Extended Whipple’s index approach to analyze proxy response and rounding in travel surveys
基于Whipple指数开发新指标，利用出发时间取整程度评估出行调查质量，发现代理回答比自答更易取整，且取整与代理报告次数负相关。
• New index to evaluate travel survey quality using departure time rounding. • Proxy responses contain more rounded reported time than self-response. • Travel log reduces rounding in reported time. • Proxies contain more instances of rounding if completers report records rounded time. • Proxy reported trips decreases with increased rounding in completers report. Proxy response bias in travel surveys affects the quality of reported data. The resulting quality deterioration should be evaluated carefully. In this study, we employed an index for evaluating survey quality using rounding in reported trip departure times. The index was developed based on Whipple’s index, which measures inaccuracies in age reporting in demography. Data from the 2017 US National Household Travel Survey was used to evaluate the effectiveness of the proposed index. The results demonstrate that 1) proxy response produced more rounded reporting than the self-response, 2) proxy-reported trips accompanied by proxy completers contained fewer rounded records, 3) proxy-reported records completed by those who reported more rounded records in their surveys often contain more instances of rounding, and 4) the number of proxy-reported trips decreases with an increase in rounding in the completers report. The findings of this study will be useful in correcting proxy response bias in future studies, and the index can be used to evaluate the quality of existing travel survey data.