Measures of difficulty in completion tests
AbstractCriticizes a paper entitled "A Tentative Standardization of the Completion Test'See PA, Vol 7(6). The method used for assigning the values to different blanks of the test was based on the frequency with which degree of difficulty was specified by the Ss. Questions the assumption of the method that an inverse rectilinear relationship exists between the frequency with which a particular judgement of difficulty is given and the amount of difficulty which that judgement implies. Proposes a more direct method involving measuring the actual time taken by Ss for the completion of each of the elisions, and on basis of these times defining values of the difficulties.