Response Errors in the Measurement of Time Use
研究回忆期长度对时间使用数据有效性和质量的影响，发现周五至周日日记的回忆期可延长至一周，而周一至周四则需一天回忆；电话访谈技术改进和面板数据能提高数据质量。
Abstract The relation between (a) length of the recall period and (b) validity and quality of data on the use of time is examined. All of the data are based on 24-hour diary constructions or reconstructions. We find that recall periods of one day are essentially equivalent, in terms of validity and quality, to recall periods of up to one week for time diaries of Fridays, Saturdays, or Sundays, but that one-day recall is significantly better than more extended recall for Monday—Thursday diaries. These findings suggest that a sample design using randomly designated rather than convenient diary dates would be of acceptable quality in the United States. We also find that technical change in telephone interviewing techniques significantly improves the quality of current data over data collected in the mid-1970s, and that longitudinal panel data have significantly higher quality than cross-sectional data.