Mediating effects of school process on the relationship between principals’ data-informed decision-making and student achievement

Abstract

We examined the path of influence from principals’ data-informed decision-making to school process and then to student achievement, by adopting the model of ‘mediated-effects with antecedent effects’. With samples of principals and teachers from Michigan, we used two different approaches to measure principals’ engagement in data-informed decision-making. Results of structural equation modelling revealed that teachers are in a better position than principals to (a) provide measures of principals’ engagement in data-informed decision-making because teacher measures of this leadership of principals accurately predicted almost all aspects of school process and (b) predict the indirect effects of principal leadership on student achievement via school process (or the direct effects of school process on student achievement). Policy implications were discussed such as the best source for measuring principals’ data-informed decision-making, and the path of influence from principals’ leadership to school process and to student achievement.

Acknowledgements

The authors wish to acknowledge the permission from the publisher, Taylor & Francis (http://www.tandfonline.com), to reuse some of the materials in Shen et al. (2012).

Notes

1. With 107 teachers coming from 98 schools (principals), the vast majority of schools had only one teacher. We, thus, did not address the issue of (within-school) inter-rater agreement. Because we adopted existing instruments that have been psychometrically validated to measure principal leadership and school process, reliabilities and confirmatory factor analyses of those instruments are well established. Consequently, we directly employed relevant scales from those instruments.

2. Sirotnik (1980) distinguished between two phases in data analysis (the psychometric phase and the study phase). Factor loadings of indicators (to constructs) are typically reported in the psychometric phase. Because our research is an empirical effort of the study phase in nature, we reported descriptive statistics of indicators following the standard statistical practice.

3. The research literature on school effects distinguishes between contextual and evaluative characteristics of schools (Teddlie & Reynolds, 2000). Contextual variables are more formative because educators have little control over school context. Evaluative variables are more reflective, describing climatic environment and interpersonal relationship within a school that are under the direct control of educators.

4. For example, TLI is calculated as (χ²/df_(null) − χ²/df_(proposed))/(χ²/df_(null) − 1).

Additional information

Notes on contributors

Jianping Shen

Jianping Shen is the John E. Sandberg Professor of Education and currently the Gwen Frostic Endowed Chair in Department of Educational Leadership, Research and Technology at Western Michigan University, Kalamazoo, MI 49008. Email: [email protected]. He earned his PhD at the University of Washington in Seattle. He published extensively on educational leadership, policy analysis and research methods.

Xin Ma

Xin Ma is a professor of Education Statistics and Mathematics Education at the University of Kentucky. Email: [email protected]. His main research interests include quantitative methods, policy analysis, school effectiveness and mathematics education. He has published numerous refereed articles in prestigious academic journals such as American Educational Research Journal, American Journal of Education, Educational Evaluation and Policy Analysis, Journal for Research in Mathematics Education and Teachers College Record.

Van E. Cooley

Van E. Cooley was a full professor, chair and interim Dean in the Department of Educational Leadership, Research and Technology, College of Education and Human Development at Western Michigan University, Kalamazoo, MI 49008. Email: [email protected].

Walter L. Burt

Walter L. Burt is an associate professor in the Department of Educational Leadership, Research and Technology, College of Education and Human Development at Western Michigan University, Kalamazoo, MI 49008. Email: [email protected]. He has had extensive experience in the k-12 system, a foundation, and the academia.