Intercorrelation of Some Measures of Time Estimation

Experiments on time estimation have used different methods to ascertain the magnitude of Ss' subjective time units or the speed of their internal clocks. The three major methods have been described and compared by Bindra and Waksberg (1956). In the method of uerbul estimation, E delimits operatively a period of time (standard), and S is asked to give a verbal estimation of its duration (judgment). In the production method, S is required to delimit operatively (judgment) an interval of time stated verbally by E (standard). In the method of reproductiolz, E delimits operatively an interval of time (standard) and then asks S to reproduce operatively the same duration (judgment). It has been assumed that all three methods yield equivalent measures of SVs subjective time units. Consequently, generalizations have been made about the nature of subjective time on the basis of results which were obtained with any one of these three methods (Wallace & Rabin, 1960). Thus, one should obtain significant correlations between subjective time-unit scores which have been obtained by any two of the three methods. The correlations should not be significantly lower than the reliability value of the least reliable of the rwo methods under comparison. The present study will report some data which bear directly on these hypotheses. Method.-Ss were 36 undergraduates at Bar-Ilan University, Israel. Repeated estimations of 5and 20-sec. intervals were obtained in order to ascertain the reliability of Ss' judgments of these time-intervals for each method. In verbal estimation, E delimited the time-intervals with the click of a stop watch. The same method was used by S in the method of production and by both E and S in the method of reproduction, the face of S's watch being covered. The interval between stimuli was 10 sec. Ss were assigned randomly to three groups which differed with respect to the serial order in which the methods were employed. Half of the groups judged 5-, 20-, 5-, and 20-sec. intervals, in that order; the other half judged in the reverse order. Thus, each S made 12 judgments, four by each of three methods. Ss were told that the purpose of the study was to determine how certain time units felt, and that no attempt should be made to count or to use other mnemonic devices. Resalts.-Since neither order of presentation of stimuli nor the serial positions of the measuring methods were significant sources of variance (Fs <