Outcome-Reporting Bias: Correlates for Incomplete Reporting in Educational Research

• In Event: 65.065 - Perspectives in Conducting and Disseminating Research

Abstract

Outcome reporting bias refers to the practice of omitting from primary studies outcomes that were actually collected. Pigott et al. (2013) found that non-statistically significant outcomes were 30% more likely to be omitted from a published version of a dissertation when compared to those outcomes that were statistically significant. The proposed paper is to extend the prior work on outcome reporting bias in educational research to examine other potential correlates for incomplete reporting. For example, outcome reporting bias may be related to characteristics of the outcome such as whether the outcome was considered a primary or a secondary focus of an intervention, whether the outcome was a potentially harmful effect and whether the measurement strategy was reliable and valid.

Authors

• Brian Trainor, Loyola University Chicago

• Joshua R. Polanin, Development Services Group

• Ryan Williams, American Institutes for Research

• Jeffrey C. Valentine, University of Louisville

• Therese D. Pigott, Loyola University Chicago