Construct Validity Evidence of the Design-based Professional Learning for Teacher Leaders Survey

Date: July 4, 2018
Author(s): Dr. Man-Wai Chu; Dr. Barbara Brown; Dr. Sharon Friesen
Presented at: bi-annual meeting of International Test Commission Conference in Montreal, QC, Canada.

Design-based professional learning (DBPL) advances principled practical knowledge and develops adaptive expertise within teachers to meet the diverse needs of students (Bereiter, 2014; Friesen & Jacobsen, 2015). However, DBPL is relatively new to the field of educational professional learning which is compounded in a field in which empirical evidence is sparse (Timperley, Wilson, Barrar, & Fung, 2007). A lack of resources to evaluate the benefits and successes that teachers may gain from DBPL sessions which are defined in terms of the impact that changed practice has on valued outcomes, is required (Coburn & Penuel, 2016; Wentworth, Mazzeo & Connolly, 2017). The DBPL for Teacher Leaders Survey (DBPLTLS) was developed by the authors to track teachers’ self-perceived gains in teaching effectiveness from DBPL. This study collected two sets of data from teacher leaders to investigate the survey’s construct validity (Kane, 2006). Both the exploratory (EFA) and confirmatory (CFA) factor analysis results indicate a six-factor design. The EFA (n=314) showed six factors (i.e., teacher understanding, authentic to the discipline, assessment, student collaboration, teacher collaboration, and teacher learning), each with three to six items that exhibited moderate to strong factor loadings (Tabachnik & Fidell, 2013). The CFA (n=257) of this six-factor model indicated strong to good model fit (RMSEA=0.042, CFI=0.968, TLI=0.958, and SRMR=0.040). The Cronbach alpha for each of the six factors indicates strong internal consistency (α=0.792-0.869). The construct validity presented in this study may allow researchers to use the DBPLTLS as a tool in a program evaluation of DBPL. The claims made regarding the six factors presented in the EFA and CFA of the DBPLTLS are supported by strong psychometric evidence. As research in the area of DBPL become popular, the DBPLTLS may become a useful tool to support claims of success or gains made from engaging in professional learning.

Chu, M-W., Brown, B., &Friesen, S. (2018, July). Construct Validity Evidence of the Design-based Professional Learning for Teacher Leaders Survey. Paper presented at the bi-annual meeting of International Test Commission Conference, Montreal, QC, CANADA