, 32-43.
Hollingworth, L., Olsen, D., Asikin-Garmager, A., & Winn, K.M. (2018). Initiating
conversations and opening doors: How principals establish a positive building culture to
sustain school improvement efforts. Educational Management Administration &
Leadership, 46(6), 1014–1034.
Honig, M.I . (2008). District central offices as learning organizations: How sociocultural and
organizational learning theories elaborate district central office administrators'
participation in teaching and learning improvement efforts. American Journal of
Education, 114(4), 627-664.
Hord, S.M. (1997). Professional learning communities: Communities of continuous inquiry and
improvement. Texas: Southwest Educational Development Laboratory.
Hoy, W.K. (2010) Organizational climate and culture: A conceptual analysis of the school
workplace. Journal of Educational and Psychological Consultation, 1(2),149-168.
Hoy, W. K., Tarter, C. J., & Hoy, A. W. (2006). Academic optimism of schools: A force for
student achievement. American educational research journal, 43(3), 425-446.
Hoy, W.K., Tarter, C.J., & Witkoskie, L. (1992). Faculty trust in colleagues: Linking the
principal with school effectiveness. Journal of Research and Development in Education, 26,
38-45.
Hu, L., & Bentler, P. M. (1999). Cutoff criterion for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55.
Humberman, A.M., & Miles, M.B. (1984). Innovation up close: How school improvement works.
New York: Plenum.
Kahne, J.E., Sporte, S.E., de la Torre, M. Easton, J.Q. (2008). Small high schools on a larger
scale: The impact of school conversions in Chicago. Educational Evaluation and Policy
Analysis, 30(3), 281-315.
Kenny, D.A. (2011). Multiple factor models: Confirmatory factor analysis. Retrieved February
22, 2012, from http://davidakenny.net/cm/mfactor.htm#DV
Lee, M. (2018). What American urban secondary schools could be: An international perspective.
Journal of Educational Administration, 56(5), 463-472.
Lee, M., & Friedrich, T. (2007). The ‘smaller’ the school, the better?: The smaller learning
communities program in U.S. high schools. Improving Schools, 10(3), 261-282.
Lee, M., & Kim, J. (2016). The emerging landscape of school-based professional learning
communities (PLCs) in South Korean schools. Asia Pacific Journal of Education, 36(2),
266-284.
Lee, M. & Louis, K.S. (January 2018). Group dominance, faultline, intragroup conflict, and
professional learning community. A paper presented at the International Congress for
School Effectiveness and Improvement. NIE, Singapore.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
31
Lee, M., Louis, K. S., & Anderson, S. (2012). Local education authorities and student learning:
The effects of policies and practices. School Effectiveness and School Improvement, 23,
133–158.
Lee, M., Walker, A., & Bryant, D. (in-press). What leadership practices are associated with
International Baccalaureate (IB) student achievement? An exploratory study of IB
schools in Southeast Asia. Peabody Journal of Education.
Lee, V.E., Ready, D.D., & Welner, K.G. (2002). Educational equity and school structure:
School size, school overcrowding, and alternative organizational structures. Los Angeles,
CA: UCLA’s Institute for Democracy, Education, & Access (IDEA).
Lee, V. E., & Smith, J. B. (1999). Social support and achievement for young adolescents in
Chicago: The role of school academic press. American Educational Research Journal,
36, 907-945.
Lieberman, A., & Grolnick, M. (1996). Networks and reform in American education.
Teachers College Record, 98, 7-45.
Leithwood, K. (2018). Leadership development on a large scale: Lessons for long-term success.
Thousand Oaks, CA: Corwin.
Leithwood, K., & Jantzi, D. (1990). Transformational leadership: How principals can help
reform school cultures. School Effectiveness and School Improvement, 1(4), 249-280.
Leithwood, K., & Jantzi, D. (2000). Principal and teacher leadership effects: A replication,
School Leadership & Management, 20(4), 415-434.
Lin, W. & Lee, M. (2018). Linking network learning capacity (NLC) to professional community
and organizational learning. Journal of Educational Administration, 56(6), 620-642.
Lin, W., Lee, M., & Riordan, G. (in-press). The role of teacher leadership in professional
learning community (PLC) in International Baccalaureate (IB) schools: A social network
approach. Peabody Journal of Education.
Little, J. W. (2002). Locating learning in teachers’ communities of practice: Opening up
problems of analysis in records of everyday work. Teaching and Teacher Education,
18, 917–946.
Lomos, C., Hofman, R. H., & Bosker, R. J. (2011). Professional communities and student
achievement - a meta-analysis. School Effectiveness and School Improvement, 22(2),
121-148.
Louis, K. S. (2006). Changing the culture of schools: Professional community, organizational
learning and trust. Journal of School Leadership, 16, 477–489
Louis, K.S., & Lee, M. (2016). Teachers’ capacity for organizational learning: The effects of
school culture and context. School Effectiveness and School Improvement, 27(4), 534-556.
Louis, K.S., & Leithwood, K. (1998). Organizational learning in schools: An introduction. In K.
Leithwood, & K. S. Louis (Eds.), Organizational learning in schools (pp. 1-16). The
Netherlands: Swets and Zeitlinger.
Louis, K. S., Leithwood, K., Wahlstrom, K. L., Anderson, S. E., & Michlin, M. (2010). Learning
from leadership: Investigating the links to improved student learning. Retrieved from
https://conservancy.umn.edu/bitstream/handle/11299/140885/Learning-from-
Leadership_Final-Research-Report_July-2010.pdf?sequence=1%26isAllowed=y
Louis, K. S., Marks, H. M., & Kruse, S. D. (1996). Teachers’professional community in
restructuring schools. American Educational Research Journal, 33(4), 757-798.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
32
Maslowski, R. (2006). A review of inventories for diagnosing school culture. Journal of
Educational Administration, 44(1), 6-35.
Marks, H.M., Louis, K.S., & Printy, S.M. (2002). The capacity for organizational learning:
Implications for pedagogy and student achievement. In K. Leithwood (Ed.),
Organizational learning and school improvement (pp. 239–266). Greenwich, CT: JAI
Press.
Masters, G. (2010). Teaching and learning school improvement framework. Melbourne:
Australian Council for Educational Research.
McLaughlin, M.W., & Talbert, J.E. (2010). Professional learning communities: Building blocks
for school culture and student learning. Voices in Urban Education, 27, 35-45
Mitra, D. (2018). Student voice in secondary schools: The possibility for deeper change. Journal
of Educational Administration, 56(5), 473-487.
Netemeyer, R. G., Johnston, M. W., & Burton, S. (1990). Analysis of role conflict and role
ambiguity in a structural equations framework. Journal of Applied Psychology, 75(2),
148-157.
Newmann, F., Smith, B., Allensworth, E., & Bryk, A. (2001). Instructional program coherence:
What it is and why it should guide school improvement policy. Educational Evaluation
and Policy Analysis, 23(4), 297–321.
Olina, E. (2010). Information seeking, trust, and turnover: Three essays examining middle
school instructional advice networks. Unpublished Ph.D. dissertation. IL: Northwestern
University, IL, USA.
Palmer, E. L. & Louis, K.S. (2017) Talking about race: Overcoming fear in the process of
change. Journal of School Leadership, 27(1), pp. 581-611.
Philpott, C., & Oates, C. (2017). Professional learning communities as drivers of educational
change: The case of learning rounds. Journal of Educational Change, 18(2), 209-234.
Poekert, P.E. (2012). Teacher leadership and professional development: examining links between
two concepts central to school improvement, Professional Development in Education,
38(2), 169-188.
Priestley, M., Biesta, G., & Robinson, S. (2015). Teacher agency: An ecological approach.
London: Bloomsbury.
Rosenholtz, S. J. (1991). Teachers' workplace: The social organization of schools. New York,
NY: Teachers College Press.
Schein, E.H. (1992). Organizational culture and leadership. San Francisco, CA.: Jossey-Bass
Publishers.
Schechter, C. (2008). Organizational learning mechanisms: Its meaning, measure, and
implications for school improvement. Educational Administration Quarterly, 44(2), 155-
186.
Schechter, C., & Qadach, M. (2012). Toward an organizational model of change in elementary
schools: The contribution of organizational learning mechanisms. Educational
Administration Quarterly, 48(1), 116–153.
Sleegers, P., EJ Thoonen, E., J. Oort, F., & TD Peetsma, T. (2014). Changing classroom
practices: The role of school-wide capacity for sustainable improvement. Journal of
educational administration, 52(5), 617-652.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
33
Starratt, R. J. (2005). Building an ethical school: A practical response to the moral crisis in
schools. New York: Routledge.
Stoll, L., & Louis, K. S. (2007). Professional learning communities: Divergence, depth and
dilemmas. Milton Keynes, UK: Open University Press.
Stringer, P. M. (2009). Capacity building for school improvement: A case study of a New
Zealand primary school. Educational Research for Policy and Practice, 8(3), 153-179.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics. Boston: Allyn and Bacon.
Tschannen-Moran, M. & Hoy, W. K. (1998). Trust in schools: A conceptual and empirical
analysis. Journal of Educational Administration, 36, 334-352.
Thapa, A., Cohen, J., Guffey, S., & Higgins-D’Alessandro, A. (2013). A review of school
climate research. Review of Educational Research, 83(3), 357-385.
Timperley, H. S., & Robinson, V. M. (2001). Achieving school improvement through
challenging and changing teachers' schema. Journal of Educational Change, 2(4), 281-
300.
Toole, J.C., & Louis, K.S. (2002). The role of professional learning communities in international
education. In K. Leithwood K. & P. Hallinger. (eds) Second international handbook of
educational leadership and administration (pp. 245-279). Dordrecht: Springer.
Venkataramani, V., Labianca, G.L., & Grosser, T. (2014). Positive and negative workplace
relationships, social satisfaction, and organizational attachment. Journal of Applied
Psychology, 98(6), 1028-1039.
Voelkel Jr, R. H., & Chrispeels, J. H. (2017). Understanding the link between professional
learning communities and teacher collective efficacy. School Effectiveness and School
Improvement, 28(4), 505-526.
Wahlstrom, K.L. & Louis, K.S. (2008). How teachers experience principal leadership: The roles
of professional community, trust, efficacy, and shared responsibility. Educational
Administration Quarterly, 44(4), 458-495.
Walker, A. & Lee, M. (in press). Weaving curriculum connections in International Baccalaureate
(IB) schools. Peabody Journal of Education.
Watson, C. (2014). Effective professional learning communities? The possibilities for teachers as
agents of change in schools. British Educational Research Journal, 40(1), 18-29.
Weinstein, J., Raczynski, D., & Peña, J. (2018). Relational trust and positional power between
school principals and teachers in Chile: A study of primary schools. Educational
Management Administration & Leadership. https://doi.org/10.1177/1741143218792912
Wu, J. H., Hoy, W. K., & Tarter, C. J. (2013). Enabling school structure, collective
responsibility, and a culture of academic optimism: Toward a robust model of school
performance in Taiwan. Journal of Educational Administration, 51(2), 176-193.
York-Barr, J., & Duke, K. (2004). What do we know about teacher leadership? Findings from
two decades of scholarship. Review of Educational Research, 74, 255–316.
Zammuto, R. F., & Krakower, J. Y. (1991). Quantitative and qualitative studies of organizational
culture. Research in Organizational Change and Development, 5, 83-114.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
34
Table 1
Sample teachers by school characteristics a
Frequency
%
Building Level
Elementary
Middle
High
School Poverty b
Low
Medium
High
School Diversity c
Low
Medium
High
School Size d
Small
Mid
Large
District Size e
Small
Medium
Large
School Performance f
Low
Mid
High
1429
1222
1242
783
2296
799
1287
1747
826
720
2305
695
756
1744
1219
1204
1242
1274
36.7
31.4
31.9
20.1
59.0
20.5
33.1
44.9
21.2
19.4
62.0
18.7
20.3
46.9
32.8
32.4
33.4
34.2
Notes
a The figures in this table are based on the data before imputation (N = 3893 teachers from 133 schools).
b School poverty is based on % of students participating in free/reduced lunch service (Low = 0%-17%, Medium =
18%-65%, High = 66%+).
c Diversity is based on % of White students (Low = 66%+, Medium = 185-65%, High = 0%-17%).

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
35
d A size of 400 was regarded as an upward limit of defining small schools (Cotton, 2001; Lee & Friedrich, 2007) and
a size of 1500 students and above was grouped as a large schools (Lee, Ready, & Welner, 2002).
e District size was grouped as small (600-2499 students), medium (2500-24999 students), and large (25000 and
above) (Louis et al., 2010).
f Based on the average school performance in language arts proficiency from 2006 and 2007, schools were
categorized into three groups—i.e., low (mean score, 39.8 on a scale of 0 to 100), mid (mean score, 71.9), and
high (mean score, 89.3) performing schools.
Table 2
Descriptive statistics of variables
Construct
Mean
S.D.
Skewness
C.R.
Kurtosis
C.R.
Organizational
Learning
4.50
1.04
-0.52
-13.35
-0.17
-2.15
3.75
1.22
-0.03
-0.79
-0.87
-11.13
3.85
1.16
-0.10
-2.53
-0.75
-9.55
Shared
Responsibility
4.48
1.28
-0.56
-14.22
-0.64
-8.15
4.21
1.19
-0.37
-9.37
-0.64
-8.17
3.99
1.12
-0.22
-5.48
-0.65
-8.26
4.10
1.15
-0.35
-8.86
-0.59
-7.46
Reflective
Dialogue
4.10
1.00
-0.92
-23.45
0.11
1.42
3.61
1.23
-0.46
-11.65
-0.85
-10.80
3.96
1.11
-0.79
-20.19
-0.30
-3.79
4.19
1.05
-1.12
-28.56
0.34
4.27
Deprivatized
Practice
2.27
1.24
0.83
21.04
-0.28
-3.51
2.66
1.23
0.40
10.06
-0.73
-9.31
2.41
1.23
0.66
16.87
-0.45
-5.73
2.34
1.28
0.71
18.19
-0.54
-6.82
Academic Press
5.06
1.11
-1.58
-40.15
2.54
32.33
4.95
1.12
-1.42
-36.16
2.14
27.24
5.09
1.05
-1.53
-38.87
2.65
33.73
5.03
1.15
-1.46
-37.30
2.09
26.64
5.14
1.10
-1.64
-41.82
2.83
36.07
4.87
1.23
-1.28
-32.55
1.31
16.63
Student Support
4.37
1.46
-0.77
-19.68
-0.29
-3.75
4.51
1.34
-0.92
-23.34
0.23
2.92
4.37
1.43
-0.77
-19.60
-0.30
-3.82
4.45
1.37
-0.84
-21.30
-0.08
-0.97
4.59
1.28
-0.94
-23.90
0.43
5.45
Trust and Respect
5.04
1.12
-1.44
-36.77
2.08
26.45

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
36
5.05
1.06
-1.46
-37.17
2.43
30.93
5.20
1.06
-1.73
-44.04
3.35
42.60
5.00
1.06
-1.42
-36.21
2.34
29.77
5.11
1.00
-1.44
-36.78
2.59
33.01
4.69
1.17
-1.11
-28.14
1.03
13.07
Negativity
3.31
1.60
0.07
1.71
-1.17
-14.94
2.40
1.44
0.82
20.97
-0.33
-4.24
2.62
1.51
0.58
14.87
-0.78
-9.99
Note: N = 3983 teachers (from the imputed data). The non-normality of the data was detected through critical ratios
(C.R. in the table). See Appendix 2 for details about each survey item.
Table 3
Correlation Matrix among Latent Constructs
Estimate
Academic Press
<-->
Negativity
-.391***
Academic Press
<-->
Trust & Respect
.552***
Academic Press
<-->
Student Support
.753***
Trust & Respect
<-->
Negativity
-.403***
Student Support
<-->
Trust & Respect
.649***
Student Support
<-->
Negativity
-.459***
Professional Learning Community
<-->
Student Support
.639***
Professional Learning Community
<-->
Negativity
-.346***
Professional Learning Community
<-->
Academic Press
.530***
Professional Learning Community
<-->
Trust & Respect
.690***
Notes: N = 3,983 teachers (from the imputed data). ***p<.001

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
37
Table 4
Latent Mean Comparison of A Strong School Cultures: Low-Performing vs. Mid-Performing
Schools
School Cultures
Latent Meana
SEa
Effect
Sizec
Professional Learning Community
0.197***
0.034
0.305
Academic Press
0.318***
0.041
0.376
Student Support
0.345***
0.043
0.375
Trust and Respect
0.286***
0.04
0.305
Negativity
-0.468***
0.054
-0.332
Notes. N = 3720 teachers (from the imputed data). ***p<.001
aThe bias in mean estimates and standard errors was corrected through bootstrapping.
cEffect sizes were calculated using bias corrected estimates and common variances of latent constructs.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
38
Table 5
Latent Mean Comparison of A Strong School Cultures: Low-Performing vs. High-Performing
Schools
School Cultures
Latent Meana
SEb
Effect
Sizec
Professional Learning Community
0.332***
0.034
0.515
Academic Press
0.528***
0.041
0.624
Student Support
0.560***
0.041
0.609
Trust and Respect
0.442***
0.039
0.471
Negativity
-0.734***
0.055
-0.521
Note. N = 3720 teachers (from the imputed data). ***p<.001
aThe bias in mean estimates and standard errors was corrected through bootstrapping.
cEffect sizes were calculated using bias corrected estimates and common variances of latent constructs.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
39
Table 6
Four Groups of Schools by Change in School-Level Achievement
Continuous Decline
Continuous Improvement
High Performing Schools
Schools 8, 19, 22, 30, 33
Schools 3, 4, 6, 9, 14, 17, 18, 21,
25, 26, 27, 31, 34, 37
Low Performing Schools
Schools 16, 39
Schools 2, 7, 10, 11, 12, 15, 20,
24, 28, 32, 35, 36
Notes: We excluded the following six schools (Schools 1, 5, 13, 23, 38, 40) from our MANOVA analysis because
their overall level of school achievement for the three years was placed in middle-performing groups (second or
third quartile in the achievement scale). For MANOVA, we needed to have a manageable number of grouping
categories.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
40
Table 7
MANOVA Test Results
Multivariate Tests
Effect
Value
F
Hypothesis df
Error df
Sig.
Intercept
Pillai's Trace
.979
8079.09
5.000
866.000
.000
Wilks' Lambda
.021
8079.09
5.000
866.000
.000
Hotelling's Trace
46.646
8079.09
5.000
866.000
.000
Roy's Largest Root
46.646
8079.09
5.000
866.000
.000
School
Types
Pillai's Trace
.143
8.68
15.000
2604.000
.000
Wilks' Lambda
.860
8.94
15.000
2391.046
.000
Hotelling's Trace
.159
9.18
15.000
2594.000
.000
Roy's Largest Root
.135
23.42
5.000
868.000
.000
Note: N = 874

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement. Teaching and Teacher Education, 81, 84-96.
41
Notes: All estimates above are statistically significant at the p<.001 level. Error terms are omitted for simplicity.
AP = Academic Press, SS = Student Support, T & R = Trust & Respect, NE = Negativity, PLC = Professional Learning Community, SR = Shared
Responsibility, RD = Reflective Dialogue, DP = Deprivatized Practice, OL = Organizational Learning.
Figure 1. CFA Measurement Model
T & R
NE
SS
AP
OL
SR
RD
DP
PLC

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
42
Appendix 1. Scaled Items for Analysis
Construct
Survey Question
α
Organizational
Learning
How many teachers in this school show initiative to identify and solve
problems?
0.87
How many teachers in this school share current findings in education
with colleagues?
How many teachers in this school seek out and read current findings in
education?
Shared
Responsibility
How many teachers in this school meet with other teachers to
collaboratively plan?
0.85
How many teachers in this school help maintain discipline in the entire
school, not just their classroom?
How many teachers in this school take responsibility for improving the
school outside their own class?
How many teachers in this school feel responsible to help each other
improve their instruction?
Reflective
Dialogue
How often in this school year have you had conversations with
colleagues about what helps students learn best?
0.83
How often in this school year have you had conversations with
colleagues about development of new curriculum?
How often in this school year have you had conversations with
colleagues about the goals of this school?
How often in this school year have you exchanged suggestions for
curriculum materials with colleagues?
Deprivatized
Practice
How often in this school year have you visited other teachers'
classrooms to observe instruction?
0.80
How often in this school year have you received meaningful feedback
on your performance from colleagues?
How often in this school year have you had colleagues observe your
classroom?
How often in this school year have you invited someone in to help
teacher you class(es)?
Academic
Press
We have well defined learning expectations for all students.
0.89
Our student assessment practices reflect our curriculum standards.
Our school's curriculum is clearly aligned with learning goals.
Our school has multiple ways of assessing student learning other than
state tests.
Academic achievement is recognized and acknowledged by the school.
The school sets high standards for academic performance.
Student
Support
Students have equal opportunities to be assigned to the best teachers.
0.86
Resources are allocated to support students who have greater needs.
All students receive the same quality of instruction.
Struggling students get the attention they need in this school.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement.
Teaching and Teacher Education, 81, 84-96.
43
In our school, problems are viewed as issues to be solved, not as
barriers to action.
Trust and
Respect
Even in a difficult situation, teachers in this school can depend upon
each other.
0.93
Most of my colleagues can be relied upon to do as they say they will do.
I can trust the people I work with to lend me a hand if I need it.
Teachers in this school respect the professional competence of their
colleagues.
Teachers in this school help and support each other.
Most teachers in our school share a similar set of values, beliefs, and
attitudes related to teaching and learning.
Negativity
Students at this school are absent habitually.
0.68
Teachers at this school are absent habitually.
There are race or cultural tensions at this school.
Note: N = 3983 teachers (from the imputed data). The Cronbach alpha of professional learning community (15 items
from organizational learning, shared responsibility, reflective dialogue, and deprivatized practice) was .884.

Lee, M. & Louis, K.S. (2019). Mapping a strong school culture and linking it to sustainable school improvement. Teaching and Teacher Education, 81, 84-96.
44
Appendix 2. Comparing School Culture by Four Types of Schools
Note: High Performing School with Sustainable Improvement (n = 411 teachers), High Performing School without Improvement (n = 79 teachers), Low
Performing School without Improvement (n = 68 teachers), Low Performing School with Sustainable Improvement (n = 316 teachers). The variable of
derivatized practice was converted from 5-points Likert scale to a 6-points Likert scale for this comparison.