Leadership for learning: Does collaborative leadership make a difference in school improvement?

Citation Hallinger, P., & Heck, R. H. (2010). Leadership for learning: Does collaborative leadership make a difference in school improvement?. Educational Management Administration & Leadership, 38(6), 654–678. Sage Publications. Sidewiki
BibDesk PDF




author = {Hallinger, Philip and Heck, Ronald H.},
date-added = {2016-06-18 23:05:31 -0400},
date-modified = {2016-06-18 23:20:42 -0400},
journal = {Educational Management Administration & Leadership},
keywords = {collaborative leadership},
number = {6},
pages = {654–678},
publisher = {Sage Publications},
title = {Leadership for learning: Does collaborative leadership make a difference in school improvement?},
volume = {38},
year = {2010},


Links here


Abstract p. 1

This longitudinal study examines the effects of collaborative leadership on school improvement and student reading achievement in 192 elementary schools in one state in the USA over a 4-year period p. 1

Using latent change analysis, the research found significant direct effects of collaborative leadership on change in the schools’ academic capacity and indirect effects on rates of growth in student reading achievement p. 1

the study identified three different growth trajectories among schools, each characterized by variations in associated school improvement processes. p. 1

The study supports a perspective on leadership for learning that aims at building the academic capacity of schools as a means of improving student learning outcomes. p. 1

Introduction p. 1

This article focuses on understanding whether and how collaborative leadership makes a difference in elementary school improvement and student learning p. 1

Distributed Leadership According to the Evidence (Leithwood et al., 2009). p. 1

we ask three related research questions p. 2

First, does collaborative leadership contribute to school improvement and student achievement? p. 2

Second, is there an identifiable set of patterns in how schools improve (or decline) in their academic performance over time and, if so, how are these patterns related to change in the schools’ leadership and school organizational processes? p. 2

We address these questions in the context of a longitudinal, time-series study of school improvement in 192 elementary schools in one state in the USA. p. 2

The study employed annual surveys of teachers and parents as a means of understanding patterns of change in the strength of collaborative leadership and academic capacity in their schools over a 4-year period p. 2

These perceptions were then compared with growth in reading achievement of a cohort of 12,480 elementary school students as they moved from Grade 3 through to Grade 5 p. 2

This approach allowed us to assess how changes in leadership were associated with patterns of change in the capacity of schools to improve, and subsequent rates of growth in reading achievement. p. 2

First, the study used student reading achievement as the focal measure of school performance p. 2

Second, we assert that a comprehensive model of leadership and school improvement must include features of the school’s external environment and internal organizational processes (Bossert et al., 1982; Hallinger and Heck, 1996a; Ogawa and Bossert, 1995). p. 2

This study employs a conceptual model in which the effects of collaborative leadership on reading achievement are subject to the influence of selected environmental factors and mediated by the school’s academic capacity (Bossert et al., 1982; Hallinger and Heck, 1996a, Pitner, 1988). p. 2

Finally, our interest in school performance lies explicitly in monitoring how leadership contributes to changes in school performance improvement over time. p. 2

We suggest that a longitudinal time series research design provides incrementally greater leverage over several methodological and theoretical problems that have persisted in this field of inquiry (Hallinger and Heck, 1996b). p. 3

Conceptual Framework p. 3

Conceptualizing ‘Leadership for Learning’ p. 3

organizational theory p. 3

defines leadership as an influence process that shapes the behavior of individuals and groups towards the attainment of goals (Yukl, 2006). p. 3

We highlight two dimensions of the conceptualization of leadership used in this study. p. 3

First, this study focused explicitly school leadership that is team-oriented or collaborative (Marks and Printy, 2003; Mulford and Silins, 2003). p. 3

school-wide leadership exercised by those in management roles (for example, the principal, assistant principals, department heads) as well as others (for example, teachers, parents, staff, students). p. 3

Shared leadership, or what we will refer to in this article as collaborative leadership, thus encompasses both formal and informal sources of leadership, and conceptualizes leadership as an organizational property aimed at school improvement (Ogawa and Bossert, 1995; Pounder et al., 1995). p. 3

The current research agenda in this field is, therefore, geared towards deepening our understanding of the means or paths through which p. 3

leadership achieves improvement in teaching and learning (Hallinger and Heck, 1996a; Leithwood et al., in press; Robinson et al., 2008). p. 4

In recent years, the phrase ‘leadership for learning’ has gained international currency (MacBeath et al., 2008; Robinson et al., 2008) p. 4

In our view, this approach to school leadership represents a blend of two earlier leadership conceptualizations: instructional leadership and transformational leadership (Hallinger, 2003). p. 4

leadership for learning signals, among other things, the critical role that leadership plays in creating and sustaining a school-wide focus on learning (Hallinger and Murphy, 1985; Hallinger et al., 1996; Heck et al., 1990; Marks and Printy, 2003) p. 4

It further highlights the importance of learning, not only for students but also for teachers and staff (Barth, 1990; Fullan, 2001; Leithwood et al., in press; Robinson et al., 2008). p. 4

This capacity-building perspective is especially supported by findings from studies of transformational school leadership (Leithwood and Jantzi, 1999; Marks and Printy, 2003; Mulford and Silins, 2003). p. 4

Three areas of focus were incorporated into the means of leadership: p. 4

Vision—making decisions to facilitate actions that focus the energy of the school on improving student outcomes and fostering commitment. p. 4

Governance—empowering staff and encouraging participation. p. 4

Resource allocation—obtaining and allocating resources to support teaching and learning. p. 4

Proponents of shared leadership also suggest that collaborative leadership has the potential to account for the broader range of naturally occurring leadership processes that exist in schools beyond the formal leadership exercised by principals (Barth, 1990, 2001; Harris, 2003; Lambert, 2002; Ogawa and Bossert, 1995) p. 4

This perspective is reflected in the growing interest expressed in ‘distributed school leadership’ (Gronn, 2002; Leithwood et al., 2009). p. 4

Scholars have suggested that acknowledging and developing the broader leadership capacity in schools may hold the key to unlocking the store of leadership potential grounded in instructional expertise that principals are often unable to provide (Barth, 1990, 2001; Crowther et al., 2008; Donaldson, 2001; Fullan, 2001; Gronn, 2002; Grubb and Flessa, 2009; Hall and Hord, 2001; Leithwood et al., 2009; Marks and Printy, 2003). p. 4

With this in mind, this study examines collaborative leadership drawn from a variety of sources, including but not limited to the principal. p. 4

Modeling Collaborative Leadership in the School Improvement Process p. 4

Students typically attend a school over a period of several years, during which they receive instruction from multiple teachers. p. 4

Student learning, therefore, depends at least in part on the quality of teaching across classrooms p. 4

From this perspective, school improvement represents a dynamic process in which schools seek to develop the breadth and density of instructional expertise among their teachers. p. 4

Although leadership is often viewed as a catalyst for change, we suggest that the ‘particular location’ of each school in its journey of school improvement creates the need for and shapes the behavior of school leadership p. 5

Leadership, therefore, not only impacts school improvement, but is also shaped by the context in which it is exercised (Bossert et al., 1982; Hallinger and Murphy, 1986; Jackson, 2000; Leithwood et al., 2006; Luyten et al., 2005; Southworth, 2003). p. 5

Increasingly, scholars have begun to view school improvement within a framework of organizational learning (Mulford, 2007). p. 5

Figure 1 presents our proposed model of how changes in school context, collaborative leadership, and school academic capacity are related to changes in student learning p. 5

The model highlights three features of data that must be incorporated into data analysis p. 5

we employ multilevel latent change analysis (LCA), a variant of structural equation modeling (SEM), to examine changes in leadership, school academic capacity and student reading outcomes over a four-year period p. 5

repeated observations of student outcomes can each be represented by two correlated latent (or underlying) factors (Raykov and Marcoulides, 2006) p. 5

These are represented as ovals in Figure 1 p. 5

The level factor represents the level of a particular variable (for example, leadership, academic capacity or student achievement) at a chosen point in time p. 5

The shape factor represents change or growth in the variable over a particular interval. p. 5

The model proposes that changes in levels of collaborative leadership in a given school will impact the school’s academic capacity p. 6

As the school builds its academic capacity over time, we would expect to see subsequent changes in teacher practices, student behavior and learning outcomes p. 6

Two-headed arrows between the level and shape factors in Figure 1 indicate expected negative correlations between the initial level of a particular variable and its subsequent change. p. 6

Research Questions p. 7

1. Does Collaborative Leadership Impact School Performance? p. 7

2. Does Collaborative Leadership Impact the Improvement of School Performance Over Time? p. 7

3. How Do Schools Differ in Their Improvement Over Time and How Are Those Differences Related to Changes in School Leadership and Capacity? p. 7

Research Design and Method p. 8

This study employed a longitudinal panel time-series design (Cook, 2002) covering a period of 4 years p. 8

Our goal in this research was to examine how changes in leadership and academic capacity over time might be related to patterns of growth in student achievement p. 8

Data were collected from students and teachers in 192 elementary schools over a 4-year period. p. 8

We captured changes in school processes through surveys given to each school’s teachers on three occasions (years 1, 3 and 4) p. 8

Achievement data from the student cohort were collected in years 2, 3 and 4. p. 8

Data p. 8

Data were collected from a random sample (N 1⁄4 12,480) of students drawn from a Grade 3 student cohort that was subsequently observed over a three-year period (that is, Grade 3 through Grade 5). p. 8

the same students were incorporated into the data analysis for the duration of the study. p. 8

The students were enrolled in 192 public elementary schools p. 8

Operationalized Model of School Improvement Effects p. 8

Collaborative leadership p. 9

Collaborative leadership was measured by a subscale of five items describing teacher perceptions of leadership exercised within the school (a 1⁄4 0.82) and a corresponding subscale of five items describing parent perceptions of school leadership and their own personal involvement in improving education at the school (a 1⁄4 0.88). p. 9

The state survey items were designed to reflect three specific aspects of collaborative leadership within each school (with items paraphrased in parentheses): p. 9

Make collaborative decisions focusing on educational improvement p. 9

Emphasize school governance that empowers staff and students, encourage commitment, broad participation, and shared accountability for student learning p. 9

Emphasize participation in efforts to evaluate the school’s academic development p. 10

Data Analysis p. 11

Data analysis proceeded in two steps. p. 11

First, we investigated the relationships among variables implied by our proposed model in Figure 1. p. 11

Second, we investigated whether variables in our proposed model might be useful in categorizing schools with different growth patterns in reading over time. p. 11

Results p. 11

Table 1 also suggests that schools made considerable growth in reading over time (averaging about 23.7 scale score points over the first growth interval). p. 11

t-tests suggested that schools changed in average academic capacity (factor means 1⁄4 0.00, 0.02, 0.13, respectively, p < 0.05) and average collaborative leadership (that is, factor means 1⁄4 0.00, 0.04, 0.05, respectively, p < 0.05) over time. p. 11

Does Collaborative Leadership Impact Initial School Performance? p. 13

The results in Figure 2 provide support for the first proposition that initial level of leadership would be related to initial levels of school academic capacity (standardized g 1⁄4 0.19, p < 0.05) p. 13

The second proposition indicated that initial collaborative leadership would be indirectly and significantly associated with initial student learning levels through initial academic capacity. We found evidence of a small indirect (but substantively unimportant) effect of initial collaborative leadership on initial reading outcomes (standardized g 1 0.02, p < 0.05). p. 14

Is the Model Useful in Classifying Schools According to Their Academic Improvement? p. 15

Overall, the variables in the model were very useful in classifying 86 percent of the schools correctly (not tabled) according to their growth in reading (that is, including 70 percent of high-growth schools, 54 percent of low-growth schools, and nearly all of the averagegrowth schools). p. 17

This supports the proposed model’s validity in explaining differences in school reading growth. p. 17

this finding suggests that it could be fruitful to focus in future research on identifying patterns of leadership practice and school improvement strategies that are linked to the different groups of schools based on where they lie in their growth trajectories. p. 17

Conclusion p. 17

Our analysis of longitudinal data supports the view that collaborative leadership positively impacted growth in student learning indirectly through building the academic capacity in schools. p. 17

The results also provide initial insight into patterns of growth that characterize different schools in their school improvement ‘journeys’. p. 17

Limitations of the Research p. 18

Implications p. 19

we suggest that the methods used in this study represent an incremental advance in the state-of-the-art of research on school leadership effects p. 19

we suggest that the ability to employ longitudinal data to model relationships as they change over time is a necessary condition in empirical studies of school improvement (Luyten et al., 2005). p. 19

First p. 20

the finding of indirect leadership effects on academic growth over time reinforces an important conclusion drawn in a series of influential reviews of research on school leadership effects (Bell et al., 2003; Bossert et al., 1982; Bush and Glover, 2003; Hallinger and Heck, 1998; Leithwood and Montgomery, 1982; Leithwood et al., 2004, 2006, in press; Robinson et al., 2008; Witziers et al., 2003). p. 20

Second p. 20

this study provides empirical support for the proposition that collaborative leadership contributes to school improvement through building the school’s academic capacity in quite specific areas (Fullan, 2001; Robinson, et al., 2008). p. 20

By academic capacity, we refer to a set of organizational conditions that impact what teachers do in classrooms to influence student learning. p. 20

we suggest that leadership acts as a driver in identifying needs and devising strategies to foster school-wide academic changes over time (that is, developing and sustaining a schoolwide focus on learning, upgrading the curriculum, providing individualized support for teachers and students, improving the monitoring of student progress). p. 20

The longitudinal analysis found that changes in collaborative leadership over time were directly associated with these types of changes in academic capacity and indirectly related to student growth in reading achievement. p. 20

Third p. 20

there are indeed several different patterns in the ‘growth trajectories’ of schools p. 20

this study further indicates that different strategies may be in order for schools that are at different points in their journeys. p. 20

Finally p. 20

these findings represent an early contribution to the emerging empirical knowledge base on the effects of shared forms of school leadership (Leithwood et al., 2009; Marks and Printy, 2003; Mulford and Silins, 2003; Pounder et al., 1995; Timperly, 2009). p. 20