Quality of Data in NFHS-4 Compared to Earlier Rounds

As the quality of data from the fourth round of the National Family Health Survey is likely to be affected both by the overstretching of the number of questions administered as well as the increased involvement of commercial agencies for data collection, rethinking both these aspects of survey management is the need of the hour.

The authors wish to thank K S James for initiating and assisting on this study.

In two articles in 2004 and 2008, which S Irudaya Rajan and K S James had attributed the extended scope, overstretched fields of enquiry, and the increasing presence and role of consultancy organisations in gathering the field-level information, to be among the reasons behind susceptible quality of data on select variables in the second and third rounds of the National Family Health Survey (NFHS). It is a common notion that with an increase in the size of the questionnaire, as well as the sensitivity of the questions asked, such as on domestic violence and sexual behaviour, the quality of the data collected is likely to take a beating, though it is difficult to establish this empirically (Bogen 1996). This is particularly true in the case of many sensitive questions that were asked in the NFHS-3 such as on domestic violence and premarital sexual behaviour. It is difficult to check the reliability and validity of the responses received from these questions, though interesting and critical articles can be and are written on the above themes.

While the NFHS-4, conducted during 2015–16, potentially provided a comparative perspective on the earlier rounds, from the organisational points of view there are some significant slippages. First, is in terms of the selection of suitable data-collecting organisations, and second, due to the overload of questions with the increasing number of donor agencies, for each donor agency would add questions in line with its agenda. In this article, we examine the quality issues of data available from NFHS-4, in the light of these aspects of data collection. A broader objective of this exercise is to help design and organise the NFHS as a credible source of valuable data on population, health and family planning parameters in India at the national, state and even district levels, over time. However, before we undertake the above analysis, a brief statement of the objectives, study design, sample size and variables on which data were collected in different rounds of the NFHS seems to be in order.