Measuring Teacher Dispositions Toward Teaching Sustainable Systems: A Mixed Methods Approach for Instrument Development and Psychometric Testing

摘要

Sustainability concerns have become prevalent in environmental, societal, and economic systems. To address education towards sustainability the need for explicit instruction in sustainable systems is apparent; however, it is underrepresented in American schools. Despite the emergence of sustainability topics in the literature, few have addressed teacher dispositions about providing this needed instruction and none have reported quantitative measures with acceptable estimates of reliability and validity.Dispositions are defined as the tendency to act in a particular manner that aligns with an individual’s belief which can develop and change over time, and are influenced by the experiences and circumstances faced by the individual. To provide the necessary instruction not only are instructors responsible for the curricular content and pedagogical content, but most importantly, they must possess positive dispositions towards providing this instruction.The purpose of this study was to construct and determine estimates of the validity and reliability of the Dispositions Toward Teaching Sustainable Systems Instrument. Using a sequential mixed methods design and Luyt’s Framework for instrument development the researcher, using qualitative methods such as interviews, identified themes that were supportive of the theoretical construct. The four themes that emerged were: administrative support, outdoor resources, collaboration, and professional development, all receiving above 80% interrater agreement (n=3 judges). The quantitative aspect identified the key indicators (items) and their estimates of reliability and validity, and their goodness of fit to the four theoretical constructs.Content validity estimates of the items generated from the interviews were explored using expert judges (n=2) in the area of sustainability education, also achieving above 80% agreement. Participants enrolled in three succeeding semesters of the Teaching Elementary Science Methods course (n=233) were then used to further estimate the reliability (α u3e.08), construct validity (normed chi-Square = 2.04; GFI=.86 RMSEA =.076; NFI=.96 CFI =.98 PGFI=.634 binomial index of model fit ppR=.271 R2=.073 (4,227)=4.5 pp