New preprint!

Hello there!

A new preprint is out! link

The concept of Math Anxiety (MA) – that is, an excessive fear or worry for math-related situations – has gathered much interest in the last few years because it hinders significantly learning and scholastic achievements. Poor numeracy is associated to early drop-out from school and difficulties in maintaining a stable employment, with particular groups (e.g., women) that are disproportionally more vulnerable. Yet, we lack an agreed upon pipeline for optimal identification of MA, including objective tools (i.e., beyond self-report measures). We are therefore witnessing a frantic race for the discovery of new biomarkers of MA, though most of them remain very unpractical (e.g., fMRI).
In this study we explore the potential of Pupil Size (PS). Pupillometry is an inexpensive, easy to administer tool that informs about the balance between sympathetic and parasympathetic autonomic systems; hence, it is in principle very well suited to measure both the broad range of autonomic reactions that defines anxiety and differences in the subjective cognitive effort invested. We assess PS in the context of a novel behavioral paradigm that accounts for several stages of mental calculation: from anticipation (the prospect of mental calculation) to the feedback (the judgment about one’s performance), all very relevant for MA and providing a very rich potential for psychophysical modelling. We wondered specifically: if MA could modulate PS beyond the impact of mathematical competence; whether PS could explain part of the variance in the scores obtained in the elective questionnaire to measure MA. We did so striving to adopt robust methodology (our sample size, N= 70, is more than twice the sample size commonly observed in similar studies) and fully open procedures: all the materials, data, and analysis scripts are openly available (https://osf.io/szb24/) alongside custom functions to analyze pupil data (https://github.com/EBlini/Pupilla). We found that MA did not modulate PS when behavioral accuracy was also used as a predictor. This can be taken as a testament of the close link between MA and math competence, which we confirmed with our psychometric battery and that may be too firmly crystalized in young adults. On the other hand, we found that the latency of PS peak dilation could explain a substantial share of variance in the questionnaires assessing math and test anxiety. Thus, not much the extent by which the pupils dilate, but rather the duration of cognitive effort was a good predictor of high anxiety levels. Results thus point to the existence of a suboptimal efficiency of information processing in anxiety that hampers performance and should therefore be the target of early educational interventions.

On a personal note: this is my first study out from beautiful Florence, and I am quite happy about how the paper currently reads. Looking forward to receive comments and reviews!

The full article is available on biorxiv: https://www.biorxiv.org/content/10.1101/2023.12.15.571819v1