Statistical tests conducted with school environment data : the effect of teachers being clustered in schools

This article discusses the effect of clustering on statistical tests conducted with school environment data. Because most school environment studies involve the collection of data from teachers nested within schools, the hierarchical nature to these data cannot be ignored. In particular, this articl...

Full description

Saved in:
Bibliographic Details
Published inLearning environments research Vol. 12; no. 2; pp. 85 - 99
Main Author Dorman, Jeffrey P.
Format Journal Article
LanguageEnglish
Published Dordrecht Springer Netherlands 01.07.2009
Springer
Springer Nature B.V
Subjects
Catholic School Environment Questionnaire (CSEQ)
Catholic teachers
Cluster Grouping
Correlation
Data Analysis
Education
Educational Administration
Educational Environment
Educational Technology
Error patterns
Government school teachers
Learning and Instruction
Learning environment
Nested data
Original Paper
Outcomes of Education
Questionnaires
School culture
School Level Environment Questionnaire (SLEQ)
Science Teachers
Secondary education
Secondary schools
Sociology of Education
Statistical data
Statistical Inference
Statistical Significance
Statistical studies
Statistics
Studies
Teacher attitudes
Teacher Educators
Teachers
Test interpretation
Tests
Work environment
Work Environment Scale
Statistical testing
School environment
Effect of clustering
Online AccessGet full text

Cover

More Information
Summary:This article discusses the effect of clustering on statistical tests conducted with school environment data. Because most school environment studies involve the collection of data from teachers nested within schools, the hierarchical nature to these data cannot be ignored. In particular, this article considers the influence of intraschool correlations on tests of statistical significance conducted with the individual teacher as the unit of analysis. Theory that adjusts t test scores for nested data in two-group comparisons is presented and applied to school environment data. This article demonstrates that Type 1 error rates inflate greatly as the intraschool correlation increases. Because data analysis techniques that recognise the clustering of teachers in schools are essential, it is recommended that either multilevel analysis or adjustments to statistical parameters be undertaken in school environment studies involving nested data. [Author abstract]
Bibliography:Refereed article. Includes bibliographical references.
Learning Environments Research; v.12 n.2 p.85-99; 2009
ISSN:1387-1579
1573-1855
DOI:10.1007/s10984-009-9054-y