An experiment to test the reliability of estimates of use of time

Abstract

81 college students were requested to estimate the amount of time spent in each of 24 activities during an average week. For 10 days after this estimate the student recorded daily the use of his time. Students who reported themselves unwilling to co-operate, or who reported that the week in which the record was made was unusual, were not used as subjects. Two measures of reliability of the estimates were found for each activity; (1) the percentage difference between the estimates and the actual records, and (2) the coefficient of correlation between the estimate and the actual record. Items of doubtful reliability included time spent in letter writing, walking, music, walking, and riding for pleasure, and doing nothing. (PsycInfo Database Record (c) 2025 APA, all rights reserved)