Construction and preliminary validation of an inconsistent responding scale for the HEXACO-100.

Abstract

Inconsistent responding poses a particular risk to self-report data and is a common phenomenon in undergraduate and community subject pools. The HEXACO-100 personality inventory (Lee & Ashton, 2018) lacks internal validity scales that could be used to identify inconsistent responding. To address this need, the present study developed a HEXACO Inconsistent Responding Test (HIRT). Highly correlated and conceptually overlapping HEXACO-100 item pairs were identified from a derivation sample and cross-validated by comparing real data to randomly generated protocols. The HIRT was highly accurate at identifying both randomly generated and half-randomly generated protocols. HIRT scores exhibited significant correlations with other inconsistency measures (i.e., the Triarchic Assessment Procedure for Inconsistent Responding and the HEXACO’s Inter-item Standard Deviation) and were as effective as these other measures at differentiating between consistent and inconsistent responders. This preliminary validation of the HIRT provides strong evidence of its ability to improve the quality of data gathered with the HEXACO-100, which may be especially useful for researchers hoping to utilize the measure in clinically relevant research and applications.

Development of an Inconsistent Responding Scale for the HEXACO Personality Inventory-Revised

“…Notably, the ability of the BRIE to effectively differentiate between random and genuine respondents parallels findings for the HIRT (Lambros et al, 2022). The two inconsistency scales were strongly correlated ( r > .70), which reflects both the substantial overlap in item content and similar methodological approaches to scale development.…”

“…Moreover, researchers interested in using the HEXACO-100 have two options to screen careless respondents. The AUC values for the BRIE (>.95) and HIRT (.98;Lambros et al, 2022) do not indicate if one necessarily performs better than the other. We suggest researchers using the HEXACO-100 consider using scales in tandem to remove potential participants, such that participants are retained if they pass both scales.…”

“…We also expected individuals with higher scores on our new inconsistency scale would self-report greater personality traits, reflecting poor impulse control (Bowling et al, 2016). Finally, we examine the overlap of our scale with the HIRT developed for the HEXACO-100 (Lambros et al, 2022).…”

“…Finally, we examined correlations between this new HEXACO scale and measures of careless responding and personality traits within undergraduate and community samples. Specifically, we correlated this scale with existing measures of careless responding, the HEXACO’s Standard Deviation index to capture incoherent responses (HEXACO- SD ; see Barends & de Vries, 2019; Lee & Ashton, 2018, p. 545; for a similar approach, see Marjanovic et al, 2015), the HEXACO-100’s HIRT (Lambros et al, 2022), and TriPM’s TAPIR (Mowle et al, 2017) as a test of convergent validity. We also expected individuals with higher scores on our new inconsistency scale would self-report greater personality traits, reflecting poor impulse control (Bowling et al, 2016).…”

“…Recently, Lambros et al (2022) developed the HEXACO Inconsistent Responding Test (HIRT)-an inconsistent responding scale for the 100-item version of the HEXACO-PI-R that effectively differentiated between true and random respondents. Lambros et al (2022) developed this scale independent of the current project. However, 13 of the 19 item pairs include one item that is not present on the 60-item version of the HEXACO inventory.…”