Statistical treatment of looking-time data. Developmental Psychology. 2016;52(4):521-36. Abstract
Statistical treatment of looking-time data
Looking times (LTs) are frequently measured in empirical research on infant cognition. We analyzed the statistical distribution of LTs across participants in order to develop recommendations for their treatment in infancy research. Our analyses focused on a common within-subject experimental design, in which longer looking to novel or unexpected stimuli is predicted. We analyzed data from two sources: an in-house set of LTs that included data from individual participants (47 experiments, 1584 observations), and a representative set of published papers reporting group-level LT statistics (149 experiments from 33 papers). We established that LTs are log-normally distributed across participants, and therefore should always be log-transformed before parametric statistical analyses. We estimated the typical size of significant effects in LT studies, which allowed us to make recommendations about setting sample sizes. We show how our estimate of the distribution of effect sizes of LT studies can be used to design experiments to be analyzed by Bayesian statistics, where the experimenter is required to determine in advance the predicted effect size rather than the sample size. We demonstrate the robustness of this method in both sets of LT experiments.
Learning in and about opaque worlds. Behavioral and Brain Sciences. 2015;38:49-50. Abstract
Learning in and about opaque worlds
We argue that direct active teaching in humans exhibits at least two properties (open-endedness and content opacity) that make the recognition of teaching episodes without ostension untenable. Thus, while we welcome Kline’s functional approach to the analysis of teaching, we think that she ignores important features of the socio- environmental niche in which human teaching likely evolved.
Giving and taking: Representational building blocks of active resource-transfer events in human infants
Active resource transfer is a pervasive and distinctive feature of human sociality. We hypothesized that humans possess an action schema of GIVING specific for representing social interactions based on material exchange, and specified the set of necessary assump- tions about giving events that this action schema should be equipped with. We tested this proposal by investigating how 12-month-old infants interpret abstract resource-transfer events. Across eight looking-time studies using a violation-of-expectation paradigm we found that infants were able to distinguish between kinematically identical giving and tak- ing actions. Despite the surface similarity between these two actions, only giving was rep- resented as an object-mediated social interaction. While we found no evidence that infants expected the target of a giving or taking action to reciprocate, the present results suggest that infants interpret giving as an inherently social action, which they can possibly use to map social relations via observing resource-transfer episodes.
Probing the strength of infants' preference for helpers over hinderers: Two replication attempts of Hamlin and Wynn (2011)
Several studies indicate that infants prefer individuals who act prosocially over those who act antisocially toward unrelated third parties. In the present study, we focused on a paradigm published by Kiley Hamlin and Karen Wynn in 2011. In this study, infants were habituated to a live puppet show in which a protagonist tried to open a box to retrieve a toy placed inside. The protagonist was either helped by a second puppet (the “Helper”), or hindered by a third puppet (the “Hinderer”). At test, infants were presented with the Helper and the Hinderer, and encouraged to reach for one of them. In the original study, 75% of 9-month-olds selected the Helper, arguably demonstrating a preference for prosocial over antisocial individuals. We conducted two studies with the aim of replicating this result. Each attempt was performed by a different group of experimenters. Study 1 followed the methods of the published study as faithfully as possible. Study 2 introduced slight modifications to the stimuli and the procedure following the guidelines generously provided by Kiley Hamlin and her collaborators. Yet, in our replication attempts, 9-month-olds’ preference for helpers over hinderers did not differ significantly from chance (62.5% and 50%, respectively, in Studies 1 and 2). Two types of factors could explain why our results differed from those of Hamlin and Wynn: minor methodological dissimilarities (in procedure, materials, or the population tested), or the effect size being smaller than originally assumed. We conclude that fine methodological details that are crucial to infants’ success in this task need to be identified to ensure the replicability of the original result.