Contents

Learning From Leadership: Investigating the Links to Improved Student Learning

Click here to download the full report:
 Learning From Leadership: Investigating the Links to Improved Student Learning

Key Findings

  • As the poverty and diversity of students served by a school increase, teachers‘ perceptions of the contexts in which they work become more negative.
  • As district and school size increases, teachers‘ perceptions of the contexts in which they work become more negative.
  • The leadership teachers experience is perceived to be more favorable in elementary as compared with secondary schools, and in small as compared with large schools.
  • Greater district size is associated with increases in shared leadership.
  • Most features of the context in which teachers work are viewed as more positive in rural as compared to urban schools.

Introduction

Why do school leaders do the things they do? When they are successful, what explains their success? Scholarly and professional discussion of these questions has consistently emphasized context as a crucial factor. As Evans notes, "school leaders negotiate multiple contexts and stakeholders, often with competing and overlapping interests" (2007, p. 159). Leadership success depends greatly on the skill with which leaders adapt their practices to the circumstances in which they find themselves, their understanding of the underlying causes of the problems they encounter, and how they respond to those problems. Context may also constrain leaders, particularly when pressures in the environment are severe.130 In education, pressures arising from rapidly changing communities challenge leaders as they work to create more effective organizations—in the presence, for example, of competition from charter schools or problems created by liberal district transfer policies.

This chapter focuses on three important topics related to context: the socioeconomic and racial mix of students who come to the school, characteristics of the community and the district, and the school‘s size and complexity.

Prior Evidence

Several strands of well-tested leadership theory acknowledge the importance of context. The multiple linkages model asserts a prominent role for "situational variables"—the size of the work group, organizational policies and procedures, the prior training and experience of members—which mediate what the leader is able to do.131 For example, the size of the school will have a significant effect on how well teachers know other teachers; it also will affect the way in which teachers form workgroups or departments to talk about their work.132 The fragmented nature of professional communities, rather than size per se, becomes a constraint on how principals try to organize professional communities to focus on instruction and student learning.

Resource dependence theory argues that organizations are dependent on obtaining resources from their environments, and that they adapt their organizational forms and functioning in order to survive in the settings in which they are located.133 This perspective is consistent, for example, with the assumption that schools in wealthier settings are likely to have better teachers, better leaders, more actively involved parents, and better results.134It also argues, however, that leaders are responsible for building bridges and adapting to the resource constraints that they experience. Schools in poor rural communities, for example, may be more likely to build bridges to the state or to other non-local funding sources, given the local constraints they face.135 Charter schools, which are particularly vulnerable to resource constraints, may need to depend more on non-educational community members than regular public schools do.136

Institutional theories take a different view, arguing that schools (like other major social service sectors) are so constrained by public expectations that they have limited options for becoming very different.137 Public agencies that have limited autonomy, owing to extensive public oversight, find it difficult to develop their own policies and initiatives for change.138 This does not mean that successful leadership activity in schools is impossible, but it does not come easily. Institutional research suggests, furthermore, that the larger set of social expectations about issues, such as how discipline should be handled or how much differentiation in curriculum is appropriate, can be more critical than local conditions.139 In the United States, for example, many parents expect that their children will have access to Advanced Placement or other advanced courses, and these expectations may constrain efforts to adopt a uniform, standards-based curriculum for all students.

Leadership research has been somewhat scattered in its examinations of context. At one extreme, researchers have claimed that local context trumps all other factors. Claims of this sort often are based on single or small-number case studies. In contrast, researchers working from quantitative studies treat contextual variables as factors to be controlled in inquiries about leadership effects. This approach essentially dismisses context as a substantive problem. Much less attention has been given to the relationship between contexts and the practice of education leaders.140 From the perspective of research design, contexts can be conceptualized as antecedents of leadership practices; they also can be conceptualized as mediators and moderators of leadership effects on organizational outcomes.

New Evidence

Equity has been a key focus in our investigations of contexts and leadership. We have sought not only to learn about leadership that might yield equitable outcomes for students (although it was beyond the boundaries of this study to look for leadership effects that were actually "closing the gap"); we also have asked whether leadership itself was equitably distributed among schools. Is the leadership that matters for student learning—shared leadership and instructional leadership—well distributed so that all teachers and students have access to its benefits? In particular, does the leadership that matters vary across contexts:

  • between schools, depending on the types of students who attend? In other words, do poorer and wealthier schools have similar levels of leadership focused on improving schools and classrooms?
  • by the size and location of school districts? We know from other studies that larger, urban districts tend to be less effective, particularly for lower-income students; but we do not know to what extent, or how, leadership effects might explain that pattern of outcomes.
  • between elementary and secondary schools? Might variability in leadership account for some of the differences we have observed in student performance on state benchmarks, where secondary schools did not score as well as elementary schools?

Method

To address these questions, we examined evidence provided by the first and second rounds of principal and teacher surveys, each of which contained measures of leadership behaviors shown elsewhere in this report to be related to student achievement. Our analysis consisted primarily of analysis of variance, in which we compared mean scores of teachers in different settings on various leadership measures.

In these efforts we emphasized our investigation of leadership variables pertaining to the distribution of leadership within a school. We examined teachers‘ perceptions of principals‘ efforts to involve others, and teachers‘ descriptions of their own leadership for improvement (measured by sense of collective responsibility and the development of shared norms and values). In addition we examined the degree to which leadership is exercised to promote a focus on improved curriculum and instruction, both at the school and district level.

Student Differences: Poverty and Diversity

Our results from Round One of the teacher survey indicate that, generally, as student poverty and diversity increase, teachers‘ experience of shared leadership devolving from the principal decreases (See Table 1.6.1 below, and C1.6.1 in Appendix C). We found teachers‘ leadership focused on collective responsibility for student learning to be more likely present in high poverty schools than in low poverty schools, but teachers are less likely in high poverty schools to share norms around teaching and instruction. Also, teachers in higher-diversity schools report that teachers‘ leadership focused on collective responsibility for student learning is lower than that found in lowdiversity schools, and, again, that teachers in low-diversity schools are less likely to share norms around teaching and instruction. Finally, the level of diversity is not statistically related to teachers‘ reports of the principal as an instructional leader (F = 0.23, p = .797; see Table 1.6.2).

Looking at teacher ratings of school climate, school openness to parents, and district support (from Round Two of the teacher survey), we find once again that as poverty and diversity increase, teachers‘ ratings of climate, openness to parents, and district support decrease (see Table 1.6.1 below, and Appendix C1.6.1).141

 

Source: 1 – 8, Teacher Survey Round One; 9 – 11, Teacher Survey Round Two. †For the planned pairwise contrasts among the means, the comparisons shown represent two means significantly different from each other at p < .05, t-test two-tailed. If a contrast is not shown, the two means in question are not significantly different from each other.

From Round Two of the principal survey, we constructed six variables that parallel the teacher survey variables or are of conceptual interest on their own. They were Principal Self-Rating on Shared Leadership Skills, Principal Self-Rating on Improvement Planning Focus, Principal Rating of District School Improvement Focus, Principal Rating of District Shared Leadership Skills, District Policies to Support Organizational Learning, and District Focus on Data-Based Decision Making (see Table C1.6.2 in Appendix C). On none of the six was there a significant main effect for poverty. Looking at the effect of diversity, we find a significant main effect for Principal Self-Rating on Improvement Planning Focus, Principal Rating of District School Improvement Focus, and District Focus on Data-Based Decision Making (see Table C1.6.3 in Appendix C). On these three variables, principals in medium-diversity buildings gave higher ratings than those in low-diversity buildings.142

Location Differences: District Size and Urbanicity

We found a significant main effect for district size on all eight variables from Round One and all three from Round Two of the teacher surveys (see Table 1.6.2). Here, large districts have significant disadvantages on all principal and teacher leadership variables: principal and teacher leadership diminishes as we move from small to large districts—with, however, a single exception. For shared leadership, there is a clear and opposite trend: the larger the district, the greater the degree of shared leadership as reported by teachers. Once again as district size increases, teachers‘ ratings of climate, openness to parents, and district support decreases.

We found a significant main effect on only two of the six variables on the second round of the principal survey: Principal rating of district shared leadership skills and District policies to support organizational learning. On both, principals from small districts gave higher ratings than principals from large districts (see Table C1.6.4, Appendix C).

Results also indicate that schools located in larger metropolitan areas exhibit significant disadvantages regarding the presence of leadership—from principals as instructional leaders and from shared norms among teachers (Table C1.6.5, Appendix C). Teachers‘ shared leadership with parents (F = 1.99, p = .113) and teachers‘ collective responsibility for student learning (F = 1.63, p = .179) were not statistically related to urbanicity. Teachers‘ ratings of climate and district support diminish as we move from rural to urban. Teachers‘ ratings on school openness to parents were not related to urbanicity (F = 1.12, p = .342).

Of the six variables from the second round of the principal survey, only one, District Focus on Data-Based Decision Making, showed a significant main effect (F = 3.45, p = .018); principals in urban districts rated it higher than principals in suburban districts.

 

Source: 1 – 8, Teacher Survey Round One; 9 – 11, Teacher Survey Round Two. †For the planned pairwise contrasts among the means, the comparisons shown represent two means significantly different from each other at p < .05, t-test two-tailed.

School Differences: School Level and School Size

Compared to high schools and middle schools, elementary schools experience higher levels of all forms of leadership associated with student learning (see Table 1.6.3). Teachers in middle and high schools are less likely to trust their principals, less likely to report that they actively involve parents in decisions, and less active as instructional leaders in their buildings. Also, teachers in elementary schools report higher ratings of climate, openness to parents, and district support. At the secondary level, high schools show a higher "leadership deficit" than middle schools, as well as lower ratings on climate, openness to parents, and district support.

School size matters, as well (see Table C1.6.6 in Appendix C). For our analysis we stratified school size (number of students) into quintiles. We found a significant main effect for school size on all eight variables from Round One and all three from Round Two of the teacher surveys. As in large districts, large schools have significant disadvantages on all principal and teacher leadership variables; principal and teacher leadership diminish as we move from small to large buildings. Also, teachers‘ ratings of climate, openness to parents, and district support diminish as we move from small to large buildings.

 

†For the planned pairwise contrasts between the means, the comparisons shown represent two means significantly different from each other at p < .05, t-test two-tailed.

Poverty and District Size

Our results indicate that student poverty and district size amount to a double disadvantage. Larger schools with high-poverty student populations are most likely to experience limited leadership—even when we control for the effects of school level and urbanicity.

Implications for Policy and Practice

Five implications for policy and practice emerged from this section of our study.

  1. Policies and programs should be developed at the state level to address leadership deficits. While the leadership deficits we have uncovered cannot account in any direct way for the achievement gap, they do provide significant evidence that leadership is unequally distributed among U.S. schools. Because leadership deficits are most apparent in schools marked by many other disadvantages known to affect student achievement, it is reasonable to argue that improving leadership capacity in these schools could also help to boost programmatic and curricular initiatives to increase equity. In particular, principals in more disadvantaged school settings are likely to need more professional development and support in their efforts to sustain practices and behaviors that will increase the involvement of others—teachers and parents—in the work of improvement. The sharing of leadership increases the total energy available to support students in learning.
  2. Policy makers and educators should avoid "one size fits all" approaches to leadership development. In their efforts to develop strong programs of instructional and shared leadership, high school principals work at a distinct disadvantage compared with elementary school principals. One-size-fits-all models of professional development for principals (widespread throughout the United States), are unlikely to work well in complex and difficult high school settings (the same point holds for some larger middle schools). This does not mean, of course, that principals in elementary and secondary schools cannot learn from one another; but general leadership models provide only a start.
  3. High-poverty schools, especially large high-poverty schools, need leadership development programs tailored to their specific needs. These are difficult leadership contexts that require additional interventions and support. While many whole-school reform models geared to urban and high-poverty contexts provide excellent professional development for teachers, few provide anything that directly address the needs and experiences for principals in high poverty settings. As we have noted in our analysis of changes in state leadership, support needs to be targeted to schools that are needy, particularly schools and districts that are not meeting AYP targets.
  4. Educators and policy makers should develop models of shared leadership and parent involvement that are context-relevant. One reason why principals in urban and highpoverty settings tend not to share leadership may be that they operate under conditions in which that kind of involvement is not rewarded. Even where urban and high-poverty school districts emphasize public engagement, the policies and preferences tend to "trickle down" to schools only in the form of mandated representation on school councils—a weak strategy for distributing leadership. Without better models and support, principals will continue to focus on the daily pressures of running the school, and not on creating a more democratic climate.
  5. Educators and policy makers should develop clearer programs to support instructional leadership, particularly in secondary schools. Many important studies of instructional leadership have been conducted in elementary school settings. As valuable as much of this work has been, we know that instructional leadership in secondary schools must differ from instructional leadership in elementary schools, simply because high school principals cannot be experts in all subject areas. Many of the strategies that seem to work well in elementary schools do not necessarily work as well in high schools. We cannot expect to see significant improvement until this issue is addressed more clearly.

< < Previous | Next > >

References

130. Ruef (1997).

131. Yukl (2002).

132. Lee, Bryk, & Smith (1993); Louis, Marks, & Kruse (1996).

133. Casciaro & Piskorski (2005); Romanelli (1991).

134. Lee et al. (1993).

135. DeYoung (1995).

136. Holyoke, Henig, Brown, & Lacireno-Paquet (2007); Renzulli (2005).

137. Rowan & Miskel (1999).

138. Boschken (1998).

139. Arum (2000).

140. Hallinger (1996); Hallinger & Murphy (1986).

141. The level of poverty, however, is not statistically related to teachers‘ reports of district support (F = 1.31, p = .272; Table 1.5.1).

142. In our examination of the leadership variable on the six context variables from Round One of the principal survey, we found only a small number of statistically significant main effects, which is not unlike what we see in Table 1.5.2. Compared with the teachers in their buildings, principals are not much attuned to their building, district, or demographic context in their experience of leadership.