HOME | CURRENT | ARCHIVES | FORUM

Research World, Volume 7, 2010
Online Version

Article S7.3

Longitudinal Survey: Issues and Challenges

Seminar Leader: R. K. Sinha
Former Professor and Head, International Institute for Population Sciences, Mumbai, India
rksinha35[at]gmail.com

A survey is a means of “gathering information about the characteristics, actions, or opinions of a large group of people, referred to as a population” (Pinsonneault & Kraemer, 1993, p.2). Surveys are conducted in different contexts, for example, market surveys, opinion surveys, and political polls. Surveys are a common method of data collection in research (Hackett, 1981). There are primarily two forms of survey: (a) cross-sectional survey and (b) longitudinal survey (Rindfleisch, Malter, Ganesan, & Moorman, 2008). Cross-sectional surveys involve observation of the population or a representative subset at a defined time. In contrast, longitudinal surveys involve repeated observations of the same items over a period, often over decades.

The seminar addressed different aspects of the longitudinal survey method, based on the seminar leader’s personal involvement in the National Family Health Surveys (NFHS) in India, which were conducted successively in 1992-1993 (NFHS-1), 1998-1999 (NFHS-2), and 2005-2006 (NFHS-3), with roughly a 5-year gap between successive surveys. These surveys, covering women and small children, were expected to provide essential data to assist in the formulation and implementation of health policies and programmes:
    The survey 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. Each successive round of the NFHS has had two specific goals: (a) to provide essential data on health and family welfare needed by the Ministry of Health and Family Welfare and other agencies for policy and programme purposes, and (b) to provide information on important emerging health and family welfare issues. (National Family Health Survey [NFHS], 2009)
The purpose of any longitudinal survey is to identify patterns and trends over time. This is achieved by comparing selected observations between successive surveys. However, in large-scale surveys such as the NFHS, where new dimensions and variables are added at each successive round, there are several methodological issues and practical challenges involved. Some of these issues and challenges are outlined here.

The first challenge is to ensure that the same respondents are being contacted again in the successive surveys. In NFHS-3, this was dealt with at two levels. The first step was to identify the primary sampling units (i.e., villages). This was achieved using the NFHS-2 sampling list and the village location maps available. In the next step, the target households within a village were identified with the help of local people. In certain cases, “Identification Modules” (a process to match a selected person with a previous survey respondent) were devised to help identify households and respondents correctly.

The survey completion rate is hardly ever 100%. In NFHS-3, the completion rates varied from state to state, from a minimum of 76% to a maximum of 94%. One of the reasons behind this is drop-out. The term refers to cases where a respondent surveyed earlier is not available for the next round of survey. Various reasons were identified in the NFHS surveys, for example, absence for attending family functions, migration due to seasonal work, relocation, and even death. Besides, if the survey team does not visit a respondent in the target sample, it also results in a lower completion rate.

A methodological issue is to establish whether this kind of incomplete data can be used for the purpose of trend estimation. The analytical task is to demonstrate that the incomplete data still refer to the same target sample.

Several practical challenges can be anticipated in carrying out such surveys. In this case, there were delays at the very inception resulting from inaccurate estimation of the data collection expenses, leading to delays in securing funding. There were failures in retaining the NFHS-2 completed questionnaires and in recording respondent identification information. The latter hampered the process of identifying respondents correctly in the follow-up survey.

Another concern can be whether it was right to return to respondents if they had not given consent during the earlier survey. In the present case, practical guidelines had to be framed to overcome such concerns. Having an ethical clearance could also be a prerequisite before one proceeds to carry out a longitudinal survey, which in this case was duly obtained.

References

Hackett, G. (1981). Survey research methods. Personnel and Guidance Journal, 59, 599-604.

National Family Health Survey. (2009). Retrieved July 18, 2010, from http://www.nfhsindia.org/

Pinsonneault, A., & Kraemer, K. L. (1993). Survey research methodology in management information systems: An assessment. Journal of Management Information Systems, 10(2), 75-105.

Rindfleisch, A., Malter, A. J., Ganesan, S., & Moorman, C. (2008). Cross-sectional versus longitudinal survey research: Concepts, findings, and guidelines. Journal of Marketing Research, 45, 261-279.


Reported by Upendra Kumar Maurya; edited by D. P. Dash. [July 18, 2010]


Copyleft The article may be used freely, for a noncommercial purpose, as long as the original source is properly acknowledged.
Xavier Institute of Management, Xavier Square, Bhubaneswar 751013, India
Research World (ISSN 0974-2379) http://www1.ximb.ac.in/RW.nsf/pages/Home