Interpretations of the inverse relation between the duration for which a printed word must be exposed visually in order to be recognized and the frequency of occurrence of the word in a large general word count are examined critically. It is found that the most satisfactory interpretation is to regard the word's frequency in a word count as an estimate of its average probability of emission by the population of subjects used in the recognition experiment (base probability). Since the threshold of recognition is defined by the probability of emission of the word following its exposure, only a short exposure will be necessary to bring up to threshold criterion a word whose base probability is almost as great as the criterion probability, while a much longer exposure will be required for a word whose base probability is low. The critical assumption of this interpretation is that the frequency of a word in a large general word-count represents its average probability of emission by the experimental subjects used in the recognition experiment. Three experiments evaluating the validity of this assump¬tion for the Lorge Magazine Count are described
展开▼
