24/10/2019 The Hindu Editorials Notes – Mains Sure Shot

Question – Analyse the reasons behind poor quality of data collected by various agencies and share the way ahead.(250 words)

Context – 25 years of National Family Health Survey.

What is National Family Health Survey (NFHS)?

• As the name suggests, NFHS is a large-scale, multi-round survey of households throughout India.
• Three rounds of the survey have been conducted since the first survey in 1992-93.
• The survey is vital to policy makers because it provides state and national information for India on fertility, infant and child mortality, the practice of family planning, maternal and child health, reproductive health, nutrition, anaemia, utilization and quality of health and family planning services.
• The survey has two specific goals – a) to provide essential data on health and family welfare and b) to provide information on important emerging health and family welfare issues.
•  The Ministry of Health and Family Welfare (MOHFW) is the nodal agency responsible for coordination and technical guidance.
• But there are certain concerns in the NFHS about India’s existing data infrastructure and data quality.

The concern:

• There are two main concerns. One is about the data infrastructure i.e. do we have a proper structure that encourages the collection of high quality data? And second is the concern about the quality of data being collected.
• In the recent meeting of demographers (those who collect data) from around the world to mark 25 years of NFHS, it was found that the quality of data collected i.e. data quality is deteriorating.

For example:

• Between 2005-06 and 2015-16, the total fertility rate (TFR) declined from 2.68 to 2.18 births. This implies that the use of contraceptives between the same period must have increased.
• But the data about contraceptive use in the same period showed a decline too, from 56.3% to 53.5%.
• Using different approaches, Prof. Tusi (at Johns Hopkin University) and Dr. Pramanik (Deputy Director at the National Council of Applied Economic Research) reach a conclusion that this is because of the declining quality of contraceptive use data collected in NFHS-4.
• Usually the discussions about data come up when data about topics like GDP and poverty are released but there is little discussion about the quality of the data released as such.
• So we need to look at the overall challenges facing our data infrastructure in a constructive manner.

How?

• First, by setting realistic goals and use creative strategies to improve the data infrastructure (i.e. the system that encourages the collection of high quality data not just numbers).
• We are talking of realistic goals because if compared with NFHS-3, then the sample size to collect data at district level increased from 1lakh households to over 6lakh households in NFHS-4. (this means that earlier 1lakh households in a district were approached and their data was collected and after an average a result (data) was prepared, that represented data at the district level). Here it is to be noted that when sample size i.e. the number of households increases so much then the load on the data collectors also increases and they tend to hurry up and in turn the quality of data collected is affected. Today there are other technical means available for collecting data from diverse sources not only households and increasing their numbers.
• Second is ensuring quality by adapting to changing institutional and technological environment for data collection. Also most of the data collection are being passed on from regular employees to contract investigators and for-profit data collectors. Unlike the veteran investigators they are not so efficient. So they need to be given proper training.
• Third, there is a need to collect data using technologically enabled procedures such as random voice recording of interviews, judicious back checks, and evaluation of agency and interviewer performance on parameters such as skipping sections, inconsistent data and consistent misreporting are needed to ensure quality.
• Fourth also state population research centres may be involved in the process and also in the process of quality monitoring.
• Fifth, there is a need to establish research units exclusively focused on data collection and research design. Because if there is proper research on data collection and practices then only we will be able to know things like whether men or women are better responders on data on household collection expenditure. Or the extent there is a discrepancy on reporting on employment data between a direct response from men and women vis-a-vis a proxy response from household head. Or how does the presence of other people when the data is being collected affect the responses?

Why is data quality or the quality of data collected important?

• Data collected guides the policies affecting millions of Indians and hence must be cautiously collected.
• Any wrong policy based on incorrect data can lead to more harm than good.
• Also data is the new oil and that must be of high quality if it has to have any market value. International investors and companies also base their decisions on data and analysis and wrong data can give wrong signals that can affect a country in both short and long term.

conclusion/Way ahead:

• While research on data collection methods in India has not improve, research methodology in other developed countries has increased phenomenally.
• For example, telephone surveys via random digit dialing or selection of respondents using voter lists are increasingly emerging as low-cost ways of collecting data.
• However without proper research we can know little about representativeness of such samples. Are men or women more likely to respond to telephonic surveys? Are migrants from other states well represented on voter list?
• So unless we pay systematic attention to data infrastructure i.e. the existing system of data collection, we are likely to have policies and debates based on poor quality data.