Citation

Rubio-Fernández P. Infant Behav. Dev. 2019; 54: 177-188.

Affiliation

Massachusetts Institute of Technology, Department of Brain & Cognitive Sciences, 43 Vassar St., Building 46, Cambridge, MA, 02139, USA; University of Oslo, Department of Philosophy, Georg Morgenstiernes Hus, Blindernveien 31, Oslo, 0313, Norway. Electronic address: prubio@mit.edu.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.infbeh.2018.09.009

PMID

30814033

Abstract

The Violation-of-Expectation paradigm is a widespread paradigm in infancy research that relies on looking time as an index of surprise. This methodological review aims to increase the reliability of future VoE studies by proposing to standardize reporting practices in this literature. I review 15 VoE studies on false-belief reasoning, which used a variety of experimental parameters. An analysis of the distribution of p-values across experiments suggests an absence of p-hacking. However, there are potential concerns with the accuracy of their measures of infants' attention, as well as with the lack of a consensus on the parameters that should be used to set up VoE studies. I propose that (i) future VoE studies ought to report not only looking times (as a measure of attention) but also looking-away times (as an equally important measure of distraction); (ii) VoE studies must offer theoretical justification for the parameters they use, and (iii) when parameters are selected through piloting, pilot data must be reported in order to understand how parameters were selected. Future VoE studies ought to maximize the accuracy of their measures of infants' attention since the reliability of their results and the validity of their conclusions both depend on the accuracy of their measures.

Copyright © 2018 Elsevier Inc. All rights reserved.

Language: en

Keywords

Accuracy; End-of-trial criteria; Eye-tracking measures; False-belief reasoning; Piloting; Surprise; p-curve analysis