The Conscientious Responders Scale

Zdravko Marjanovic,C Ward Struthers,Robert Cribbie,Esther R Greenglass

doi:10.1177/2158244014545964

Zdravko Marjanovic, C Ward Struthers + Show 2 more

Open Access

https://doi.org/10.1177/2158244014545964

Copy DOI

Journal: SAGE Open	Publication Date: Jul 1, 2014
Citations: 91	License type: cc-by

Affiliation: Thompson Rivers University, York University

Abstract

This investigation introduces a novel tool for identifying conscientious responders (CRs) and random responders (RRs) in psychological inventory data. The Conscientious Responders Scale (CRS) is a five-item validity measure that uses instructional items to identify responders. Because each item instructs responders exactly how to answer that particular item, each response can be scored as either correct or incorrect. Given the long odds of answering a CRS item correctly by chance alone on a 7-point scale (14.29%), we reasoned that RRs would answer most items incorrectly, whereas CRs would answer them correctly. This rationale was evaluated in two experiments in which CRs’ CRS scores were compared against RRs’ scores. As predicted, results showed large differences in CRS scores across responder groups. Moreover, the CRS correctly classified responders as either conscientious or random with greater than 93% accuracy. Implications for the reliability and effectiveness of the CRS are discussed.

Highlights

When a self-report psychological inventory1 is administered, the expectation is that respondents follow testing instructions and answer its items as honestly and accurately as possible
The tool we developed for this investigation is called the Conscientious Responders Scale (CRS; see the appendix), which is a five-item variant of a traditional validity scale
Unlike the CRS, the size of the group difference in Pettit Random Responding Scale (PRRS) scores was inconsistent across studies

Summary

Introduction

When a self-report psychological inventory is administered, the expectation is that respondents follow testing instructions and answer its items as honestly and accurately as possible. For example, purposefully distort their responses to be perceived more positively or negatively than they really are This is known as faking good and faking bad, respectfully (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989). Bruehl, Lofland, Sherman, and Carlson (1998) showed this possibility in a clever study using a widely used pain inventory They concluded that if the measure was administered to a group of RRs in a clinical setting and their random responding went unidentified, 35% of them would be classified as having elevated levels of interpersonal distress and another 35% as being highly adaptive copers. It increases measurement error, making it more difficult to identify significant relations when they are present in data. An exemplar of an infrequency scale is the Minnesota Multiphasic Personality Inventory–2’s (MMPI-2) F Scale (Butcher et al, 1989)

Methods

Results

Discussion

Conclusion