2 Components of Social Research: Theory, Logic, and Method

N. Jayaram

Contents

1.      Objective

2.      Introduction

3.      Learning Outcome

4.      Definition of Research

5.      Research Problem

Self-Check Exercise 1

6.      The Three Components of Social Research

Component 1: Theory

7.      Meaning and Types of Theory

8.      Theory as Embodying Assumptions

9.      Paradigm and Paradigm Shift

10.  Assumptions and Hypothesis

11.  Ontology and Epistemology

11.1 Ontology: Objectivism and Constructivism

11.2 Epistemology: Positivism and Interpretivism

Self-check Exercise- 2

Component 2: Logic

12.  Deductive Reasoning

13.  Inductive Reasoning

14.  Abductive Reasoning

Component 3: Method

15.  Methods and Data

16.  Triangulation

Self-Check Exercise 3

17.  Summary

Notes

1.        Objective

In this module you will learn about the three components of social research, namely, theory, logic, and method. Each of these components is discussed separately and the relationship of each component with the other two is analysed. This module explains to you how the methodology of social research is more than mere application of methods.

2.        Introduction

Scientific knowledge is not only distinguished from all other forms of knowledge, but it is also the most privileged form of knowledge. Its claim for this privileged status is based on the fact that the conclusions of science, unlike common-sense beliefs and religious dogmas, are the outcome of research. Research in science, as also in social sciences, including sociology, involves the application of institutionalised principles, procedures, and techniques of acquiring new knowledge as also for refining existing knowledge. Popularly, this is called the scientific method. The rigour with which a discipline practises scientific method depends on (a) the nature of the subject matter that it studies, and (b) the methods and techniques that are used for studying it.

In this context, a distinction is often made between natural sciences and social sciences. The natural sciences, for various reasons, have made rapid strides and great advances in generating reliable knowledge about several aspects of the physical and the organic world we inhabit. The social sciences, again for various reasons, have not been as successful in generating reliable knowledge about the various aspects of the social world we inhabit. Not surprisingly, therefore, the practitioners of natural sciences claim superior status for the knowledge they generate vis-à-vis that generated by their counterparts in the social sciences; some even deny scientific status to social sciences. Be that as it may, what is common to both natural sciences and social sciences is their acceptance of research as the pathway to reliable knowledge.

3.        Learning Outcome

This module would acquaint you with the three major components of social research, namely, theory, logic, and method. As these components are intricately linked to each other, you would also learn about their relationship. By doing so, this module would make your understanding of theory, logic and methods and techniques of collecting data as also analysing them more concrete and deep.

4.      Definition of Research

There are many definitions of research. Tracing the origin of the word ‘research’ to 16th century France, the Concise Oxford English Dictionary defines research as ‘the systematic investigation into and study of materials and sources in order to establish facts and reach new conclusions’ (Soanes and Stevenson 2004: 1222). This noun usage of the term emphasises the nature (‘systematic’) and objective (‘to establish facts and reach new conclusions’) of the activity called research. This terse definition is, no doubt, too general and, as a consequence, many an activity that is systematic is described as ‘research’. If we shift our attention to what natural scientists and social scientists do when they engage in an activity called research, we arrive at a better understanding of that term.

In simple terms, research may be defined as an activity that consists of two interrelated activities, namely,

(i)   asking questions and (ii) attempting to answer them. The attempt at answering questions does not mean that the researcher will eventually find a satisfactory answer. It is likely that the researcher may end up with more questions about the phenomenon than he/she had set out to answer. This must not be viewed as a negative reflection of the way the research was conducted; rather, it is a valuable contribution to the discipline(s), as questions direct the practitioners of the discipline(s) to hitherto unknown aspects of the phenomenon in question. After all, questioning is the driving force of any science.

5.      Research Problem

Thus, asking meaningful questions (that is, formulation of the research problem) is the fundamental step which determines how the researcher goes about studying a phenomenon and the eventual usefulness of her/his findings. Intellectual curiosity (that is, the desire to gain knowledge for its own sake), practicality (that is, the urgency to solve a problem), and intrinsic orderliness (that is, observation of a regularity of pattern) of a phenomenon may motivate a researcher to ask questions. Personal experience, state of knowledge on the subject of research, social premiums (that is, ‘hotness’ of a topic and availability of money for researching it), everyday life and personal values of the researcher may also play an important role in the selection of the research problem (see Neuman 1994: 110).

‘A problem well put’, as the aphorism goes, ‘is half solved’. However, there is no formal recipe for the formulation of a research problem. Systematic immersion in the subject matter and training in the art and craft of interrogating the existing stock of knowledge about a phenomenon will help the researcher in this (see Greer 1977).

What are the characteristics of ‘good’ researcher questions? According to Keith F. Punch, ‘good’ research questions are:

Clear: They can be easily understood, and are unambiguous.

Specific: Their concepts are at a specific enough level to connect to data indicators.

Answerable: We can see what data are required to answer them, and how the data will be obtained.

Interconnected: They are related to each other in some meaningful way, rather than being unconnected.

Substantively relevant: They are interesting and worthwhile questions for the investment of

research efforts (1996: 49).

Self Check Exercise 1:

1.   What is a scientific method?

Scientific method involves the application of institutionalised principles, procedures, and techniques of acquiring new knowledge as also for refining existing knowledge. Unlike common-sense beliefs and religious dogmas, scientific method relies on clearly defined theory, logic and methods to study a problem.

2.   What is common among natural and social sciences?

Despite the fact that the practitioners of natural sciences claim superior status for the knowledge they generate vis-à-vis that generated by their counterparts in the social sciences, both accept research as the pathway to reliable knowledge. In other words, both are involved in asking questions and answering them. Questioning is the driving force of any science.

3.   How does a researcher formulate her/his research problem?

There is no formal recipe for the formulation of a research problem. Systematic immersion in the subject matter and training in the art and craft of interrogating the existing stock of knowledge about a phenomenon will help the researcher in this. On the whole, Intellectual curiosity, practicality, and intrinsic orderliness of a phenomenon may motivate a researcher to ask questions. Personal experience, state of knowledge on the subject of research, social premiums, everyday life and personal values of the researcher may also play an important role in the selection of the research problem.

6.      The Three Components of Social Research

Be it in the natural sciences or in the social sciences, the questions we ask about the world do not suddenly appear out of the blue. Rather, our questions only arise within the context of general interpretations of what the world is or, what John A. Hughes (1976) calls, our ‘meaning systems’. Different models of reality will lead to different propositions about what reality is, and so different ways of establishing what can be accepted as real, different ways of justifying the data relevant to reality, and different strategies for collecting such data. These aspects of research and understanding are built into all meaning systems. They can be broadly discussed as three components of social research, namely, theory, logic, and methods.

In actual research, these three components of social research are interrelated; each one of them is implicated in the other. The nature of their relationship is reciprocal, as shown in Figure 1. Theory has a bearing on logic and method; logic, on theory and method; method, on theory and logic; and each one is, in turn, influenced by the other two. Thus, they can be discussed in isolation of the other two only for analytical convenience.

Figure 1: Theory, Logic, and Method

Logic                               Method

Such an isolated discussion of any one component may clarify its dimensions as well as nuances. However, they are best discussed in a sequence, and the sequence that is appropriate is theory, logic, and method. This is because theory is implicated in and independently influences both logic and method more distinctly than the other way round.

Component 1: Theory

7.        Meaning and Types of Theory

The term ‘theory’ is one of the most amorphous, misused, and misleading terms in the vocabulary of social scientists. It has a range of meanings: it is used broadly to encompass all thought and narrowly to refer to a single thought. It has been applied to thinking process per se or only to its results and conclusions. It may vary from complete conjecture to solid confirmation; from an unarticulated impression to a precisely defined prediction. Percy S. Cohen compares the term ‘theory’ to a ‘blank cheque’: ‘its potential value depends on the user and his use of it’ (1968: 1).

According to Robert K. Merton, the term theory refers to ‘logically interconnected sets of propositions from which empirical uniformities can be derived’ (1968: 39). Given the imponderable complexity of socio-cultural reality, sociologists, as most other social scientists, have developed innumerable theories. Sociology, accordingly, has been described as a multiple paradigm science’ (Ritzer 1975) or being in a ‘multi-paradigmatic stage’ (Bryant 1976: 15). Of course, not all theories in sociology have the same analytical status; they differ in terms of such attributes as generality, precision, testability, elegance, predictive power, and the nature of their postulates. In terms of their generality, scholars like Merton have tried to place sociological theories on a continuum: from ‘general theories of social systems’ at one end to ‘detailed orderly descriptions of particulars that are not generalized at all’, ‘middle-range theory’ being intermediate to these (Merton 1968: 39; emphasis added).

Sociological theories can be and have been classified in many ways. For the discussion on the nature of theory as a component of research and its relation to other two components, namely, logic and method, it is appropriate to distinguish between ‘substantive theories’ and ‘general theoretical orientations’. Substantive theories deal with particular aspects or spheres of social phenomena. Thus, we come across theories of crime and delinquency, child development, fertility transition, migration, poverty, stratification, suicide, and so on. Overarching these theories are the general theoretical orientations which define the methodological perspective or approach to social phenomena. These are often described as paradigms of sociological inquiry. It is in the latter sense of theoretical orientations briefing sociological inquiry that the relationship between theory, on the one hand, and logic and method on the other are delineated in this module.

8.      Theory as Embodying Assumptions

Underlying all research, and therefore, all knowledge, are assumptions. Assumptions are propositions that are accepted as true without proof. They are the basis on which we start the process of acquisition of knowledge about anything; they are also the basis on which we accept or reject knowledge about a given thing. Scientifically speaking, assumptions are by themselves neither true nor false. Propositions in the form of assumptions cannot be tested for their truth content. They are accepted or rejected based on their capacity to explain observed things or reality.

If the assumptions have to do with knowledge in which faith is primary, then such assumptions will be believed to be true, notwithstanding any evidence to the contrary. The ‘theory of creation’ in founded religions, as a body of knowledge, which rests on belief, is an apt illustration of this. On the other hand, if the assumptions have to do with knowledge in which doubt is primary, then such assumptions will be discarded if they repeatedly fail to explain a given phenomenon, or an alternative set of assumptions explain that phenomenon more convincingly. The ‘theory of evolution in biology’, as a body of knowledge, which rests on doubt, is an apt illustration of this.

Even in science, often no set of assumptions convincingly explains all instances of a given phenomenon. In such a situation, plurality of assumptions may persist for a long time, even if some of them contradict each other. For example, in physics, there are plurality of assumptions to explain the phenomenon of light; hence plurality of theories of light. In view of the nature and complexity of social reality, in sociology and other social sciences, plurality of assumptions is the norm and well accepted. It is for this reason, sociology, as noted earlier, is often described as a multi-paradigmatic discipline.

9.      Paradigm and Paradigm Shift

In its sociological use, the term paradigm derives from the work of the well-known philosopher of science, Thomas Samuel Kuhn (1922–1996). Writing on the nature of scientific knowledge in his book The Structure of Scientific Revolutions (1970), Kuhn argued that scientists work within paradigms. Paradigms are general ways of seeing the world (assumptions) which point to the nature of scientific work to be done and the kinds of theory that are acceptable. These paradigms provide what Kuhn calls ‘normal science’ (ibid.: 10), the kind of science pursued routinely. Over a time, normal science produces anomalies which cannot be resolved within the paradigm. Then, the given paradigm is replaced by a new one, leading to a new period of normal science, often expressed as ‘paradigm shift’. In sociology, as with many other scientific terms, the term paradigm is used vaguely. It denotes, ‘schools of sociological work or meta-theories, each of which is relatively self-contained, with its own methods and theories’ (Abercrombie et al.2000: 253).

10.  Assumptions and Hypothesis

Assumptions are often confused with hypotheses (singular, hypothesis). Assumptions, to reiterate, are never tested empirically; they are, therefore, never accepted or rejected based on empirical evidence. Hypothesis, on the other hand, is a tentative answer to the research question. In other words, it is a proposition or set of propositions put forward for empirical testing; it is accepted or rejected based on empirical test. Often, hypotheses are derived from assumptions. Thus, the strength of an assumption lies in its capacity (a) to generate testable hypotheses and (b) to explain observations made or results arrived at.

A mere rejection of one hypothesis does not automatically result in the rejection of the assumption from which it is derived. However, if repeatedly hypotheses derived from an assumption are rejected, or if the assumption does not adequately explain the observations made or results arrived at, then it is bound to be discarded.

11.  Ontology and Epistemology

In doing research, we make two sets of assumptions. The first set of assumptions is about the nature of reality that exists, or we believe exists. It is concerned with how we define the subject matter of our discipline or the particular object of our research. These assumptions are called ontological assumptions. The second set of assumptions is about the nature of knowledge about what exists or is believed to exist. It is concerned with the type of knowledge that is possible to acquire about the subject matter of our discipline or the particular object of our research. These assumptions are called epistemological assumptions. Briefly stated, ontological assumptions have to do with reality and the epistemological assumptions have to do with knowledge about that reality. Obviously, these two sets of assumptions are intrinsically related.

11.1       Ontology: Objectivism and Constructivism

In sociology, broadly, there are two contrasting ontological positions: objectivism and constructivism. Objectivism assumes that reality has an existence independently of the human beings living it and the sociologist studying it. That is, reality has an objective existence. Thus, it is possible for us to acquire objective knowledge about it. Sociologist, no doubt, is essentially a subjective creature, with her/his own values and biases. Hence, the emphasis on objectivity in research requires two things: first, a belief that it is possible to acquire such knowledge, and second, we must separate values from facts.

In this context, one is reminded of Émile Durkheim’s (1858–1917) ‘first and most fundamental rule’ of sociological method: ‘Consider social facts as things’ (1966: 14; emphasis original). Durkheim elaborated this rule by highlighting its three corollaries as follows:

i.     All preconceptions must be eradicated.

ii.   The subject matter of every sociological study should comprise a group of phenomena defined in advance by certain common external characteristics, and all phenomena so defined should be included within this group.

iii.     When … the sociologist undertakes the investigation of some order of social facts, he must endeavour to consider them from an aspect that is independent of their individual manifestations (ibid.: 31, 35, 45).

How rigorously can a sociologist adhere to these rules in the study of social reality and what compromises he or she makes is a different issue.

Constructivism, on the other hand, believes that human beings are meaning-making individuals and the transaction of meanings among individuals takes places in the context of everyday life. Thus, sociologists following this line of ontological assumption – called constructivists – emphasise that reality is socially constructed. They are influenced by the overarching theoretical framework called sociological phenomenology. The chief proponent of this was the Austrian sociologist Alfred Schütz (1899–1959), well known for his book The Phenomenology of the Social World. It was refined by his disciples Peter L. Berger (1929–) and Thomas L. Luckmann (1927–), well known for their book The Social Construction of Reality, and Harold Garfinkel (1917–2011), who articulated it in its extreme anti-objectivist form in his book Studies in Ethnomethodology.

Constructivism shifts the attention from the objective social reality ‘existing out there’ to social reality being constructed in everyday life. Entailed in this approach is privileging the people as the generators of social knowledge; they are participants in the study, rather than being mere subjects of the study. Understanding social construction of reality can take place in the context of everyday life of small groups spread over a long period. The type of knowledge generated by constructivists is remarkably different from that generated by objectivists.

11.2    Epistemology: Positivism and Interpretivism

In sociology, broadly, there are two contrasting epistemological positions: positivism and interpretivism. Positivism assumes that science can deal only with observable entities known directly to experience. Based on concrete objective evidence, it assumes, we can arrive at generalisations about social reality. Further, based on such generalisations, we can construct general laws or theories which express relationships about phenomena. Through empirical research, it is assumed, we can show that the phenomena are or are not related in the predicted way. Explanation in positivism consists in showing that the observed relationships are instances of the general laws or regularities.

In positivism, answer to research questions is posited in advance; the relationships about phenomena are predicted and research consists of testing these predictions. These apriori predictions are called hypotheses, and these hypotheses are tested through the collection and analysis of appropriate data. These hypotheses are generally deduced from some general propositions about the phenomenon, that is, a substantive theory. In other words, positivism as an epistemology, in general, implies the primary use of deductive logic (discussed later).

Moreover, positivism is committed to empiricism, the doctrine that assumes that the only source of knowledge is observation and obtaining response to externally induced stimulus (e.g., response obtained through interview schedule or questionnaire). To arrive at generalisations based on observation or responses to stimuli, it uses measurement and numerical analysis.

In brief, in following objectivist ontology and positivist epistemology, a researcher is inevitably committed to quantitative approach as the main plank of study. Furthermore, the objective of arriving at valid and reliable generalisation requires rigorous formulation of the research design. That is, decisions regarding various steps of the study are meticulously taken and adhered to. This makes the design of a positivist (quantitative) study linear and ‘hard’.

Markedly different from positivism, is interpretivism, which is also called as hermeneutics and Verstehen in the literature of sociological theory and research methodology. Interpretivism as an epistemological position is implicated in constructivist ontology. It assumes that, as a social science, sociology can only deal with actions (i.e., subjectively meaningful behaviour) and interactions (sequences of action– reaction). Beyond these, are abstractions (relationships, institutions, society, etc.), which can hardly be dealt with as empirical entities; they cannot be observed.

Furthermore, instead of seeking to arrive at generalisations about social reality, interpretivism seeks to interpretively understand the meanings that individuals attach to behaviour, things, events, etc. In other words, it seeks to understand the process by which reality is socially constructed. In the interpretive tradition, substantively, theorising about any social phenomenon has to be grounded in the context in which it is located. Hence, it is called ‘Grounded Theory’ (Glaser and Strauss 1967). A phenomenon taken out of its context loses meaning.

In interpretivism, answer to research questions is not posited in advance; in fact, engagement with the field can change the research questions themselves. So, hypothesis, in the sense it is used in positivism has no meaning in interpretivism. From the data collected in the field, the researcher may arrive at hypotheses and grounded theory. Thus, interpretivism as an epistemology, in general, primarily implies the use of inductive logic (discussed later).

The interpretivist epistemology does not depend on a single source or type of data. Anything (oral, visual, documentary, and unobtrusive measures) that can help in understanding the reality in question will be used as data. Thus, the use of multiple methods for collection and analysis of data is the norm. Much of this data is descriptive in nature, but, if available, numerical data is also used. The process by which different types of data – collected from different sources and using different methods – are put together to arrive at an understanding or explanation of a phenomenon is called triangulation (discussed later). Research in the interpretivist tradition, thus, can be likened to solving a puzzle.

In brief, in following constructivist ontology and interpretivist epistemology, a researcher is inevitably committed to qualitative approach as the main plank of study. Furthermore, the objective of arriving at an understanding of and explaining a complex phenomenon by locating it in its context requires the research design to be flexible. That is, decisions regarding various steps of the study cannot be taken in one go or adhered to meticulously. This makes the design of an interpretivist (qualitative) study cyclical and ‘soft’ (Hennink et al. 2011).

Self-check Exercise- 2

1. What is theory?

The term theory has a range of meanings: it is used broadly to encompass all thought and narrowly to refer to a single thought. Robert K. Merton has defined as ‘logically interconnected sets of propositions from which empirical uniformities can be derived’.

2. What is paradigm?

The term ‘paradigm’, according to Thomas Samuel Kuhn, is general ways of seeing the world (assumptions) which point to the nature of scientific work to be done and the kinds of theory that are acceptable. In sociology, as with many other scientific terms, however, the term paradigm is used vaguely. It denotes, ‘schools of sociological work or meta-theories, each of which is relatively self-contained, with its own methods and theories’

3. How do assumptions differ from hypothesis?

Assumptions are often confused with hypotheses. Assumptions are never tested empirically; they are, therefore, never accepted or rejected based on empirical evidence. Hypothesis, on the other hand, is a tentative answer to the research question. In other words, it is a proposition or set of propositions put forward for empirical testing; it is accepted or rejected based on empirical test. Often, hypotheses are derived from assumptions. Thus, the strength of an assumption lies in its capacity (a) to generate testable hypotheses and (b) to explain observations made or results arrived at.

4.   How do ontological and epistemological assumptions are related to each other?

These two sets of assumptions are intrinsically related. Ontological assumptions have to do with reality and the epistemological assumptions have to do with knowledge about that reality.

Component 2: Logic

Answers to research questions take the forming of reasoning. Reasoning is the process of using knowledge, existing or newly acquired, to formulate hypotheses, make predictions, draw conclusions, or construct explanations. The branch of philosophy concerned with the use and study of valid reasoning is called logic (from Ancient Greek logike). There are three methods of reasoning: deductive, inductive, and abductive; hence, there are three types of logic: deduction, induction, and abduction.

12. Deductive Reasoning

Deductive reasoning starts with given premises (or propositions) to arrive at a guaranteed specific conclusion: if the original premises are true, the conclusion must also be true. That is, it moves from the general rule to the specific application. The language of formal, symbolic logic looks more like a mathematical equation. A deductive syllogism (plain-English version of a mathematical equation) is expressed in ordinary language.

The process of linking the conclusion from the premises, called deductive inference, could be either valid or invalid. However, whether the conclusion arrived at through deductive reasoning is sound (true) or unsound (false) depends on the truth of the original premises, as any premise may be true or false. Thus, inferential process can be valid even if the starting premise is false. But the conclusion arrived at is invariably false if the premises are false, even if the inferential process is valid.

Assuming the propositions are sound, the rigorous logic of deductive reasoning can give us absolutely certain conclusions. However, deductive reasoning by itself cannot increase human knowledge because the conclusions it yields are tautologies – they are contained within the premises and virtually self-evident. That is, tautologies are statements that are true by necessity or by virtue of their logical form.

Deductive reasoning is behind the formulation of hypotheses, which are tentative answers to the research question. These hypotheses are tested with the help of appropriate data. This makes deduction integral to objectivist ontology and positivist epistemology. Accordingly, studies based on these ontological and epistemological positions are often described as following the ‘hypothetico-deductive model’.

13. Inductive Reasoning

Inductive reasoning begins with observations that are specific and limited in scope. Based on these observations, it arrives at a conclusion that is likely, but not certain. In other words, inductive reasoning moves from the specific to the general. Conclusions reached by the inductive method are not logical necessities; no amount of inductive evidence guarantees the conclusion. This is because there is no way to know that all the possible evidence has been gathered, and that there exists no further bit of unobserved evidence that might invalidate one’s conclusion.1 This cautions us against the language of inductively reached, probable conclusions.

The fact that inductive reasoning cannot yield absolutely certain conclusions does not mean that they are useless. In fact, evidence-based inductive reasoning can actually increase human knowledge. It can make predictions about future events or as-yet unobserved phenomena. It is useful in sociological research as the basis of grounded theory and a source for the formulation of hypotheses in areas where there is little generalised knowledge.

14. Abductive Reasoning

Deduction and induction are the two main and contrasting types of logic in research. A third type, abduction, is often mentioned in discussions on reasoning in research. Abductive reasoning typically begins with an incomplete set of observations and proceeds to the likeliest possible explanation for the set. It is akin to routine decision-making that does its best with the information at hand, which often is incomplete. For this reason, it is described as ‘taking one’s best shot’.

Apparently, abductive reasoning is similar to inductive reasoning. The difference between the two lies in that inductive reasoning requires the evidence that might shed light on the phenomenon to be fairly complete, whether positive or negative, but abductive reasoning is characterized by lack of completeness, either in the evidence, or in the explanation, or both. In science, often deductive and inductive reasoning are not just enough; a creative leap of imagination and visualisation that is apparently scarcely warranted by the mere observation may result in a path-breaking discovery. Albert Einstein’s theory of relativity, based on ‘thought experiment’ (with moving trains and falling elevators), is a good example of how abductive reasoning can be creative, intuitive, and even revolutionary.

In brief, in research, theory and logic are closely related. Commitment to objectivist ontology and positivist epistemology entails deductive reasoning, whereas commitment to constructivist ontology and interpretivist epistemology entails inductive (or sometimes abductive) reasoning.

Although logic is an important and necessary component of research, it alone or in combination with theory, is not sufficient. As Morris R. Cohen and Ernest Nagel put it succinctly, ‘Logic cannot guarantee useful or even true propositions dealing with matters of fact …’ (1968: 23). Besides theory (ontology and epistemology) and logic (deduction, induction, and abduction), what is necessary to answer our research questions is data or evidence, as it is called in common parlance.

Component 3: Method

15. Methods and Data

Having zeroed in on her/his research problem, and formulated her/his research questions and hypotheses (if any) to be tested, a sociologist will have to address issues relating to data. There are different types of data: primary and secondary; quantitative and qualitative; and objective and subjective.

Primary data are those required to be collected afresh by the researcher; secondary data are those already available to the researcher, though not in the form in which he/she wants it, as the same would have been collected for purposes not necessarily the same as those of the researcher. Data collected through interviews and questionnaires are an example of primary data, and data collected as part of the decennial census and large-scale sample surveys conducted by various organisations are an example of secondary data.

Quantitative data are numerical in form and they can be subjected to quantitative analysis to find out the distribution of traits, relationship between variables, and trends over time. Some of the variables chosen for study may be numerical by nature, e.g., age, income, family size, etc.; others may be nominal in nature to which numbers are assigned to facilitate quantitative analysis, e.g., caste, gender, religion, linguistic background, etc. Qualitative data are descriptive in nature; they are expressed through language and its nuances. Narratives of participants in the study and thick descriptions of events are examples of qualitative data.

Objective data are those that are external to the individual and to which the researcher has assigned meaning a priori. Dimensions of a house, age of a respondent, education of the family members, etc. are examples of objective data. Subjective data are those relating to the subjective makeup of the individuals studied. Attitudes and opinions are typical illustrations of subjective data.

There are different types of data and there are different methods of collecting and analysing them. Apparently, the choice of the type of data and the method adopted to collect them is a technical matter: that is, what is most helpful in answering the research questions on hand. On a closer examination, it will be revealed that this is not a technical matter alone. Both the research design and the choice of the type of data and the method for collecting them are significantly influenced by theory (ontology and epistemology) and logic (deduction and induction). This is illustrated in Table 1.

A research design based on objectivist ontology and positivist epistemology warrants the researcher to objectively collect numerical data that are statistically analysable. It prescribes quantitative approach. Secondary data are privileged in such a design as they are objective.2 Primary data collected from observation (when controlled for ‘observer effect’) is also privileged. Interestingly, primary data, collected though the interview method (using an interview schedule or questionnaire), are also used in studies designed in this tradition. Data collected from the interview method raises the problem of validity and reliability.3 These issues are generally addressed through pre-test of the interview schedule or questionnaire, as the case may be.

A research design based on constructivist ontology and interpretivist epistemology prescribes qualitative approach. It warrants the researcher to objectively collect different types of data in collaboration with the participants of the study. It lays emphasis on sustained and prolonged engagement (called ‘immersion’) of the researcher in his ‘field’. Ethnography is qualitative method par excellence. It privileges ‘thick descriptions’ and detailed narratives. Reflexivity of the researcher plays an important role in it. It permits multiple sources of data, including quantitative data.

16. Triangulation

Tapping of data from multiple sources and using multiple methods for collecting them necessitates integrating and synthesising them. The process of doing this called triangulation (read Module RMS 7 for details). Triangulation is ‘a method used by land surveyors and map-makers to locate a spot, by taking bearings from three known points and plotting their intersection’ (Abercrombie et al. 2000: 364). Since more than three methods may be used, the term ‘triangulation’ may not be the most accurate descriptor of mixing methods in sociology. Nevertheless, the underlying idea is important.

Norman Denzin (1989) has identified four forms of triangulation: (i) data triangulation – use of a number of types of data in a project; (ii) investigator triangulation – use of several different researchers; (iii) theory triangulation – application of multiple perspectives to interpret the data; and (iv) methodological triangulation – use of multiple methods to study a single issue. Although triangulation is widely practised in qualitative approach to research, to some extent it could also be useful in quantitative approach.

In triangulation or mixing methods, it is important to bear in mind that different sources of data and different methods of collecting them are premised upon certain ontological and epistemological assumptions. For example, observation and the use of secondary data are premised upon objectivist ontology and positivist epistemology. In-depth interview, participant observation, focused group discussion, participatory research appraisal, case study, life history, etc., are premised upon constructivist ontology and interpretivist epistemology. Certain methods of analysis – like interaction process analysis, content analysis, and conversation analysis – can be used in both the traditions. However, the epistemological assumptions they make and procedures they adopt are quite different.

A sound understanding of (a) theory (ontological and epistemological assumptions), (b) logic (deduction, induction, and abduction), and (c) methods and techniques of collecting data (observation, interview, ethnography, case study, life history, etc.) as also analysing them (quantitative and qualitative methods) is a prerequisite for a competent and skilful sociological researcher.

Self-check Exercise- 3

1. How does deductive reasoning differ from inductive one?

Deductive reasoning starts with given premises (or propositions) to arrive at a guaranteed specific conclusion. On the contrary, inductive reasoning begins with observations that are specific and limited in scope. Based on these observations, it arrives at a conclusion that is likely, but not certain. In other words, inductive reasoning moves from the specific to the general where as the reverse is followed in case of deductive reasoning. Commitment to objectivist ontology and positivist epistemology entails deductive reasoning, whereas commitment to constructivist ontology and interpretivist epistemology entails inductive (or sometimes abductive) reasoning.

2. A research design based on constructivist ontology and interpretivist epistemology would prescribe what type of research?

Such a research design would follow qualitative approach. This is because such research would warrant the researcher to objectively collect different types of data in collaboration with the participants of the study. It would put emphasis on sustained and prolonged engagement (called ‘immersion’) of the researcher in his ‘field’. Ethnography, for instance, is a qualitative method par excellence. It privileges ‘thick descriptions’ and detailed narratives.

17. Summary

Research in science, as also in social sciences, including sociology, involves the application of institutionalised principles, procedures, and techniques of acquiring new knowledge as also for refining existing knowledge. Notwithstanding differences between natural and social sciences, both accept research as the pathway to reliable knowledge. The nature of relationship among the three components of social research, namely, theory, logic, and method is reciprocal. Theory has a bearing on logic and method; logic, on theory and method; method, on theory and logic; and each one is, in turn, influenced by the other two. Hence, the methodology of social research is more than mere application of methods. Thus, questions related to research design, the choice of the type of data and the method for collecting them are deeply influenced by the ontological and epistemological positions of a research as well as the type of reasoning used to explain data. Hence, a sound understanding of theory, logic and methods and techniques of collecting data as also analysing them is a prerequisite for a competent and skilful sociological researcher.

Notes

1. In philosophy of science, this is discussed as ‘Verification’ versus ‘Falsifiability’ (see Taleb 2008).
2. One may recall here that Durkheim’s classical study on suicide was based exclusively on an analysis of secondary data.
3. Validity, which can take several forms, refers to ‘the capacity of a research technique to encapsulate the characteristics of the concepts being studied, and so properly to measure what the methods were intended to measure’ (Payne and Payne 2004: 233). That is, do the research techniques or instruments ‘capture the essence of what they are intended to represent’ (ibid.: 234). Reliability is ‘that property of a measuring device for social phenomena (particularly in the quantitative methods tradition) which yields consistent measurements when the phenomena are stable, regardless of who uses it, provided the basic conditions remain the same’ (ibid.: 195). In other words, ‘reliability is about being confident that the way data were gathered could be repeated without the methods themselves producing different results’ (ibid.: 196; emphasis original).

Reading Material

References

1. Abercrombie, Nicholas; Stephen Hill and Bryan S. Turner. The Penguin Dictionary of Sociology (Fourth edition). London: Penguin Books, 2000.
2. Bryant, Christopher G. A. Sociology in Action: A Critique of Selected Conceptions of the Social Role of the Sociologist. London: George Allen and Unwin, 1976.
3. Cohen, Morris R. and Ernest Nagel. An Introduction to Logic and Scientific Method (First Indian reprint). New Delhi: Allied Publishers, 1968.
4. Cohen, Percy S. Modern Social Theory. London: Heinemann Educational Books, 1968.
5. Denzin, Norman. The Research Act (Third edition). Englewood Cliffs, NJ: Prentice Hall, 1989.
6. Durkheim, Émile. The Rules of Sociological Method (Eighth edition, translated by Sarah A.
7. Solovay and John H. Mueller and edited by George E.G. Catlin). New York: The Free Press, 1966.
8. Glaser, Barney G. and Anselm L. Strauss. The Discovery of Grounded Theory: Strategies for
9. Qualitative Research. New Brunswick and London: Aldine Transaction, 1967.
10. Greer,  S.  ‘On  the  Selection  of  Problems’,  in  Martin  Blumer  (ed.):  Sociological  Research Methods: An Introduction. London: Macmillan, 1977, pp. 55–64.
11. Hennink, Monique; Inge Hutter and Ajay Bailey. Qualitative Research Methods. London: Sage Publications.
12. Hughes, John A. Sociological Analysis: Methods of Discovery. London: Nelson, 1976.
13. Kuhn, Thomas  Samuel.  The Structure of Scientific Revolutions (Second edition, enlarged). Chicago: The University of Chicago Press, 1970.
14. Merton, Robert King. Social Theory and Social Structure (Enlarged edition). New York: The Free Press, 1968.
15. Neuman, W. Lawrence. Social Research Methods: Qualitative and Quantitative Approaches (2nd edition). Boston: Allyn and Bacon, 1994.
16. Payne, Geoff and Judy Payne. Key Concepts in Social Research. London: Sage Publications, 2004.
17. Punch, Keith F. Introduction to Social Research: Qualitative and Quantitative Approaches (2nd edition).London: Sage Publications, 1996.
18. Ritzer, George. Sociology: A Multiple Paradigm Science. Boston: Allyn and Bacon, 1975.
19. Soanes, Catherine and Angus Stevenson. Concise Oxford English Dictionary (11th  edition). Oxford: Oxford University Press, 2004.
20. Taleb, Nassim Nicholas. The Black Swan: The Impact of the Highly Improbable. London: Penguin Books, 2008.

Suggested Readings

1. Blaikie, Norman. Approaches to Social Enquiry. Cambridge: Polity Press, 1993.
2. Cohen, Morris R. and Ernest Nagel. An Introduction to Logic and Scientific Method (First Indian reprint). New Delhi: Allied Publishers, 1968.
3. Denzin, Norman. The Research Act (Third edition). Englewood Cliffs, NJ: Prentice Hall, 1989.
4. Lessnoff, M. The Structure of Social Science: A Philosophical Introduction. London: George Allen and Unwin, 1974.
5. Merton, Robert King. Social Theory and Social Structure (Enlarged edition). New York: The Free Press, 1968, pp. 139–171.
6. Mukherji, Partha Nath. ‘Introduction: Methodology in Social Research – Dilemmas and Perspectives’, in Partha Nath Mukherji (ed.): Methodology in Social Research. New Delhi: Sage Publications, pp. 13–84.
7. Nagel,  Ernest.  The  Structure  of  Science:  Problems  in  the  Logic  of  Scientific  Explanation. London: Routledge and Kegan Paul, 1961.
8. Neuman, W. Lawrence. Social Research Methods: Qualitative and Quantitative Approaches (2nd edition). Boston: Allyn and Bacon, 1994.