Skip to Main Navigation

A Comparison of CAPI and PAPI through a Randomized Field Experiment (English)

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


  • Author

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

  • Document Date


  • Document Type

    Working Paper

  • Report Number


  • Volume No


  • Total Volume(s)


  • Country


  • Region

    The World Region,

  • Disclosure Date


  • Disclosure Status


  • Doc Name

    A Comparison of CAPI and PAPI through a Randomized Field Experiment

  • Keywords

    consumption; Computer-Assisted Personal Interviewing; nature of measurement error; total consumption; Poverty & Inequality; household characteristic; household size; missing value; sample size reduction; unit of measurement; interaction effect; acres of land; types of consumption; food energy intake; years of schooling; lack of evidence; information technology revolution; use of time; standard error; fixed effect; household head; income elasticity; consumption datum; data quality; consumption pattern; explanatory variable; poverty headcount; interview time; standard deviation; home production; sample bias; household survey; survey data; marital status; female head; data tables; survey period; regression analysis; low education; measured inequality; education level; conversion rate; regression coefficient; poor household; measured poverty; dependency ratio; consumption aggregate; smaller households; in school; touch screen; farm household; data loss; variable list; attitudinal factor; warning messages; farming household; start school; fertility rate; landless household; data transfer; robustness analysis; learning effect; transport data; data processing; self-help group; married man; field work; data capture; household asset; per household; natural experiment; missing observation; annual consumption; several months; cooking oil; calorific value; upper bind; enumeration area; empirical evidence; consumption measure; unit price; small island; average consumption; mean differences; look-up table; survey methods; estimation equation; summary statistic; survey methodology; research agenda; school expenditure; children of ages; education component; measured consumption; productive asset; food recall; independent variable; instrumental variable; market place; negative effect; formal schooling; positive correlation; negative coefficient; modern technology; measuring consumption; household questionnaire; consumption module; household consumption



Official version of document (may contain signatures, etc)

  • Official PDF
  • TXT*
  • Total Downloads** :
  • Download Stats
  • *The text version is uncorrected OCR text and is included solely to benefit users with slow connectivity.


Caeyers,Bet Chalmers,Neil De Weerdt,Joachim

A Comparison of CAPI and PAPI through a Randomized Field Experiment (English). Washington, D.C. : World Bank Group.