Pupillometry of taste: Methodological guide – from acquisition to data processing - and toolbox for MATLAB

Anaïs Lemercier,Thierry Baccino,Geneviève Guillot,Pascal Schlich,Claire Garrel,Philippe Courcoux

doi:10.20982/tqmp.10.2.p179

Abstract

The food sensory science community has recently shown an increasing interest towards the measurement of emotions induced by product consumption. In this paper, we propose a standard method to objectively and quantitatively explore emotions in tasting situations through measurement and analysis of pupil diameter. Pupillometry is a well-knownmethod for investigating cognitive load and emotional effects. However, since a standard method to analyze pupil response data is lacking, we provide here a guided methodology, from data acquisition to dataprocessing. An example of an application is also provided. The advantages and major drawbacks of this method are discussed.

Highlights

Pupil studies have been used in psychology for many years
We propose guidelines for conducting a pupillometry study, some of which are specific to the product tasting situation when using an on-screen eye tracker
Computing a covariance Principal Component Analysis (PCA) where subject x condition are the observations and where the pupil diameter for the significant times (280 ms-1600 ms) are the variables, we identified one principal component (PC) to retain according to the elbow criterion

Summary

Introduction

Pupil studies have been used in psychology for many years. In her seminal work, Loewenfeld (1993) provides a history of pupil use in psychology, including research about voluntary pupil movements in the 18th century, “conditioned” and “orienting” reflexes at the beginning of the 20th century, and “attitude,” “interest,” etc., taken up by Hess in the 1960s. Since the 1960s, various pupillometry studies examined cognitive and emotional stimuli. Researchers have reported that pupil size is correlated to levels of cognitive effort (Hess & Polt, 1966; Bradshaw, 1968; Ullwer, Ries, Foth, & van det Meer, 2010), and to emotional load or preferences (Barlow, 1969; Partala & Surakka, 2003; Bradley, Miccoli, Escrig, & Lang, 2008; Laeng, 2010).

Objectives

Results

Conclusion