Measurement issues associated with conditional reasoning tests: An examination of faking

• Conditional reasoning tests measure personality by assessing implicit cognitions.
• Responses on CR-based measures are typically unaffected by response distortion.
• Two studies assessed the fakability of a CR measure of addiction proneness.
• Results indicate that the CR measure is resistant to response distortion.

Conditional reasoning (CR) is a new item format that measures personality by indirectly assessing reliance upon the cognitive biases associated with specific traits (James et al., 2005). Previous research suggests that, relative to self-report measures, responses on CR-based measures are more difficult to distort (LeBreton, Barksdale, Robin, & James, 2007). The issue of response distortion in the context of CR-based measures was evaluated in two studies. Study 1 (within-subjects) and Study 2 (between-subjects) both investigated whether responses on a CR-based test of addiction could be faked when indirect assessment was upheld. Results of both studies indicated that, unlike a self-report measure of a similar construct, the CR-based measure was unaffected by response distortion.