Wait wait, don't tell me: Handedness questionnaires do not predict hand preference for grasping

ABSTRACT

Handedness questionnaires are a common screening tool in psychology and neuroscience, used whenever a participant's performance on a given task may conceivably be affected by their laterality. Two widely-used examples of such questionnaires are the Edinburgh Handedness Inventory and the Waterloo Handedness Questionnaire. Both instruments ask respondents to report their hand preference for performing a variety of common tasks (e.g., throwing a ball, or opening a drawer). Here we combined questions from the two instruments (E-WHQ; 22 questions total) and asked participants to report their preferred hand for each via a five-point scale. The purpose of this study was to determine whether responses on the E-WHQ are accurate, reliable, and/or predictive of hand-preference for a simple grasp-to-construct task. Regarding accuracy, handedness scores were 5% lower when participants used a scrambled response key versus a consistent one. Test-retest reliability of the questionnaire was weak, with any given inventory item eliciting a different response from 34% of respondents upon retesting. Neither was the E-WHQ predictively useful—although both left- and right-handers preferred their dominant hands, E-WHQ score did not correlate with overall percentage of dominant-hand grasps in either group. We conclude that the E-WHQ is unsuited for predicting hand preference for grasping.

KEYWORDS:

• Handedness questionnaire
• hand preference
• grasping
• Edinburgh
• Waterloo

This article refers to:

Don’t expect the unexpectable: Commentary to Flindall and Gonzalez (2018)

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Information on age and gender were reported at each participant's own discretion; as such, gender information was available for 90% of participants, while age information was only available for 50% of participants. However, as all participants were recruited from the same pool of psychology, kinesiology, and neuroscience undergraduate students at the University of Lethbridge, the demographic information reported here is assumed to be representative of our overall sample.

2 We are aware that past research has recommended polychoric correlation to assess the relationship between items on the EHI (Dragovic, 2004). In this case, however, our dependent variable (probability of right hand use in a grasping task) is measured on a continuous scale, and our independent variable (E-WHQ score) is reported on an 88-point ordinal scale (which may safely be treated as continuous; see Rhemtulla, Brosseau-Liard, & Savalei, 2012); for these reasons (and again, because our data fails the test of normality), we have instead used Kendall's tau to estimate rank correlation.

Additional information

Funding

This work was supported by Natural Sciences and Engineering Research Council of Canada [grant number CGSD2 - 476054 - 2015].