Invalid responding in questionnaire-based research: Implications for the study of schizotypy.

Abstract

Data collected through self-report questionnaires are particularly susceptible to inappropriate or random responding, and such invalid data increase noise and attenuate true statistical relationships. While many researchers studying schizotypy have employed infrequency measures to exclude participants, such measures are not universally employed. Moreover, some researchers have even outright questioned whether validity scales are warranted. Here, we show the effect of invalid responses on the relationship between schizotypy and hedonic reaction. For valid responders, negative schizotypal traits were inversely related to both anticipatory and consummatory pleasure (p < .01). Invalid responses were found for 23% of respondents, and within these subjects, no relationship was found between any of the measures. When the valid and invalid respondents were pooled, the relationship was dampened. Furthermore, linear multiple regression modeling showed that validity trended toward moderating the relationship between the variables of interest. These data highlight the importance of screening for, and excluding, invalid responses in schizotypy research. Our results also affirm that screening for random responding is effective and warranted. Implications for future studies employing questionnaire-based methods are discussed.