Notes on Methods

Mind the Gap?: Navigating the Studies in Indian Politics

6(1) 140–145
Quantitative and the Qualitative © 2018 Lokniti, Centre for the
Study of Developing Societies
in Survey Research SAGE Publications
sagepub.in/home.nav
DOI: 10.1177/2321023018762827
http://journals.sagepub.com/home/inp

Ankur Datta1
Divya Vaid2

One of the persistent concerns and conflicts faced by social scientists who study different aspects of
Indian society and politics pertains to a methodological divide. This divide is constituted by a separation
of methods into the quantitative and qualitative broadly. While these two categories cover a wide range
of methods and techniques of research and data collection, they have also been treated as representing
two polarities of the social sciences, often seemingly opposed. This note will explore the division
between the quantitative and qualitative.3 While we feel that any divide in method and theory is counter-
productive to the larger project of studying political life, we want to try and see how this separation can
be bridged without necessarily forcing cohabitation. By looking at the survey as a method, this note is an
initial attempt to engage in a dialogue between quantitative- and qualitative-oriented scholars and to
perhaps consider how the two groups are far closer than they think.

The Survey
Look Closely or examine (someone or something)
Examine and record the area and features of (an area of land) so as to construct a map or a plan
Investigate the opinions or experience of (a group of people) by asking them questions
—Survey, Oxford English Dictionary

The survey is a tool that is ubiquitous with any act of gathering data on collectives of human beings,
whether one is interested to know of the composition of a group in a given area, their place in networks
of relatedness, their attitude to politicians or brands of soft drinks. As the definition from the Oxford
English dictionary indicates, ‘to survey’ is essential to what any social scientist seeks to achieve in their
work. In its current form as popularly understood, the survey represents one of the most well-known

1
Department of Sociology, South Asian University, Akbar Bhavan, Chanakyapuri, New Delhi, India.
Centre for the Study of Social Systems, Jawaharlal Nehru University, New Delhi, India.
2

3
  We do not cover the much treaded ground of the advantages and limitations of each method (refer Babbie, 2007; Creswell, 2009).
Note: This section is coordinated by Divya Vaid. E-mail: [email protected]

Corresponding author:
Ankur Datta, Department of Sociology, South Asian University, Akbar Bhavan, Chanakyapuri, New Delhi, India.
E-mail: [email protected]
Datta and Vaid 141

forms of large-scale data mining from a given population especially in the period after World War II
(Schuman, 2002). It is also one of the main forms of data collection tied to particular ends, all the way
from market research to enabling government policy.
The forms the survey takes, especially in providing data in its quantitative form in India, can be
connected to other projects most notably the census which was considered an important tool in generat-
ing objective knowledge about a population (Cohn, 1996). Even before the introduction of the modern
census in colonial India, colonial surveyors’ produced documents of objective knowledge but which
took interesting forms, drawing on both numerical data, geographical mapping and descriptive sections
that can be regarded as a precursor to ethnographic studies.4
Yet, the survey also remains an instrument that is treated with suspicion by social scientists, especially
to try and capture data on beliefs and opinions. Yogendra Yadav locates this suspicion in the critique
made by Indian political scientists who argued that the survey is an instrument that in practice had been
transplanted from the West into Indian settings and caught in a positivist framework (2008, pp. 8–11).
While Yadav locates this problem in the division between empiricists on one side and theory builders on
the other, there appears to be a divide even among empiricists of surveyors on one side versus qualitative
researchers on the other side. Part of the problem comes from a perceived sense of scale with which
methods are treated. Surveys and other quantitative methods which appear to generate data that speaks
for large populations enable an analysis of trends and patterns, whereas qualitative methods such as
interviewing, life histories or case studies enable the accessing of detail, texture and nuance of the lives
of people we are studying. Ideally, both are essential to good analysis in the social sciences.

An Ethnographic Detour
The tense relationship between quantitative and qualitative research has an odd history, especially if we
take the case of Social Anthropology and especially researchers practising ethnography. In the present,
the divide seems significant when ethnographers are suspicious of anything quantifiable. For instance,
Benei, whose work looks at the connections between nationalism and schooling in India, argues for the
utility of life histories over quantifiable data ‘as if “collective” bodies were but the total sum of individu-
als’ (2010, p. 199). Contemporary practitioners of ethnography emphasize that their method enables a
nuanced engagement with a given population under study in the context of everyday life, which quantifi-
able methods cannot enable. If we look at recent publications by ethnographers who study political
life, there appears to be an engagement with different disciplines and methods often outside the social
sciences, such as Roxanne Varzi’s (2006) remarkable study of youth and martyrdom in Iran, which
makes use of poetry and fictional approaches.
The ethnographic monograph, and ethnography as a method, has changed over time in styles of
doing fieldwork, collecting data and especially in writing and presentation.5 However, what seems to be

4
  Accounts by colonial officials such as W. W. Hunter come to mind whose Statistical Accounts of Bengal are remarkable for
extensive prose descriptions of districts and their inhabitants.
5
  Compare Varzi’s monograph to Edmund Leach’s Political Systems of Highland Burma (1954), which is regarded as a classic in
political anthropology. In any case within Anthropology there have been occasionally critiques of the ethnographic enterprise, first
in the 1980s with the Writing Culture group which treated ethnographies as texts that can claim truth partially (Clifford & Marcus,
1986) to a recent debate started by Tim Ingold (2014) who has criticized the over dependence of anthropology as a discipline
defined by ethnography and simultaneously preventing any discussion of the politics of doing ethnographic research especially in
relations between ethnographers and respondents.
142 Studies in Indian Politics 6(1)

forgotten is how the survey was often an important part of the ethnographer’s tool kit in the field.
Most ethnographers may locate themselves in a particular location or set of locations and work with a
given identified population, but in practice generate much of their data in interaction with a close set of
respondents/informants. Hence, the household survey was an important part of fieldwork especially as it
gave the ethnographer the accurate picture of the composition of a given population, to the extent that
many students of social anthropology for a few decades drew on the Rhodes-Livingstone Institution
Census card. If one looked at the early ethnological work of the late nineteenth century, regardless of
how problematic they were, the survey and other means of securing quantifiable data were part of
ethnography.6 Even anthropologists writing in the middle of the twentieth century would insist that quan-
titative methods were ‘aids to description’ and facilitated establishing of correlation between variables
(Mitchell, 1978, p. 20).
The problem however remains in a perceived hierarchy of methods in the ethnographic enterprise.
As Elizabeth Colson had written in reflections on fieldwork among dam displaced groups in Africa in
the 1950s, surveys provided important quantifiable material to back and support qualitative data (1978,
p. 4). Quantitative data in other words supported description but otherwise remained a supporting cast
member, whereas description remains centre stage. Further, surveys here are essentially about obtaining
basic background data about kinship, property and networks of respondents under study. In some cases,
surveys are often a useful pretext to meet more respondents. Hence, the traditional ethnographic survey
is not about seeking information about opinions and attitudes. Perhaps, this is why some qualitative
researchers continue to underestimate the survey, and in turn other quantitative methods, even when they
are amenable to the idea of survey work. On the other hand, if we look at reports as produced by the
United Nations or the World Bank which often privilege quantitative measures, qualitative research such
as case studies are often featured as supporting the data, ostensibly giving the numerical a human face.

Back to Surveys
It will be wrong to presume that the divide between the quantitative and the qualitative remains insur-
mountable. One can look at the work of social anthropologists such as Veena Das’s (2015) work on
health seeking behaviour among the poor in New Delhi or Peter Loizos’s (2008) monograph on Cypriot
refugees which draws on longitudinal data. Das addresses the health-seeking behaviour and concerns
and approaches to suffering of the urban poor by drawing on a mix of participant observation, case stud-
ies, working with teams of interviewers and surveyors. Loizos’s work is especially interesting as it builds
on early work on refugees in Cypress (Loizos, 1981). His monograph is quite rare as it builds on several
decades of research on the same community and makes use of ethnography, case studies and surveys to
collect detailed and accurate longitudinal data on health and well-being. It appears that health studies
have overcome the divide of techniques. As Carla Pezzia (2016) suggests in her work on alcoholism in
Mexico, both quantitative and qualitative methods are needed and yet both sets of methods may often
need to be revised in the manner of their implementation in data collection in the field.
The question then remains with regard to the survey itself. Why is the survey that seeks to elicit opin-
ions and attitudes subject to critique? Perhaps, one reason could be the discomfort sociologists and

6
  The work of Imperial officers and ethnologists such as Risley were surveys in their fullest form comprising anthropmetric
measurements, population statistics, inventories of practices and so on. While some of this work has been criticized for being
embedded in colonial and racial politics (Guha, 1999), they were marked by a full sense of survey being a body of knowledge on
a given population.
Datta and Vaid 143

political scientists who employ qualitative techniques feel is with regard to data itself. Many scholars by
and large rely on published secondary data whenever they feel the need to supplement their material with
macro-data. Surveys then, if at all conducted, become a stop gap in data collection to satisfy the needs of
faculty and grant committees, which increasingly look for ‘mixed-method’ approaches in applications.
In fact, we observe that ‘mixed-method’ is greatly misunderstood, where the survey is thrown in as an
afterthought in a basket of methods without considering how it can complement or supplement other
methods. In the process, we miss out that even qualitative methods may generate numerical data and
hence a stop-gap survey might do more damage to the research project. For example, a detailed descrip-
tion of activities at a voting booth will include noting the number of staff present and the number of
people coming to cast a vote even if the researcher’s final aim is to present a prose-based description
of political activity. Hence, a survey would really only be useful here if the researcher wants to go
beyond these basic observations and instead wants to elicit opinions and attitudes that can perhaps be
generalizable, adding depth to their study. A survey that explores topics such as political opinion is
ultimately addressing choice and preference which are subjective phenomena as opposed to the basic
household survey. Finally, researchers seem to miss another important dimension of surveys, that is, that
data gathered by surveys can help in supporting and questioning the results from qualitative research.
It must be admitted that like any other method the survey needs to be modified over time. As Nooruddin
argues, the introduction of survey research in political science was an important moment and continues
to form an important part of the methodological arsenal (2014, p. 105). What he suggests is a need to
improve the survey as a method in different ways by drawing on experimental methods as part of a larger
project of methodological innovation (Nooruddin, 2014, p. 108). The call for methodological innovation
or for developing research that involves a mix of methods is one that is made fairly often. Even if
contemporary qualitative researchers use surveys, grudgingly or with enthusiasm, there is the unforeseen
advantage that can be afforded. For example, Datta’s (2017) ethnography of a displaced community in
Jammu and Kashmir involved observation and description gradually built with a core group of respond-
ents. As part of data collection, a small household survey was conducted which first confirmed the size
of households in the camp and allowed for mapping camp residents in the context of kinship networks.
However, the survey also allowed interactions with people outside the core set of respondents and
interlocutors thereby generating further narratives and also allowing the researcher to develop new
relationships. During the administering of surveys, Datta observed a certain reticence of respondents
when they would see the survey sheets; the respondents pointed out that they were used to being
surveyed constantly by students, researchers, NGOs and government staff. This realization speaks to the
larger politics and ethics of doing research and the relationship between researchers as data collectors
and respondents as sources of data to be mined. Nevertheless, the survey in practice regardless of its size
and form offers unexpected possibilities.

Towards Dialogue
Eventually any discussion between practitioners of particular methods requires a common ground
rather than forcing collaboration. The three points of commonality between quantitative and qualitative
methods comprise the following:

1. Sampling: All social scientists have to identify with a sample, whether the sample can consist of
a thousand respondents from a metropolitan city to research work in a small village. It will be
interesting to have a conversation between the two sides of the divide of their views on sampling
144 Studies in Indian Politics 6(1)

especially as there is an assumption of scale at play when survey research in political science is
regarded to approach large-scale data as opposed to data generated by qualitative methods.
The question of scale is further complicated when ethnographers incorporate survey techniques
in their work. Or, indeed when survey researchers simply sample a few respondents, rather
than in large numbers. Further, the randomness and non-randomness of sampling too can have
overlaps in qualitative and quantitative approaches.
2. The use of semi-structured interviewing: Most researchers who employ survey research will
attest to the use of semi-structured questions in questionnaires to generate data. Most researchers
using qualitative methods also use semi-structured interview techniques, which may often be
shaped by the setting of the interview and the context. For example, if a researcher is interviewing
prominent members of a political party of government functionaries, the presence of a question-
naire sheet is an indication of structure to the encounter, especially if only one encounter is
possible with a respondent. This eases the collection of data and clearly signifies intent of the
researcher.
3. The possibility of manipulation: All data can be manipulated. While quantitative information can
be tweaked or suppressed, qualitative data also has the potential of being used selectively.
An ethnographer can present certain vignettes from his/her field notes while withholding others.

There can be many other common points of concern. It should also be clear that researchers should not
necessarily be forced into a mixed methods framework and that good research and scholarship is possi-
ble if a researcher remains on either side of the fence. What ultimately matters is the question or topic
that the researcher is seeking to answer or explore. If we return to some of the examples we have
mentioned in this discussion, it is obvious that the demands of the research question influenced the methods
employed (Lieberson, 1988). This does not mean that the use of a quantitative method renders, for exam-
ple, Veena Das’s work to be ‘un-anthropological’ for using mixed-methods and for not relying solely on
ethnography. What makes Das’s work anthropological is that it is driven by a question that seeks to raise
a ‘sustained and disciplined inquiry into the conditions and potentials of human life’ (Ingold, 2011, p. 1).
Hence, it is the specific approach to a question, rather than a strict adherence to a particular method that
is critical. By that extension, the same can be said of any good piece of social science research, whether
it is anthropology, sociology or political science.

References
Babbie, E. (2007). The practice of social research (12th ed.). Belmont, CA: Wadsworth.
Benei, V. (2010). To fairly tell: Social mobility, life histories, and the anthropologist. Compare: A Journal of
Comparative and International Education, 40(2), 199–212.
Clifford, J., & Marcus G. (Eds). (1986). Writing culture: The poetics and politics of ethnography. Berkeley, CA:
University of California Press.
Cohn, B. (1996). Colonialism and its forms of knowledge. New Delhi: Oxford University Press.
Colson, E. (1978). The intensive study of small sample communities. In A. L. Epstein (Ed.), The craft of social
anthropology (pp. 3–15). Delhi: Hindustan Publishing Corp.
Creswell. J. W. (2009). Research design: Qualitative, quantitative, and mixed methods approaches. Thousand Oaks,
CA: SAGE Publications.
Das, V. (2015). Affliction: Health, disease, poverty. New Delhi: Orient Blackswan.
Datta, A. (2017). On uncertain ground: Displaced Kashmiri pandits in Jammu and Kashmir. New Delhi: Oxford
University Press.
Guha, R. (1999). Savaging the civilised: Verrier Elwin, his tribals and India. New Delhi: Oxford University Press.
Ingold, T. (2011). Being alive: Essays on movement, knowledge and description. London: Routledge.
Datta and Vaid 145

Ingold, T. (2014). That’s enough about Ethnography! Hau Journal of Ethnographic Theory, 4(1), 383–395.
Leach, E. (1954). Political systems of Highland Burma: A study of Kachin Social Structure. Cambridge: Harvard
University Press.
Lieberson, S. (1988). Asking too much, expecting too little. Sociological Perspectives, 31(4), 379–397.
Loizos, P. (1981). The heart grown bitter: A chronicle of Cypriot War Refugees. Cambridge: Cambridge University
Press.
———. (2008). Iron in the Soul: Displacement, Livelihood and Health in Cypress. London: Berghahn.
Mitchell, C. J. (1978). On quantification in social anthropology. In A. L. Epstein (Ed.), The craft of social anthro-
pology (pp. 17–45). Delhi: Hindustan Publishing Corp.
Nooruddin, I. (2014). Making surveys work better: Experiments in public opinion research. Studies in Indian
Politics, 2(1), 105–108.
Pezzia, C. (2016). When your method fails: Stumbling upon new questions. MAT: Medicine Anthropology Theory,
special section on dissertating. Retrieved 16 March 2018, from http://www.medanthrotheory.org/read/5841/
when-your-method-fails
Schuman, H. (2002). Sense and nonsense about surveys. Contexts, 1(2), 40–47.
Varzi, R. (2006). Warring souls: Youth media and martyrdom in Post-Revolutionary Iran. Durham, NC: Duke
University Press.
Yadav, Y. (2008). Whither Survey Research? Reflections on the state of survey research on politics in most of the
World. New Delhi: Malcolm Adeseshiah Memorial lecture.