A Comparison of Capi and Papi through a Randomized Field Experiment

Bibliographic Data

ID19902201
AuthorsBet Caeyers (0000-0003-3654-6462), Neil Chalmers, Joachim De Weerdt (0000-0002-5572-3332)
Year2010
Publication date2010-11-01
Open AccessNo
TypeBOOK
PublisherWorld Bank, Washington, DC (PUBLISHER)
DOI10.1596/33699
OpenAlexW2287690061
LanguageEN
Citations received6
References cited4

This paper reports on a randomized \n survey experiment among one thousand eight hundred and forty \n households, designed to compare pen-and-paper interviewing \n (PAPI) to computer-assisted personal interviewing (CAPI). \n The authors find that PAPI data contain a large number of \n errors, which can be avoided in CAPI. The authors show that \n error counts are not randomly distributed across the sample, \n but are correlated with household characteristics, \n potentially introducing sample bias in analysis if dubious \n observations need to be dropped. The authors demonstrate a \n tendency for the mean and spread of total measured \n consumption to be higher on paper compared to CAPI, \n translating into significantly lower measured poverty, \n higher measured inequality and higher income elasticity \n estimates. Investigating further the nature of PAPI’s \n measurement error for consumption, the authors fail to \n reject the hypothesis that it is classical: it attenuates \n the coefficient on consumption when used as explanatory \n variable and the authors find no evidence of bias when \n consumption is used as dependent variable. Finally, CAPI and \n PAPI are compared in terms of interview length, costs and \n respondents’ perceptions

Field (mathematics) · Gender, Labor, and Family Dynamics · Housing Market and Economics · Mathematics · Urban, Neighborhood, and Segregation Studies

Tools

Open DOIOpen Access