A Comparison of CAPI and PAPI through a Randomized Field Experiment

Caeyers, Bet ; Chalmers, Neil ; De Weerdt, Joachim

World Bank, Washington, DC

关键词: COMPUTER-ASSISTED PERSONAL INTERVIEWING;    CAPI;    SURVEY METHODOLOGY;    CONSUMPTION;    HOUSEHOLD WELFARE;   

RP-ID  :  148049

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

This paper reports on a randomizedsurvey experiment among one thousand eight hundred and fortyhouseholds, designed to compare pen-and-paper interviewing(PAPI) to computer-assisted personal interviewing (CAPI).The authors find that PAPI data contain a large number oferrors, which can be avoided in CAPI. The authors show thaterror counts are not randomly distributed across the sample,but are correlated with household characteristics,potentially introducing sample bias in analysis if dubiousobservations need to be dropped. The authors demonstrate atendency for the mean and spread of total measuredconsumption to be higher on paper compared to CAPI,translating into significantly lower measured poverty,higher measured inequality and higher income elasticityestimates. Investigating further the nature of PAPI’smeasurement error for consumption, the authors fail toreject the hypothesis that it is classical: it attenuatesthe coefficient on consumption when used as explanatoryvariable and the authors find no evidence of bias whenconsumption is used as dependent variable. Finally, CAPI andPAPI are compared in terms of interview length, costs andrespondents’ perceptions.