Key points are not available for this paper at this time.
Visual duration thresholds for words were measured under three conditions: (i) with a highly predictive context; (ii) with a lesser predictive context; (iii) a control condition with no context. The results show that the visual duration threshold for a word is reduced in the presence of a context by an amount depending on the probability of the word given the context. This result agrees with previous related findings. The errors made by the subjects could be classified into three main groups. The first of these groups showed the effect of expectancy, or response bias. Further errors showed the influence of previous stimuli or responses and the third group indicated the role of the stimulus in determining the response. A model for the recognition of words is outlined which accounts for these results and attempts to define what is meant by ‘the effects of response bias in the perception of words’.
John Morton (Fri,) studied this question.