Assessing the measurement invariance of latent job satisfaction ratings across survey administration modes for respondent subgroups: A MIMIC modeling approach

Assessing statistical aspects of test fairness with structural equation modelling

“…It is also possible to assess whether what a test measures depends on its format, such as 206 R.B. Kline paper-and-pencil administration versus computer administration of the same items in the same group (e.g., Whitaker & McKinney, 2007). Predictive bias is usually estimated at the level of test summary (total) scores, and it concerns the regression of scores from an external criterion on test summary scores.…”

A comparison of stimulus ratings made online and in person: Gender and method effects

“…Whitaker and McKinney (2007) and Norris et al (2007) demonstrated that even after completely removing what is often the largest single source of variance in psychological studies (i.e., individual differences), a method effect can be observed. However, in none of these studies was the form of the instrument held constant across testing conditions.…”

“…We included participant gender in our statistical models, since differences in the performance of subgroups of participants have not been explored extensively (Whitaker & McKinney, 2007). There is some evidence of differential gender effects in how participants make ratings such as these (e.g., Hewson & Charlton, 2005;Ullman et al, 2002) and also in how they respond to them (e.g., Wurm, Whitman, Seaman, Hill, & Ulstad, 2007).…”

“…The results suggested that the new battery was perfectly usable: "Results indicated comparability across the two batteries" (p. 940). Whitaker and McKinney (2007) had participants complete Internet and paper-and-pencil versions of a job satisfaction instrument. They found that job satisfaction ratings were quite comparable across methods (demonstrating measurement invariance), but they also found that a correlation between age and job satisfaction held only for the paper-and-pencil data.…”

Assessing Measurement Invariance of the Survey of Perceived Organizational Support (SPOS): Paper-and-Pencil vs. Online Administrations

“…measurement equivalence), especially if the researcher is planning to integrate data gathered via online surveys with those collected with paper-and-pencil based surveys. If a test of measurement invariance would reveal that the measurement instrument does not function in the same way across administration modes, it would not be appropriate to combine the data collected using these distinct methods (Fang et al, 2014;Teo, 2013;Whitaker and McKinney, 2007). Even if only online mode of data collection is employed through the research, it is still crucial to assure that the assumption of measurement invariance is not violated, so that one can justify the utilization of one specific mode of scale administration instead of another.…”