Schools As Learning Organizations: Evaluating Factor Structure and Measurement Invariance across Professional Role

Background and Purpose: Public schools face the challenge of meeting increasing performance standards without significant changes in support. One way to improve student achievement is to promote working relationships among school employees. Organizational learning—a special type of organizational culture in which the organization values, acquires and uses information from stakeholders to plan, implement and evaluate strategies to achieve goals—represents one way of conceptualizing the working relationships among school employees. This study uses data from the School Success Profile Learning Organization (SSP-LO) to assess the factor structure of organizational learning.

Methods: The SSP-LO was administered to 3,065 employees at 68 schools from 2005 to 2009.  The SSP-LO used a 36-item learning organization measure and a six-point response scale to assess 12 organizational dimensions (common purpose, respect, cohesion, trust, mutual support, and optimism, team orientation, innovation, involvement, information flow, tolerance for error, and results orientation). Confirmatory factor analysis (CFA) were used to validate the two-correlated factor model proposed by Bowen, Rose and Ware (2006). The weighted least squares estimator adjusted for means and variances (WLSMV) in Mplus 7.11 was used. Invariance across participant role (i.e., teaching versus non-teaching staff) was assessed.

Results: The model with a two correlated first-order factor structure had a good fit (χ2=1496.554, df=241, p<.001, RMSEA=.043, 90% CI [.041-.045], CFI=.963, TLI=.958, two factors correlation=.459). The invariance test results show that the separate models for teaching and non-teaching staff had good model fit with the data with factor loadings ≥ .30. The configural model showed excellent fit with the data (χ2=1603.08, df=478, p<.001, RMSEA=.041, 90% CI [.038-.043], CFI=.974, TLI=.963), suggesting that teaching and non-teaching staff had similar factor structure. Similarly, the factor loadings were invariant (χ2=1634.26, df=501, p<.001, RMSEA=.040, 90% CI [.038-.042], CFI=.968, TLI=.964). The chi-square difference test between the configural and model with all factor loadings invariant yielded statistically nonsignificant results (Δχ2 = 20.23, Δdf = 23), suggesting that the data support the equivalence of factor loadings for teaching and non-teaching staff.

Discussion: Although school organizational features have long been deemed critical to students’ success, measurement tools to assess these features of schools are underdeveloped. This study found that the SSP-LO validly and reliably measures two domains of organization learning regardless of professional role of the organizational staff.

Implications: The SSP-LO has wide application and may be used with teaching and non-teaching staff. Thus, researchers and practitioners do not have to invest time or resources in developing role-specific instruments, thereby addressing some of the financial challenges to the widespread use of validated instruments.