Improving Tools for the Analysis of Brain
Based Measures of Infant Information
Processing

Abstract

While current methodological approaches to data-acquisition and -analysis used in ERP-studies have shown to be appropriate for adult populations, their applicability with infant populations is debatable. Many researchers agree that they are, in fact, unsuitable for infants. However, due to a lack of alternatives, traditional ways of data-collection and -analysis originally designed for adults are still used with developmental and clinical populations. The four studies reported in the current thesis propose novel approaches to methodological issues involved in infant ERP-studies. Limiting the number of artefact-free trials included in the average, Study 1 shows that fewer trials from earlier stages of a test-session can convey meaningful information and provide new insights into infant cognition. Additionally, it is shown that large amounts of usable data are often discarded in a traditional data-analysis. Study 2 aimed to provide a general overview on attrition in infant ERP-studies – which often seems to be accepted as a given among developmental psychologists – and to find underlying causes for attrition which may be based on study-design features that are common to all studies. The majority of investigated features did not impact attrition. Therefore, further factors outside the scope of the current consideration are responsible for an increase in attrition. Studies 3 and 4 built on the findings of the first two. Presenting infants with eight experimental conditions, Study 3 illustrates a novel approach to study-design which both yielded more data from individual participants and decreased attrition. Finally, using the data-analysis outlined in Study 1 on the data collected for Study 3, Study 4 both substantiates the claims made in Study 1 and provides further, previously unanticipated insights into infant cognition. These findings illustrate that more infant-suitable strategies for study-design and data-analysis are possible and that they can enhance our knowledge about infant cognitive processes.