Left-Handed People Are Being Unnecessarily Excluded From Neuroimaging Research, Study Finds

Detail of a no left turn road sign and a hanging yellow traffic light in downtown Manhattan

By Matthew Warren

In a world made for right-handed people, life can sometimes be frustrating if you are among the 10% or so who are “adextral” — that is, left-handed or ambidextrous. Now a new grievance can be added to the list. Brain imaging researchers are systematically excluding adextrals from participating in their studies, according to an analysis of recent research papers published in top neuroimaging journals. Yet there’s no good reason to exclude this population, say the authors — and in fact, the practice could be detrimental to research.

The tendency to include only right-handed people in neuroimaging research stems from the finding that certain processes are represented differently in the brains of adextrals. In particular, while the left hemisphere of the brain dominates language processing in almost all right-handed people, in about 30% of adextrals this processing occurs predominantly in the right hemisphere (or both hemispheres). Handedness also influences how the brain represents sensation and movement of the hands. Many neuropsychologists have therefore avoided recruiting adextrals for fear of affecting their data.

But Lyam Bailey and colleagues at Dalhousie University noticed that researchers were excluding adextrals from a range of studies unrelated to language or movement, even though there is little evidence that handedness influences other processes. So in their new paper, published in the European of Journal of Neuroscience, the team investigated whether there is evidence for this bias in the literature. The researchers looked at 1,008 articles published in three major neuroimaging journals across the course of 2017, noting whether each study reported participants’ handedness — and if so, how many adextral participants were included.

The team found that a little over half of the studies reported participants’ handedness. But of the almost 32,000 people in those studies, only 3.2% were adextral — considerably lower than the rate of 10-13% in the general population. Overall, fewer than one in five of the studies that reported handedness included any adextrals at all, suggesting that most were actively excluding these people.

Intriguingly, studies looking at clinical populations and children were more likely to include adextral participants than those on non-clinical and adult populations. The authors suggest that this may be because it is harder to recruit from these populations, so researchers are more willing to include non-right-handed people.

The findings are not necessarily that surprising, though it is interesting to see the hard data to back up what many neuropsychologists may have already expected. But the key question is — does it matter that adextrals are underrepresented?

The authors make a convincing case that it does. Handedness is just another kind of individual difference, the researchers write, much like memory capacity or socioeconomic status. Both of those factors can also influence how language is represented in the brain, but are not normally used as a basis for excluding people from research. And we can only fully understand whether there are any differences in the brains of right- and left-handed people if both groups are actually included in research. “Handedness should be treated with the same measured scientific curiosity that is now afforded to individual differences such as language proficiency, working memory capacity [and] SES,” write the researchers. Instead of excluding adextrals, they add, statistical methods can be used to control for any potential effects of handedness.

And if that’s not enough, the logic for excluding adextrals even from language studies is flawed, say Bailey and colleagues. Although 30% of adextrals don’t show the normal left hemisphere language lateralisation, the same is true for about 4% of right-handed people. And because adextrals already represent a much smaller proportion of the population, including them will have little impact on how many participants have this “atypical lateralisation”.

The researchers say they hope their work will encourage others to take a hard look at whether they really need to be excluding adextrals from their studies. “Such a change in attitudes will surely benefit both our understanding of how handedness relates to brain function, organization and structure and, more broadly, our understanding of the brain at the whole-population level,” they write.

A sinister subject: Quantifying handedness-based recruitment biases in current neuroimaging research

Matthew Warren (@MattbWarren) is Editor of BPS Research Digest

7 thoughts on “Left-Handed People Are Being Unnecessarily Excluded From Neuroimaging Research, Study Finds”

  1. Not to mention that this might also affect what we know about the unique positives that left-handed people can contribute, For example, many teachers I know, and I have heard similar proclamations from organizational consultants that “lefties” often come up with unique and helpful solutions that “righties” do not come up with.
    Wouldn’t it be great to see if we can understand IF that is true, and if so, what specific brain functioning and where in the brain that happens?

  2. Congrats to the authors! This is an excellent review article.
    I would like to take a step back and point to a more fundamental underlying problem: The problem of handedness classification in empirical studies. Standard procedures use self-declared handedness of participants often measured by means of established inventories (e.g., Edinburgh Handedness Inventory). The problem with these preference-based approaches is the implicit risk of misclassification: Self-declaration is an unreliable measurement of handedness. In other words: it is very likely that the group of right-handers in the sample of Neuro-Imaging studies will be contaminated by false negative participants (so-called right-preferred non-right-handers) and the group of excluded left-handers could also include false positive participants.

    How can the severe risk of misclassification be reduced? As addressed in a previous paper (Kopiez et al. (2010). https://www.tandfonline.com/doi/abs/10.1080/13576500902885975), an objective procedure for the determination of the “true” or designated handedness should use a performance-based handedness classification procedure (e.g., speed tapping with measurement of hand performance differences). Surprisingly, based on the paradigm of hand performance differences, the proportion of non-right-handers is always significantly higher when compared to a classification methods based on self-reports or inventories. The reported extremely low proportion of 3-4% of non-right handers (which is only one third of the estimated proportion of non-right-handers in the general population) in Bailey’s review, could indeed be based on cultural differences (as assumed by the authors), however, this surprisingly low proportion could also be explained as a result of false negative misclassifications. As a consequence, we cannot exclude that an unkown number of unidentified non-right-handers have been included in the studies, causing an unkown degree of contamination.

    To put it simply: for the control of handedness in experimental studies, I am arguing for the future use of (time-critical and computer-based) hand performance measurements and a farewell from fuzzy classification criteria derived from self-reported (inventory-based) preference handedness.

Comments are closed.