1 Department of Psychology, Faculty of Social Sciences, Aarhus University, Aarhus University2 Department of Psychology and Behavioural Sciences, Aarhus BSS, Aarhus University3 Department of Psychology and Behavioural Sciences, Aarhus BSS, Aarhus University
The concept of ecological validity is central when considering an ecological approach to experimental research. However, the concept of ecological validity seem to occur rather seldom in discussion sections in journal articles on experimental infant reseach. In this conference presentation I shall attempt to apply an analysis based on the concept of ecological validity on a specific experimental study within the field of infant cognition. I will attempt to show that such an analysis can be worthwhile.The presentation falls in three sections. The first section is a presentation of the concept of ecological validity. The second section reports an original empirical study investigating whether familiar objects make a difference when 8-month-old infants attempt to individuate objects. Object individuation refers to the ability to decide the number of distinct objects present in a given scenario. It has become almost a convention to use novel objects when investigating infants' understanding of the physical world. Such a strategy is easely legitimized on methodological grounds. However, the exclusive use of novel objects prevent us from exploring the possibility that familiar objects might make a difference when infants reason about the physical world. For example, it might be argued that infants would probably pay more attention to ‘disappearing' familiar objects compared to novel ones. Using the so-called "violation-of-expectation" method, the study employed a design that recently (Krøjgaard, 2003) has been used to replicate the results obtained in a seminal study by Xu and Carey (1996) with 10-month-old infants. On alternating trials infants are presented with ‘expected' test events (in which the number of objects present behind an occluder corresponds to the preceeding introduction) and ‘unexpected' test events (in which the number of objects present behind the occluder violates the available information regarding the number of objects present during the introduction). Two groups of 8-month-old infants were tested with a toy teddy bear as target object. For the infants in the first group the teddy bear was novel, whereas for the infants in the second group the teddy bear was familiar (each infant in the second group had received an exemplar of the teddy bear three months prior to the test). Based on rationale that infants, all things equal, would look longer at the unexpected test compared to the expected test events, the looking times obtained in the test events were averaged and compared. The results revealed that infants from both groups failed to individuate the objects. Thus, being familiar to the objects that (apparently) disappeared in the ‘unexpected' test events had no effect on the infants' ability to individuate objects in the present design.In the third and final section the results obtained in the experiment are discussed, not only in relation to the existing litterature within the field of research of object individuation, but also by specificly applying the concept of ecological validity. It is argued, that the negative results obtained may at least in part be explained by incorporating the concept of ecological validity into the analysis, and hence that the concept of ecological validity can indeed be a usefull analytical tool when discussing experimental infant research.