Viswanathan, M., Kayande, U., Bagozzi, R. P., Riethmuller, S., & Cheung, S. Y. Y.
Impact of questionnaire format on reliability, validity, and hypothesis testing
Researchers use a variety of questionnaire formats to collect data on measures of constructs for theory testing. For example, researchers may label measures of constructs, present measures of different constructs on different pages, or intersperse items from different constructs. Such questionnaire format choices are often guided by commonly held beliefs and conventions. However, there is little if any empirical research evaluating such different formats, precluding an informed view of the appropriateness of a format for a study. To our knowledge, our research represents the first systematic empirical investigation of how different formats affect research outcomes. We conduct a series of studies to systematically examine the effects of questionnaire format on the psychometric properties of measures of constructs and the relationships between constructs. Using multiple group confirmatory factor analysis (MGCFA), we find that the measures are largely invariant to questionnaire format when using student samples, and recommend that researchers should reconsider the rationale provided for choosing specific formats.
Testing, Psychometrics, Methodology in Applied Psychology, 2017, Vol. 24, pp. 465-498, DOI: 10.4473/TPM24.4.1Back