Comparison of IVR and CATI Surveys in Bangladesh

Comparison of Interactive Voice Response (IVR) and Computer-assisted Telephone Interview (CATI) Mobile Phone Surveys for Non-communicable Disease Risk Factor Surveillance in Bangladesh

This study focuses on mechanisms to adapt the performance of interactive voice response (IVR) and computer assisted telephone interviews (CATI) surveys conducted in low-and middle-income (LMIC) setting (Bangladesh) and evaluates how the two survey modalities (IVR and CATI) affect survey metrics, including response, completion and attrition rates.

Using random digit dialing (RDD) sampling technique, participants will be randomized to one of two arms : 1) IVR or 2) CATI. Participants in the first study arm will receive an IVR survey. Participants in the second study arm will receive a CATI survey. Each questionnaire contains a set of demographic questions and one non-communicable disease (NCD) module (alcohol, or tobacco, or diet, or physical activity, or blood pressure and diabetes). We will examine contact, response, refusal and cooperation rates and demographic representativeness by each study arm.

As defined by American Association for Public Opinion Research, cooperation rate is defined as I/(I+P+R) where I is complete interviews, P is partial interviews, and R is refusals and breakoffs

As defined by American Association for Public Opinion Research, response rate is defined as (I+P)/(I+P+R+eU) where I is complete interviews, P is partial interviews, R is refusals and breakoffs, and eU is the estimated eligible proportion of unknowns

As defined by American Association for Public Opinion Research, contact rate is defined as (I+P+R)/(I+P+R+eU) where I is complete interviews, P is partial interviews, R is refusals and breakoffs, and eU is the estimated eligible proportion of unknowns

As defined by American Association for Public Opinion Research, refusal rate is defined as (R)/(I+P+R+eU) where R is refusals and breakoffs, I is complete interviews, P is partial interviews, and eU is the estimated eligible proportion of unknowns

Inclusion Criteria: Access to a mobile phone Greater or equal to 18 years of age In Bangladesh, conversant in Bengali language Exclusion Criteria: Less than 18 years of age

Pariyo GW, Greenleaf AR, Gibson DG, Ali J, Selig H, Labrique AB, Al Kibria GM, Khan IA, Masanja H, Flora MS, Ahmed S, Hyder AA. Does mobile phone survey method matter? Reliability of computer-assisted telephone interviews and interactive voice response non-communicable diseases risk factor surveys in low and middle income countries. PLoS One. 2019 Apr 10;14(4):e0214450. doi: 10.1371/journal.pone.0214450. eCollection 2019.

Gibson DG, Wosu AC, Pariyo GW, Ahmed S, Ali J, Labrique AB, Khan IA, Rutebemberwa E, Flora MS, Hyder AA. Effect of airtime incentives on response and cooperation rates in non-communicable disease interactive voice response surveys: randomised controlled trials in Bangladesh and Uganda. BMJ Glob Health. 2019 Sep 6;4(5):e001604. doi: 10.1136/bmjgh-2019-001604. eCollection 2019.

Gibson DG, Pereira A, Farrenkopf BA, Labrique AB, Pariyo GW, Hyder AA. Mobile Phone Surveys for Collecting Population-Level Estimates in Low- and Middle-Income Countries: A Literature Review. J Med Internet Res. 2017 May 5;19(5):e139. doi: 10.2196/jmir.7428.

Hyder AA, Wosu AC, Gibson DG, Labrique AB, Ali J, Pariyo GW. Noncommunicable Disease Risk Factors and Mobile Phones: A Proposed Research Agenda. J Med Internet Res. 2017 May 5;19(5):e133. doi: 10.2196/jmir.7246.