• wednesday, 18 october 2017—12:15

    The contribution of big data tot word processing: Some recent developments

    Marc Brysbaert, Ghent University

    Word recognition has been investigated mostly with small factorial designs, in which one or two variables were manipulated and other variables controlled. Usually, the number of stimuli per condition was rather small (some 20), in line with the habit in psychology to run underpowered studies. In addition, the words often came from the extremes of the variables, meaning that participants were asked to respond to a series of very unusual words. Another approach is to collect data for thousands of words and to run multiple regression analyses. In this talk I discuss the latest discoveries, with particular attention to the effects of word frequency, word prevalence, and individual differences in word processing.

    external seminar