Validation and DIF Analysis of a School Climate Measure

• In Event: We don't need no Stinking Reliability! Data Assessments of Reliability and Validity

Abstract

Research suggests school success is in part determined by a school’s climate and should be considered in improvement/reform strategies (Cohen et al., 2009). One of the largest and least addressed problems within the field of school climate is the lack of commonality in the instrument, scaling and treatment of data. A Rasch model allows for the purposeful examination of a measure to provide validity and stability for future use. An Andrich-Rasch rating scale (RS) model was applied on a 23-item school climate scale for unidimensionality, item fit and difficulty, and reliability with a portion (n=500) of the larger sample (n=4974). A DIF analysis will be conducted to determine if differences exist between 3rd and 8th grade teachers.

Author

• Audrey Conway Roberts, University of Kentucky