Parsing eye-tracking data of variable quality to provide accurate fixation duration estimates in infants and adults

Abstract

Researchers studying infants’ spontaneous allocation of attention have traditionally relied on hand-coding infants’ direction of gaze from videos; these techniques have low temporal and spatial resolution and are labor intensive. Eye-tracking technology potentially allows for much more precise measurement of how attention is allocated at the subsecond scale, but a number of technical and methodological issues have given rise to caution about the quality and reliability of high temporal resolution data obtained from infants. We present analyses suggesting that when standard dispersal-based fixation detection algorithms are used to parse eye-tracking data obtained from infants, the results appear to be heavily influenced by interindividual variations in data quality. We discuss the causes of these artifacts, including fragmentary fixations arising from flickery or unreliable contact with the eyetracker and variable degrees of imprecision in reported position of gaze. We also present new algorithms designed to cope with these problems by including a number of new post hoc verification checks to identify and eliminate fixations that may be artifactual. We assess the results of our algorithms by testing their reliability using a variety of methods and on several data sets. We contend that, with appropriate data analysis methods, fixation duration can be a reliable and stable measure in infants. We conclude by discussing ways in which studying fixation durations during unconstrained orienting may offer insights into the relationship between attention and learning in naturalistic settings.

Disentangling the mechanisms underlying infant fixation durations in scene perception: A computational account

“…Eye-tracking data from infants may contain considerably higher levels of noise than data from more compliant participants such as adults due to various factors including their high degree of movement, lack of compliance to the task, poor calibration and corneal reflection disturbances due to the underdeveloped cornea and iris (Hessels, Andersson, Hooge, Nyström, & Kemner, 2015;Saez de Urabain, Johnson, & Smith, 2015;Wass, Smith, & Johnson, 2013). To account for this potential quality/age confound, dedicated in-house software for parsing and cleaning eye tracking data has been developed (GraFix, Saez de Urabain et al, 2015).…”

An Open Conversation on Using Eye-Gaze Methods in Studies of Neurodevelopmental Disorders

“…Although a number of published empirical articles (Nyström, Andersson, Holmqvist, & van de Weijer, 2013;Wass et al, 2014;Wass, Smith, & Johnson, 2013), methodological reviews (Gredebäck, Johnson, & von Hofsten, 2010;Karatekin, 2007;Oakes, 2012), publication guidelines (e.g., Oakes, 2010), and conferences (e.g., EyeTracKids) have addressed the use of eye-gaze measures to study typical development, methodological issues specific to children with NDD have received less attention (but see Kylliäinen et al, 2014;Sasson & Elison, 2012). Discussing this topic is worthwhile because the issues that arise in studies of individuals with NDD may differ from those that arise in studies of typically developing individuals.…”

“…Manual coding is neither automatic nor free of the risk of bias because it relies on human coders to make decisions about gaze location (Oakes, 2012;Wass et al, 2013). In addition, manual coding takes time (e.g., 1-2 hr for a 4-min video) as does training coders to lab standards and coordinating lab-wide agreement checks (see "Manual Eye-Gaze Coding").…”

“…One option is to use default algorithms to define the temporal and spatial parameters of these behaviors; however, default processing algorithms may make assumptions of which the researcher is not aware (Oakes, 2010). Default algorithms may also produce undesirable systematic relationships between data quality and fixation length, potentially allowing poor measurement quality to obscure true gaze patterns (Wass et al, 2013). Another option is to customize algorithms.…”

“…Clear reporting of the parameters used in specific studies is a necessary first step in better understanding what constitutes a fixation in different populations of individuals with NDD (Gillespie-Lynch, Elias, Escudero, Hutman, & Johnson, 2013). It is particularly beneficial to the research community when researchers share new procedures they have developed to address limitations of default processing algorithms (Wass et al, 2013).…”