Sociological Research

Chapter 3

Sociological research
Methods are not simply neutral tools: they are linked with the ways in which social scientists envision the connection between different viewpoints about the nature of social reality and how it should be examined. (Bryman 2008: 4)

Key issues
➤ What is sociological research?
➤ What different research methods are available to sociologists?
➤ What are the philosophies that underlie the collection and analysis of data?
➤ Why and in what ways have feminists criticized conventional sociological research? Introduction
We all engage in some form of research in our everyday lives by collecting and processing information and coming to conclusions about a product, service or

decision. You may have been stopped in the street and asked to take part in market research, testing the latest brand of a particular product. You may have heard research findings discussed in the news. Indeed, research is part of our everyday lives.
In this chapter we look at sociological research and explore why and how sociologists do research.
Sociological research can provide explanations for issues that affect us both as individuals and as members of larger groups. It can help us to make the links between personal troubles and public issues, understanding, for example, how your social background can affect your educational attainment and why people in some countries die from diseases that have long since been eradicated in other parts of the world. People hold a vast range of views on social issues, such as why certain people become criminals, why women are massively underrepresented in positions of power in the political and business world, and why fewer people attend religious services now than in the past. The findings of
1 09

Part 1 PT Introduction to the sociological imagination

sociological research should lessen the misconceptions and prejudices that often form the basis of commonsense views of many important issues such as these.

Why do sociologists do research? To help us to answer this question, we can ask another
– what does sociological research produce? Some typical answers may be ‘facts’, ‘knowledge’, and ‘ideas’. Each of these has a particular meaning but can be seen as dimensions of a larger concept – ‘evidence’.

We can see evidence as information that supports a statement, but we can also see it as a form of knowledge that stems from a range of sources. In every society there are many ways that we know what we know – many of which we take for granted. Scott (2002) identifies six basic categories of knowledge:
●
●
●

Stop and think
Jim and Sarah are ordering food in a restaurant.
Sarah wants to order the soup but is unsure whether it is vegetarian. Think about the way evidence is used in their conversation:
Sarah (to waiter) – Is the soup vegetarian?
Waiter – I’ll check for you . . . [comes back from the kitchen and tells her it is made with vegetable stock. She orders the soup. It arrives]
Sarah – It is definitely vegetarian – I can taste it.
Jim (to Sarah) – But how do you know it really is and they aren’t just saying that? [What proof is there?] Sarah – I’ve had it here before anyway.
This is a very ordinary everyday example of how evidence is part of our lives and how we constantly use evidence to make judgements and to give support to what we are saying. In this example we can view evidence as understood and used in a number of ways:
➤ Facts – The soup was made with vegetable stock.
➤ Confirmation – Checking this fact when the soup is served.
➤ Proof – At what point, and why, do you believe the evidence? Sarah wanted the soup so wanted to believe the waiter and the chef. Jim, on the other hand, wanted to tease Sarah so implied that he did not believe it.
Now think of a situation in which you have used evidence and reflect on why you accepted or rejected the evidence.

110

●

●
●

Common-sense knowledge – Refers to something everyone knows to be true, for example fire burns.
Authority-based knowledge – We tend to give a lot of credence to expert sources, such as doctors.
Experiential knowledge – We develop knowledge based on our own experiences, which can at times differ from knowledge from experts. For example, some parents believe, from their own experience, that autism is linked to the MMR (measles, mumps and rubella) immunization, but most medical experts deny such a link.
Traditional knowledge – Knowledge can also be based on practices passed down through generations to explain and justify many aspects of their lives, for example the Countryside Alliance, a UK organization supporting working and living in the countryside, often refers to ‘tradition’ in support of its beliefs.
Non-rational knowledge – Based on faith, for example a belief in God.
Scientific knowledge – Based on systematic, rigorous testing. We can identify a range of different forms of knowledge that we use in different ways. What we can also begin to see is that they are not all given equal weight: there is a hierarchy, with some forms of knowledge being privileged over others. In some societies, knowledge based on tradition is valued most highly, while in others, non-rational, faith-based knowledge is the dominant form. The nature of what is considered to be privileged knowledge will vary from time to time and from place to place. If we look at
Western Europe and trace back the forms of knowledge considered most accurate, and around which people lived their lives, we can see that religious belief (nonrational knowledge) was dominant for centuries. In the sixteenth century it was argued that the earth did not move and that people were the centre of the universe.
Copernicus devised a theory that challenged this, claiming that the earth in fact revolved around the sun.
He would not allow his theory to be published until he was dying because he was aware that the Roman
Catholic Church would treat such a claim as heresy.

Chapter 3 Sociological research

Indeed, when Galileo developed the ideas of
Copernicus, he was imprisoned by the church
(McIntyre 2005).
In contrast to earlier eras, in modern Western societies scientific knowledge is privileged. Science is generally accepted as the most valid form of knowledge, the most accurate way of knowing.

common sense, personal experience, faith in God, the word of an authority or tradition? Sociologists try to ensure that sociological knowledge goes beyond these forms. Instead, sociology derives its knowledge, its evidence, from research. Sarantakos defines research as the purposive and rigorous investigation that aims to generate new knowledge . . . Social research is about discovery, expanding all horizons of the known, confidence, new ideas and new conclusions about all aspects of life.

Stop and think
➤ Why do you think scientific knowledge is the dominant form of knowledge in Western societies? ➤ What alternative ways of knowing exist in these societies? (Sarantakos 2004: 4)

If we pick out some key phrases in this, we can further define sociological research and note the differences with knowledge based on common sense, faith or personal experience. Sociological research:
●

We could consider that scientific knowledge is privileged because it works; there are indeed many examples of the triumphs of science, space flight to name but one. Nevertheless, the effectiveness of a particular form of knowledge does not always relate to its general acceptance in society. In the chapter on health, we can see that the successes of many forms of complementary and alternative medicines have not resulted in their widespread inclusion in medical practice within Western societies. Forms of knowledge tend to be regarded as valid only if they fit with the ways in which people already make sense of the world.

●

Defining sociological research

●

So where does sociology fit in this discussion: what form does evidence take in sociology? Would it be sufficient for sociologists to base their claims on

Generates new knowledge – Sociological research creates new knowledge about a vast range of topics, some of which you can read about in the chapters of this book. Sociological research produces data
(for example, statistics, interview recordings and observations) and from these data new understandings of the world around us are built.
Expands horizons of the known – Research does not happen in a vacuum. Sociologists build on and develop the research of others. The critical evaluation of previous research is fundamental to sociology. Research should not produce findings because we want a particular result, or because we have a feeling that it should be that way. Rather, it has to be able to stand up to challenges from others who may disagree.
Involves purposive and rigorous investigation –
Sociological research is carried out in a systematic way. There is a clear plan concerning what the research involves and how it will be carried out, and this is documented to enable those reading the

A closer look
Primary and secondary data There are two forms of data that sociologists work with – primary and secondary.

●

Primary data are data collected first hand through a range of methods, including interviews, observations, experiments and questionnaires. ●

Secondary data are collected by someone other than the researcher, for example official documents, statistics and media reports. 111

Part 1 PT Introduction to the sociological imagination

findings to evaluate the research. All sociological research is planned around a particular methodological approach – a way of producing knowledge. In this chapter we will explore the key methodologies used in sociology and the implications of this for research practice. The foundations of sociological research
Some of you may be familiar with certain ways of doing research, for example using questionnaires, and you may have learnt the advantages and disadvantages of such tools. However, the choice of research tool is really the end point in planning research. To understand why researchers choose particular methods, and indeed for you to begin thinking about how you could design your own research, we need to consider the foundations of sociological research. In doing so, we will be using four key terms – methods, methodology, theory and epistemology. Crotty outlines their meaning:
●

●

●

●

Methods: the techniques or procedures used to gather and analyse data related to some research question or hypothesis
Methodology: the strategy, plan of action, process or design lying behind the choice and use of particular methods and linking the choice and use of methods to the desired outcomes.
Theoretical perspective: the philosophical stance informing the methodology and thus providing a context for the process and grounding its logic and criteria
Epistemology: the theory of knowledge embedded in the theoretical perspective and thereby in the methodology. (Crotty 2003: 3)

These definitions indicate how the four aspects relate, but this can be seen more clearly through a diagram. Figure 3.1 shows the movement from epistemology to methods.
Starting from the bottom and working up, we can see that, in order to explain the reason for choosing a particular method, for example interviews, the researcher must have a strategy or plan of action – a methodology. The methodology is influenced by the researcher’s theoretical perspective, which in turn is shaped by their wider beliefs about the status of the knowledge that we, as researchers, can access. For
112

Figure 3.1 The movement from epistemology to methods

example, if, as a researcher, I believe that the meanings and interpretations people give to the world are the fundamental aspects of life that sociologists should study (epistemology), then I am likely to engage with sociological theories such as symbolic interactionism, which reflect those ideas (theoretical perspective).
When carrying out research, I need to develop a strategy that fits with my interest in meaning (methodology) and finally I will need to choose tools to enable me to explore meanings (methods).
One of the most important things you learn from studying sociology is that people see and understand the world around them in many different ways. This is reflected in the different theoretical approaches you may have read about in Chapter 2. These different viewpoints have an impact on research because they shape the researcher’s views on the type of knowledge that research can produce, and so on the best way to study social life. Here we will focus on the most fundamental difference raised by Bryman’s question:
Whether social entities can and should be considered objective entities that have a reality external to social actors, or whether they can and should be considered social constructions built up from the perceptions and actions of social actors.
(Bryman 2004: 18)

Two positions are reflected within that question – objectivism and constructionism – and can be illustrated through an example. Social class can be defined and measured in many different ways. From an objectivist position, social class can be defined through a particular measure, which is likely to involve

Chapter 3 Sociological research

occupation, skill level and so on. This measure and indeed the class positions themselves exist independently of the perceptions of the people located within them. From a constructionist position, social class is regarded as having no independent, external reality. Rather, it is defined by perceptions, experiences and actions.
These positions are reflected in theoretical perspectives. Marxism and functionalism tend to reflect objectivist beliefs, while symbolic interactionism takes a more constructionist position. In relation to research, this fundamental difference also leads on to a key question – should sociology adopt the same principles for studying the social world as those used by scientists in their study of the natural world? The basic issue is how far the study of human behaviour and social life is fundamentally different from the study of the natural world. Stop and think
➤ Think back to lessons you have had in chemistry, physics or other sciences. How were you taught to find out about things?
➤ How might the methods you were taught in those subjects be used to find out about social issues, such as (a) why some groups get better educational qualifications than others; and (b) why men commit more crime than women do?

There are two very broad theoretical approaches in sociological research: positivism and interpretivism.
A positivist approach advocates the application of scientific methods to sociological research, while an interpretivist approach stresses the difference in studying human beings and the need to develop more applicable research strategies. We will look at each in more detail.

Positivism and sociology
Science is usually taken to refer to the natural sciences and (in the educational context) to subjects such as chemistry, physics and biology, which aim to explain the natural world in a logical manner by using specific techniques – the ‘scientific method’. Science aims to produce knowledge that can be trusted because it is known to be true in all circumstances and at all times.
It produces knowledge that has been empirically discovered and tested, rather than knowledge based on belief or faith.
Whether the scientific method can be applied to sociological research is a question that has excited considerable debate and divided opinion in sociology.
Positivism takes an objectivist stance and this is reflected in its support of the scientific method of research. Positivist research in sociology tried to discover ‘scientific laws’, which could explain the causes, functions and consequences of social phenomena, such as rates of crime and suicide. The term ‘laws’ reflects a fairly hard-line position; many

A closer look
Positivism and interpretivism Positivism and positivist research are based on the logic and method of science and scientific enquiry.
Positivism sees empirical science
(science based on experiments that are testable) as the only valid form of human knowledge. Auguste
Comte coined the term when

arguing that the application of the methods of the natural sciences to sociology would produce a ‘positive science of society’. In contrast, the interpretivist perspective maintains that there is a fundamental difference between the subject matter of the natural and the social sciences.
Interpretivism can be defined as the study of the ways in which people understand and interpret

the world in which they live. This perspective derives from the work of Max Weber. Although adopting a ‘grand theoretical’ approach, he argued that people cannot be studied in the same manner as the physical world. People attach meanings to what they do, and sociology has to acknowledge this and attempt to interpret those meanings. 113

Part 1 PT Introduction to the sociological imagination

positivists would aim to discover ‘tendencies’ rather than laws.
The sociologist [must] put himself [sic] in the same state of mind as the physicist, chemist or physiologist when he probes into a still unexplored region of the scientific domain. When he penetrates the social world, he must be aware that he is penetrating the unknown; he must feel himself in the presence of facts whose laws are as unsuspected as were those of life before the era of biology; he must be prepared for discoveries which will surprise and disturb him.
(Durkheim 1964: xIv)

Karl Popper (b. 1902) has been one of the foremost supporters of the scientific approach to research. In his view, scientific knowledge is the only valid form of knowledge. The development of knowledge is dependent on mutual criticism in that we learn about the world only by testing ideas against reality. Science proceeds by the disproving of generalizations, by refutation, according to Popper. Science should make generalizations or hypotheses that are open to testing.

Scientists should be detached observers, suspicious of common-sense ideas and intuitions and able to reject theories that they may hold when there is evidence against them.
Popper’s view has been widely accepted as an accurate account of what scientists do and has been instrumental in establishing the ‘hypothetico-deductive’ method: scientific knowledge and theory develop from the deducing and testing of hypotheses. The procedure is essentially a set of steps that describe how a particular piece of research is carried out. These steps are illustrated using the example of football hooliganism
(Figure 3.2):
1. Identification of a specific social issue or phenomenon that is to be investigated: football hooliganism.
2. Formulation of a hypothesis: football hooliganism is caused by young people in ‘dead-end’ jobs that have little future and allow no scope for creativity and self-expression. 3. Selection or design of a particular research method by which the hypothesis might be tested: checking of

Figure 3.2 Should football hooligans be subjects for ‘scientific’ research?
© Corbis

114

Chapter 3 Sociological research

4.

5.

6.

7.

police records of people arrested at football matches, followed up by asking those arrested what they feel about their employment situation.
Collection of information: examine police records and interview or give questionnaires to known football hooligans. Interpretation and analysis of the information gained: relating the data gathered to the hypothesis being investigated, how many football hooligans were in
‘dead-end’ jobs?
Formulation of a theory based on the tested hypothesis and the interpretation of the data collected: there is (or is not) a causal link between employment situation and football hooliganism.
Reporting the findings and conclusions, which must be open to discussion and retesting by others who may be interested: in some cases, the findings might be used in the formulation of policy, perhaps in deciding whether to introduce identity cards as a requirement for entry to football grounds or whether to segregate and fence groups of football supporters into self-contained areas of the grounds.

The steps need not be followed in the exact order indicated above. For instance, a scientist may observe something happening and examine it without having any clear hypothesis in mind as to why it occurred; the hypothesis may emerge later in the investigation, perhaps after some information has been collected.
In reality, research is rarely as clearcut as a textbook suggests. Stop and think
➤ Select another issue or phenomenon that sociologists might investigate and consider how steps 2–7 of the scientific method detailed above could be applied to it. Consider how this scientific method could be applied to an examination of (a) the extent to which the television portrayal of women influences children’s attitudes towards the role of women; and (b) the decline in attendance at religious services since the nineteenth century.
➤ Can you identify any weaknesses with this approach? Consideration of the scientific method and its applicability to sociology highlights the relationship between sociological theory and method and the difficulty of looking at research methods in isolation.
The formulation of a hypothesis and the type and style of questions asked will depend on the theoretical perspective favoured by the researcher. This theoretical perspective is also likely to guide the researcher towards certain ‘facts’. Hypotheses do not appear from nowhere: they might derive from beliefs and theories that are already held. In our example of football hooliganism, the hypothesis emphasizing the employment situation could derive from a criticism of government economic

A closer look
Durkheim and ‘social facts’ In reality there is in every society a certain group of phenomena which may be differentiated from those studied by the other natural sciences. When I fulfill my obligations as brother, husband, or citizen, when I execute my contracts, I perform duties which are defined, externally to

myself and my acts, in law and in custom. Even if they conform to my own sentiments and
I feel their reality subjectively, such reality is still objective, for
I did not create them; I merely inherited them through my education . . . Here, then, are ways of acting, thinking, and feeling that present the noteworthy property of existing outside the individual consciousness.

These types of conduct or thought are not only external to the individual but are, moreover, endowed with coercive power, by virtue of which they impose themselves upon him, independent of his individual will . . . If I do not submit to the conventions of society, if in my dress I do not conform to the customs observed in my country and in my class, the ridicule I

115

Part 1 PT Introduction to the sociological imagination

A closer look continued provoke, the social isolation in which I am kept, produce, although in an attenuated form, the same effects as a punishment in the strict sense of the word. The constraint is nonetheless efficacious for being indirect . . . Here, then, is a category of facts with very distinctive characteristics: it consists of ways of acting, thinking, and feeling, external to the individual, and endowed with a power of coercion, by reason of which they control him. These ways of thinking could not be confused with biological phenomena, since they consist of representations and of actions; nor with psychological phenomena, which exist only in the individual consciousness and through it.
They constitute, thus, a new

variety of phenomena; and it is to them exclusively that the term ‘social’ ought to be applied. And this term fits them quite well, for it is clear that, since their source is not in the individual, their substratum can be no other than society . . .
It is true that, when we define them with this word
‘constraint,’ we risk shocking the zealous partisans of absolute individualism. For those who profess the complete autonomy of the individual, man’s dignity is diminished whenever he is made to feel that he is not completely self-determinant. It is generally accepted today, however, that most of our ideas and our tendencies are not developed by ourselves but come to us from without. How can they become

policy as having caused an increase in the sense of frustration felt by certain groups of people or from a wider theory about the alienating nature of work under capitalism. As we noted above, positivist views about the applicability of scientific methods to the study of the social world have been disputed. We can now go on to explore a very different set of ideas about sociological research. Interpretivism and sociology
Interpretivism, influenced by the work of German social scientist Max Weber, is sometimes presented as a reaction to positivist conceptions of science. However, we should not jump to the conclusion that this leads to an abandonment of science. Rather, it is a way of redefining science to enable the development of the best methods of studying the social world. The social world
116

a part of us except by imposing themselves upon us? This is the whole meaning of our definition. And it is generally accepted, moreover, that social constraint is not necessarily incompatible with the individual personality. (Durkheim 1964: 1–4)

Questions
1. Using Durkheim’s broad definition, what ‘social facts’ have had an impact on your actions? 2. What do you think Durkheim means when he says that
‘social constraint is not necessarily incompatible with the individual personality’?
3. Compare Durkheim’s ideas on
‘social facts’ with Mills’ concept of the ‘sociological imagination’
(see pp. 00–00)

is considered to be very different from the natural world and so, it is argued, cannot be studied in the same way.
Human beings live and act in the world in a conscious manner; a crucial difference from rocks, for example.
Interpretivism takes a constructionist position, recognizing that people act consciously in order to create and re-create their worlds. We experience the world around us subjectively, and ‘society’ is not an objective entity; it is nothing more than the perceptions and actions of the people who exist within it. It is these perceptions and actions that are central to understanding human behaviour from an interpretivist stance. Rather than seeking causal relationships or laws that were said to determine people’s behaviour, sociology should seek to understand how and why people interpret the social world in various ways. This is a very different form of science from that advanced by positivists, and we can highlight this difference by looking at the issue of value freedom in sociological research.

Chapter 3 Sociological research

Can sociological research be value-free? The positivist argument that sociology should attempt to be as scientific as possible is based on the belief that only science can provide the ‘truth’. Scientists discover this truth by being completely objective, by dealing with facts. In their research, sociologists must be objective and neutral, must not take sides and should adopt an approach based on a position of value-freedom. This

idea of value-freedom in sociological research – or indeed in scientific research in general – has not been universally accepted by sociologists. The facts collected in research depend on the questions asked, and it has been argued that sociological research is inevitably directed by values – which are cultural products. From this perspective, knowledge is a cultural product. What a society defines as knowledge reflects the values of that society; another society and culture will accord other things the status of ‘knowledge’.

Case study
Value-freedom in sociology?
This extract is taken from Alvin Gouldner’s attack on the model of objectivity promoted by positivist sociology.
Does the belief in a value-free sociology mean that sociology is a discipline actually free of values and that it successfully excludes all non-scientific assumptions in selecting, studying and reporting on a problem? Or does it mean that sociology should do so? Clearly, the first is untrue and I know of no one who even holds it possible for sociologists to exclude completely their non-scientific beliefs from their scientific work; and if this is so, on what grounds can this impossible task be held to be morally incumbent on sociologists?
Does the belief in a value-free sociology mean that sociologists are or should be indifferent to the moral implications of their work? Does it mean that sociologists can and should make value judgements so long as they are careful to point out that these are different from ‘merely’ factual statements? Does it mean that sociologists do not or should not have or express feelings for or against some of the things they study?
Does it mean that sociologists should never take the initiative in asserting that some beliefs which laymen hold, such as the belief in the inherent inferiority of certain races, are false even when known to be contradicted by the facts of their discipline? Does it mean that sociologists should never speak out, or speak out only when invited, about the probable outcomes of a public course of action concerning

which they are professionally knowledgeable? Does it mean that social scientists should never express values in their roles as teachers or in their roles as researchers, or in both? Does the belief in a value-free sociology mean that sociologists, either as teachers or researchers, have a right to covertly and unwittingly express their values but have no right to do so overtly and deliberately?
I fear that there are many sociologists today who, in conceiving social science to be value-free, mean widely different things, that many hold these beliefs dogmatically without having examined seriously the grounds upon which they are credible. Weber’s own views on the relation between values and social science, and some current today are scarcely identical. If Weber insisted on the need to maintain scientific objectivity, he also warned that this was altogether different from moral indifference.
(Adapted from Gouldner 1973: 5–6)

Question
1. In what ways might the values of a researcher influence the following stages of research:
(a) the choice of research issue;
(b) the formulation of a hypothesis;
(c) the choice of research method;
(d) the choice of questions asked (if any);
(e) the interpretation of results;
(f) the presentation of findings?
117

Part 1 PT Introduction to the sociological imagination

Howard Becker is an advocate of the view that sociological research need not, and often cannot, be value-free. In his classic study of deviance, Outsiders
(1963), he argues that it is difficult to study both ‘sides’ involved in deviance objectively – the rule-breakers and rule-enforcers – and that, whichever group is chosen to study, there will inevitably be some bias. Becker suggests that there is a strong case for sociologists representing the views and attitudes of the deviants as it is their views that will be least known about and therefore most open to misrepresentation. C. Wright Mills, in his renowned introduction to sociology, The Sociological Imagination
(1970), also makes the point that social scientists cannot avoid choices of values influencing their work. Political and moral concerns are central to sociology and value-freedom is, therefore, impossible. In a similar vein to Becker and Wright Mills, Erving Goffman, in reflecting on his study of mental patients, Asylums, argued that it was unrealistic to aim to be value-neutral:
To describe the patient’s situation faithfully is necessarily to present a partisan view. For this bias
I partly excuse myself by arguing that the imbalance

is at least on the right side of the scale, since almost all professional literature on mental patients is written from the point of view of the psychiatrist.
(Goffman 1968: 8)

These arguments contrast with the positivist view that scientists must aim to produce value-neutral knowledge and that sociology should aim to be value-free. As we have seen, the debates raised by this question relate to the particular epistemological and theoretical position of the researcher. If a researcher begins with a view that the social world is objective and external, this will lead them towards a theoretical perspective that reflects this belief – that is, towards a positivist position. If, on the other hand, a researcher begins with a view that the social world is constructed by the perceptions and actions of those within it, then they will favour an approach that sees perceptions and actions as central to understanding behaviour.
In the discussion in this section, we have laid the foundations for understanding how research strategies
(methodologies) are developed.

A closer look
Weber’s concept of verstehen Social scientists cannot hope to formulate general laws on the basis of observing (patterns of behaviour) . . . It is clearly impossible to find out about a person’s motivations and intentions simply by watching their outward behaviour – the ideas people have about the society they live in and the machinations of their inner world cannot be directly observed. What people do has to be interpreted in the light of the meanings, motives and intentions behind their action . . .

118

Instead, (social scientists) have to ask questions about the beliefs people hold and the meanings they attach to action.
They have to concern themselves with the inner world of their subjects in order to understand why they act as they do . . .
Verstehen essentially involves the attempt to understand social action through a kind of empathetic identification with the social actor. The researcher must try to see the world through the eyes of the research subject in order to grasp the meanings, motives and intentions behind their action . . . through rigorous study, the

researcher can build up a picture of how the world appears to others and of the choices and constraints that they perceive . . . When interpretative understanding is achieved, the meaning of a given form of social action becomes clear.
(O’Connell Davidson and Layder
1994: 31)

Questions
1. Contrast this statement with
Durkheim’s ideas on ‘social facts’ on pp. 000–000.
2. What methods do you think are appropriate to Weber’s ideas on social action and verstehen?

Chapter 3 Sociological research

Methodologies: how to produce sociological knowledge Methodology refers to the strategy or plan for carrying out the research that allows the researcher to answer their research questions. It should be clear that the methodological approach of a researcher is not plucked from the air; it is very much shaped by their epistemological and theoretical stance. Positivists argue that it is possible to measure human behaviour in an objective sense and to identify cause and effect relationships. A suitable methodology is one that is deemed to be as objective as possible. Therefore knowledge is best produced by developing hypotheses that can be tested against empirical observations
(Popper’s hypothetico-deductive method). In contrast, an interpretivist approach would stress the need to understand how people see and interpret the world in which they live. A suitable methodology must facilitate this understanding – that is, it must enable the researcher to get the ‘insider’s view’. The most significant distinction between methodological strategies is between quantitative and qualitative research.

Quantitative and qualitative research
Although it is not possible to establish hard and fast connections between methodologies and particular theoretical positions, a quantitative methodology is likely to be used by those who favour a ‘macrosociological’ perspective and a qualitative methodology by those favouring ‘microsociology’. Many of the early, ‘grand’ social theorists adopted a macro-, or large-scale, approach in their writings. Karl Marx, for instance, set out to describe and explain the origins and development of modern industrial capitalist society.
He examined different types of society – tribal, feudal, capitalist and communist – and explored how one type of system evolved from another. Consequently, he had little interest in the personal beliefs of individuals or their motives for action and spent little time engaged in trying to discover what individuals actually thought.
Marx and other classical sociologists who are regarded as the founders of the discipline, including Durkheim and Weber, based their analyses of society on evidence from second-hand, general and historical sources rather than on original, first-hand research. At a similar period to the writing of these early sociologists, around the

turn of the century, social reformers such as Charles
Booth and Seebohm Rowntree were engaged in quantitative research in the form of large-scale surveys.
The concern of qualitative research tends to be with the small scale; a close-up view of society is taken. Such sociological research might focus on one aspect of social behaviour, perhaps a religious group or a juvenile gang.
Although this form of research might consider broad issues, the emphasis is different; there is less concern with generalizing about whole societies from particular instances and more focus on interpreting the behaviour of individuals and groups.
McNeill (1990) emphasizes the cyclical nature of trends in sociological research. After the Second World
War, the importance of objective data and statistical proof was stressed, particularly in the sociological work pursued in the USA and the UK. During the 1960s a reaction against this kind of sociological research developed, and the qualitative approach became the vogue. While various methodological approaches tend to be popular at different times, the extent of the division into an ‘either one or the other’ approach should not be overstated. Pawson (1989) suggests that the supposed conflict between positivists and interpretivists has led to certain methodological myths being propagated, in particular that the two traditions or approaches are mutually incompatible and in a state of permanent dispute. As he puts it, no good sociologist should get his or her hands dirty with the paradigm wars . . . both qualitative and quantitative approaches face identical problems and need to adopt common solutions.
(Pawson 1989: 31–2)

Pawson argues that quantitative sociology can be based on non-positivist lines. Sociologists should not be afraid to admit that research is influenced by theory: the fact that sociologists pick and choose the evidence that they will examine according to their theoretical interests should not be seen as a failure of positivism.
For example, in the sociological study of religion, one researcher might focus on the high proportions of people who believe in God and the growing numbers who express an interest in superstition and new forms of religious expression as evidence to demonstrate the continuing importance of religion in modern society.
Another might emphasize declining attendance at mainstream churches as the key evidence on the importance of religion in society. Pawson also points out that, even in ‘respected’ sciences such as physics,
119

Part 1 PT Introduction to the sociological imagination

A closer look
Quantitative and qualitative research
Quantitative research involves the collection and presentation of numerical data, which can be codified and subjected to detailed statistical testing. It follows the scientific method in so far as it attempts to discover and measure facts about society and social behaviour. Information is collected

and analysed in order to test a specific hypothesis (see p. 000).
Methods of research include gathering social data through questionnaires and structured interviews. These techniques usually involve studying large numbers of subjects so that the findings can be used as a basis for presenting general conclusions about social behaviour. the data collected are influenced by theory; even the instruments used by physicists, such as thermometers, have been developed and constructed on the basis of complex theories. Similarly, Pawson argues that it is mythical to see qualitative research as a coherent and superior alternative that can get to grips with the special character of human meaning.
Therefore, it is important to highlight the limitations of dividing all research into either quantitative or qualitative. There are various subdivisions within the two broad approaches, overlaps between quantitative and qualitative research, and many examples of sociological research that have adopted aspects of both approaches. Nonetheless, this broad division provides a structure for examining the various research methods used by sociologists. Before going on to discuss these methods in more detail, we need to consider the general methodological issues that all researchers need to address – sampling, gaining access and ethics.

Sampling
Researchers, including sociologists, cannot collect data from everyone they wish to survey. A 100 per cent sample would be the ideal, but for practical reasons researchers have to sample – that is, select people or information – from the population they are researching.
All researchers are systematic in their approach to sampling, but the particular approach taken will depend on the methodological strategy and research questions.
For example, you may want to find out how many
120

Qualitative research focuses on smaller units of society and on the understanding of social situations and the meanings that individuals attach to behaviour. It is a more subjective approach, whereby the researcher aims to understand and interpret the experiences of the individuals involved by viewing the world through the eyes of the individuals being studied. Methods of research include various forms of interviews and observation.

people in Scotland use complementary therapies, what kind of therapies, how often and for what conditions.
Your aim is to be able to say that x% of the population uses complementary therapies so it is very important to be able to achieve a sample that is representative of the whole population. However, if your research is a qualitative in-depth exploration of people’s experiences using complementary therapies, then your sample will be much smaller and so it is less likely that you would try to achieve a representative sample. Instead, the focus would be on the depth of understanding provided by the study of a particular group.

Sampling methods
●

●

●

Random or probability sampling is where all the members of a population have a chance of selection
– for instance, all schoolchildren of a particular age in a particular area – and perhaps one in every 100 is selected. Quota or stratified sampling is the major form of non-probability sampling. Here, the technique is to make the sample non-random deliberately by splitting it up beforehand, usually into categories such as sex, age or class, and then selecting a certain number for investigation from each category.
Random sampling and quota sampling are the most commonly used methods of sampling.
Snowball or opportunity sampling is where one person selected and questioned recommends another person, and so on.

Chapter 3 Sociological research

A closer look
Growing Up in
Scotland survey
Growing Up in Scotland is a large-scale longitudinal social survey designed to examine the characteristics, circumstances and behaviours of children from birth to late adolescence.
It is being carried out by the
Scottish Centre for Social
Research in collaboration with the Centre for Families and
Relationships at Edinburgh
University. The sample of children was drawn from child benefits records (CBR) held by the Department for Work

and Pensions. In order to accommodate a representative random sample of children that could be accessed efficiently by the interviewers from the study, a process known as primary sampling units (PSUs) was selected. These units enable the fieldwork to be managed efficiently by the study’s interviewers, since the selected families are clustered in particular geographic areas.
The PSUs are based on aggregations of data zones, which are newly geographic units being increasingly used as the basis of neighbourhood

Gaining access
When we decide to do research on a particular topic and with a particular group of people, we need to consider how to access this group. How will we make contact with the group, inform them and ensure there are enough people willing to take part in the research?
This involves knowing what sample you want to end up with and identifying places where it is most likely that you will find the kind of people you are aiming to recruit into the research. For example, if you are carrying out research with parents, you could start with plans to contact them through schools, parent and toddler groups and parenting organizations. Often a wide range of recruitment measures are needed; these may include strategies such as advertising in local newspapers and handing out flyers in shopping centres.
Each research project will have its own recruitment challenges. If a topic is very sensitive or not considered important, people may be less willing to participate.
The method of research may also have an impact on recruitment. If we want to do fact-to-face interviews, then there may be more issues than if we want to do questionnaires. If we want people to be involved repeatedly in experiments or in repeat interviews, that

statistics in Scotland. These data zones were randomly selected within a stratification scheme based on region and deprivation.
Within these sample PSUs, all eligible children are then sampled for the 0–1 years cohort, and three-fifths of all eligible children are selected for the 2–3 years cohort. Where a 0 to 1-year-old child has a sibling aged 2–3 years, one child is selected at random.
A total of 8000 children were enrolled into the study in
2005–06.
Source: Growing Up in Scotland
(www.crfr.ac.uk/gus/index.html).

might raise issues about commitment over a period of time. Finally, access to certain groups considered to be
‘vulnerable’ will require negotiation with ‘gatekeepers’.
For example, in the UK, if you are carrying out research with children and want to access them through schools, you need first to contact the local council to seek permission to approach the school, then the head teacher must be willing to allow you to contact the children in her school, then parents must give permission for their children to participate, and finally you must ask the child whether they want to participate in the research.

Ethical issues
Sociologists have a responsibility to ensure that the physical, social and psychological well-being of research participants is not adversely affected by the research. They should strive to protect the rights of those they study, their interests, sensitivities and privacy, while recognising the difficulty of balancing potentially conflicting interests.
(British Sociological Association 2002)

Ethical standards require that researchers do not put participants in a situation where they might be at risk of
121

Part 1 PT Introduction to the sociological imagination

Case study
Participant observation
Participant observation involves the researcher becoming a part of a group or community in order to study it; initially, then, there is a need to gain access.
Roy Wallis’s (1976) research into the Church of
Scientology illustrates the problems associated with this.
Scientology was founded by L. Ron Hubbard, a science-fiction writer, who developed a religious doctrine based on the idea that humans can regain the spiritual powers they have lost through a process of training that unlocks the doors to these powers. The movement has attracted considerable controversy since the 1960s.
Wallis’s interest in studying Scientology was hampered by its reluctance to be investigated. As a result he joined and followed an introductory course put on by
Scientologists and became a participant observer. As a secret or covert participant observer, Wallis had difficulty in showing his support for and commitment to the beliefs of Scientology and felt uncomfortable deceiving members of the sect, saying ‘Disguised or covert participant observation is easy when no participation beyond mere presence is required. When the interaction moves on to a more personal basis and participation of a more active kind is necessitated, role-playing dilemmas present themselves.’ He did not complete the course, but the participant observation gave Wallis an insight into
Scientology from the perspective of a potential recruit and provided the basis for another approach – this time he adopted an overt strategy and applied directly to the leaders for permission to interview members of the
Church. He was given permission but, probably due to his earlier deception, his research was hampered by a series of ‘mysterious and unpleasant happenings’ and his first attempts to publish his findings met with threats of legal action. He was forced to concede some editorial control to the Church and over 100 amendments were made before final publication.
At the end of his study Wallis reminds us that research takes place within the context of power:

122

The sociologist’s interaction with his subjects forms a part of, and takes place in the context of, the overall interaction between those subjects and the wider society. He may be seen as a potential legitimator or defender of their public image or a threat to it.
Those groups which are, or have been, in conflict with agencies of the wider society are likely to view a potential threat to their public image with hostility
. . . The sociologist who undertakes to study the social structure and dynamics of powerful groups . . . must expect his revelations to be met with hostility and the mobilisation of strategies to censor or even prohibit his work . . . While in the past sociologists have . . . displayed concern over the dangers of harming the interests of the powerless groups they have chosen to study, they should not altogether forget the problems of the relatively powerless sociologist faced with the threat of censorship.
(Wallis 1976)

Questions
1. What problems of access would be faced by research into (a) inmate culture in prisons; (b) the division of household tasks between partners; (c) sexism and/or racism in the school playground; and (d) the decision-making processes of business and companies?
2. Following the Macpherson Report in 1999 (see
Chapter 6), BBC reporter Mark Daly joined the
Manchester police force to examine the extent to which ‘institutional racism’ still existed and kept an undercover record of his experiences at the Police
Training Centre in Warrington using hidden cameras.
Despite pressure from the government and Mark
Daly’s subsequent arrest, the documentary The
Secret Policeman was broadcast in 2003.
(a) Why do you think Mark Daly chose this method of investigation?
(b) Do you think he was justified in using this technique? (c) What problems do you think this posed for him?

Chapter 3 Sociological research

harm as a result of their participation. Although arising in relation to concerns about medical research, the need to protect human subjects from being used as ‘guinea pigs’ in scientific research is also applied to social science research. Harm can be defined as both physical and psychological, for example causing undue stress, anxiety or emotional distress. The now infamous experiments by Milgram (1974), which explored our capacity for cruelty, illustrate the potential that social science has for causing psychological harm to participants.
The volunteers in Milgram’s experiments were asked to play the role of either ‘teacher’ or ‘learner’; the teachers were then told to deliver electric shocks to the learners as punishments when they made errors. Although the
‘teacher’ volunteers did not know the experiment was set up so that no electric shock was transmitted, the majority of them were prepared to deliver what they felt to be severe shocks. If you were one of these volunteers, how would you feel about yourself after you had taken part? For many people, participation in such an experiment may lead to self-doubt and a questioning of identity and so can be seen as causing harm.
Milgram’s experiment also raises the issue of what it means to be a research ‘volunteer’. The principle of voluntary participation requires that people not be coerced into participating in research. Fundamental to voluntary participation is the requirement for informed consent. Essentially, this means that prospective research participants must be informed fully about the procedures and risks involved in research and must give their consent to participate. However, any group of people who are used to others speaking for them and making decisions for them, such as children, pose a more difficult challenge when it comes to consent
(see Chapter 7 for more discussion of this issue). In addition, giving consent should be seen as an ongoing process, and consent can be withdrawn at any time.
In addition to gaining informed consent, researchers must ensure the particpants’ rights to privacy. There are two standards that are applied in order to help protect the privacy of research participants. Research guarantees participants’ confidentiality and anonymity: the participants are assured that their responses, their interviews and notes about them will not be shown to others until the information has been anonymized – that is, until all identifying features, most obviously names, are removed. Researchers need to anticipate any issues that may arise in order to ensure that confidentiality and anonymity are maintained. In many studies, not only is the identity of individual respondents

kept secret, but also the names of schools, organizations and towns are changed in order to avoid identification.
When discussing research ethics, there is an even more fundamental question that we can raise about academic research – research for whom? Who benefits from the research that sociologists do, and what responsibility do we have to those who participate in research with us? The British Sociological Association
(2002) ethical code states that
Sociologists, when they carry out research, enter into personal and moral relationships with those they study, be they individuals, households, social groups or corporate entities.
It is the very nature of the relationship we have with the people who participate in research that is the focus here. However, we can really consider the nature of this relationship only by looking in more detail at the issues raised in relation to specific research methods.

Methods of research
We mentioned at the start of the chapter that the choice of method is often the last point in planning research and is shaped by the methodology and in turn by the epistemological and theoretical position of the researcher. When investigating a particular issue or phenomenon, the sociologist is not limited to any one method. What is important is that the method fits – that is, it reflects the research strategy and it helps the researcher to answer their research questions.

Quantitative research
Surveys
The survey is usually a large-scale method of research that involves collecting information from large numbers of people. Although this information is typically gathered from questionnaires or interviews, a survey is not limited to any one technique of collecting information. In contrast to qualitative research, which provides a more in-depth study of social life, surveys tend to produce information that is less detailed but that can form the basis for making statistical generalizations over broad areas.
There are many well-known examples of large-scale surveys that have been used in sociological research, including the studies of poverty in the early years of the
123

Part 1 PT Introduction to the sociological imagination

A closer look
Reliability and validity
Examination of the different methods of sociological research should consider the concepts of reliability and validity. The degree of reliability and validity acts as a sort of quality control indicator in the assessment of any particular research method.
Reliability refers to consistency in research; however, the way that consistency is interpreted depends on the researcher’s epistemological and theoretical stance. From a positivist approach, reliability refers to when repeated applications of the same technique of collecting or analysing information produce the same results. To ensure this is possible, data collection methods should be highly structured so that any researcher would collect the same types of data from the same respondent. Unreliability would be a human error that could be

eliminated. In contrast, an interpretivist approach focuses on the fluid nature of reality based on our perceptions and actions and so would reject the notion that it is possible to repeatedly ‘capture’ that reality and expect it to be the same. Again, this difference of opinion relates to fundamental differences of understanding about the nature of the social world and how it can be studied.
Rather, interpretivists would seek to ensure reliability through consistency, though not necessarily standardization, in research procedures. Validity refers to the degree to which the findings of research can be relied on, and it involves an evaluation of all the methodological objections that can be made about the particular research. This is an obvious requirement for good research; however, as with

twentieth century undertaken by Charles Booth and
Seebohm Rowntree, the Oxford Mobility Studies of the
1970s, and the British Crime Surveys of the past 20 plus years. The British Crime Survey was set up in 1982 to investigate crime through a sample survey of 11,000 households that asked people about their experiences of crime. It was established to address the limitations of official crime statistics, particularly the fact that those statistics include only crimes that are known to the police. The British Crime Survey asked this large sample of people whether they had been victims of crime and whether they had themselves committed any crimes over the previous year. The British Crime Survey has shown that there is a great deal of crime, including serious crime, that does not appear in the official statistics.
The British Crime Survey also provides information on public fear of crime, perceptions of antisocial behaviour and attitudes to the criminal justice system.
124

reliability, the notion of validity applies differently to different approaches to research. Positivist quantitative research tends to refer to the extent to which a technique measures what the researcher intends it to measure. Qualitative research does not ‘measure’ social phenomenon, so for some issues we are better asking ‘are we accurately reflecting the phenomena under study as perceived by the study population?’
(Lewis and Ritchie 2003: 274).
Others argue that reliability and validity are inherently positivist concepts, and not applicable to interpretivist qualitative research, and suggest that new evaluative concepts should be developed.
Guba and Lincoln (1994) developed alternative criteria to evaluate the trustworthiness of research: credibility, transferability, dependability and confirmability.

Stop and think
➤ What kinds of questions could be asked to find out whether people have been victims of crime?
➤ What do you think might be the problems with and limitations of surveys that ask about people’s experiences of crime?

Surveys can be distinguished from other research methods by the forms of data collection and data analysis. Surveys produce structured or systematic sets of data, providing information on a number of variables or characteristics, such as age, sex and political affiliation. As questionnaires are the easiest way of getting such structured data, they are the most common

Chapter 3 Sociological research

A closer look
Can surveys measure social change?
Surveys involving the collection of information at one point of time are referred to as cross-sectional surveys; they provide a snapshot picture. The data gathered provide information such as who would vote for a particular political party or who belongs to a particular occupational grouping. Crosssectional surveys that are repeated at different times, such as the

British Crime Survey and the
General Household Survey, allow some analysis of change over time.
Longitudinal surveys provide data that enable the analysis of change at the individual or microlevel. One of the best-known longitudinal surveys in the UK is the cohort studies of the National
Child Development Study, in which a sample of children born in April
1958 have been followed from birth and interviewed at various

technique used in survey research. The analysis of data produced from a survey will provide standardized information on all the subjects being studied, for example how much television people watch a week or how people intend to vote. Surveys can also provide detail on the causes of phenomena such as variations in age, and suggest the extent to which this influences television-watching or voting behaviour.
Surveys are one method of collecting and analysing data that usually involve large numbers of subjects.
They are seen as highly reliable in that the data collected can be easily coded and analysed and should not vary according to the person or people collecting it. In conducting a large-scale survey, it is clearly impossible to investigate every single case or person, which raises the issue of sampling, and whether a smaller number can be used to represent a larger population. However, the fact that the data gained from surveys are necessarily restricted can be seen as a strength, in that it enables the analysis of these data to focus on standardized questions.

Criticisms of surveys
According to de Vaus (1986), the major criticisms of surveys can be classified as philosophical, techniquebased and political. Philosophical criticisms suggest that surveys cannot uncover the meanings of social action: they neglect the role of human consciousness, goals and values as important sources of action. To some extent these kinds of criticism are linked with more general

stages of their lives. More recently, the British Household
Panel Survey has been established at the University of Essex. This is made up of 10,000 people who were interviewed annually throughout the 1990s. Such longitudinal surveys are concerned with the behaviour of people over time and are therefore well suited to the analysis of change. (Adapted from Rose and Gershuny
1995: 11–12)

criticisms of quantitative sociology as being too rigidly scientific, being too focused on hypothesis testing and the collection of facts and statistics, and neglecting imaginative and creative thinking. Some surveys do, however, try to discover what people (say they) think; for example, the British Crime Surveys ask respondents about their attitudes to crime.
Technique-based criticisms emphasize the restrictiveness of surveys due to their reliance on highly structured ways of collecting data. The statistical emphasis is seen as reducing interesting issues to sterile and incomprehensible numbers. Political criticisms of surveys see them as being manipulative: the aura of science surrounding them gives power to those who commission and use the data. Survey data can be used to justify and further particular political interests.
However, it is important to remember that surveys can provide unreliable information and are sometimes wrong. The surveys carried out before the 1992 general election in Britain predicted a very close-run election; however, the Conservatives were elected for a fourth term of office with an overall majority. The pre-election opinion poll surveys overestimated the Labour vote by
4 per cent and underestimated the Conservative vote by
4.5 per cent. Much of this difference was attributed by the Market Research Society to ‘fundamental problems’ in the way opinion polls are conducted. With regard to voting intentions, there would seem to be a persistent and growing exaggeration of the Labour vote, which
125

Part 1 PT Introduction to the sociological imagination

suggests that the sampling procedures used are missing a significant proportion of Conservative voters, or that some Conservatives lie about their voting intentions.
Concern about the accuracy of survey polls has led to them being outlawed during election campaigns in some countries: in France, for example, they are banned for a period just before elections.

In defence of surveys
Despite the criticisms, O’Connell et al. defend the use of surveys as being part of the armoury of techniques available to sociologists. Surveys may have limitations, but they are a useful means of producing certain types of data:
Surveys are not the or best method for acquiring objective knowledge. But we should also recognise that survey methods are a valuable adjunct to other techniques. Surveys can play a vital role in confirming more qualitative research, in highlighting gaps in knowledge or issues that require further investigation, and in revealing broad patterns that might be missed if researchers relied solely upon qualitative methods.
O’Connell Davidson and Layder (1994: 114–115)

Collecting quantitative data: questionnaires and structured interviews
Although a number of techniques for collecting information are available to sociologists who conduct surveys, information is usually gathered from questionnaires and/or structured interviews. While there are clearly differences between questionnaires and structured interviews – in terms of the way they are administered and the issues and problems they raise – we shall focus here on some of the common elements.
Respondents (the subjects from whom information is sought) can be asked questions, which are either written down in a questionnaire or presented verbally in an interview. The questions used in survey research are always structured – that is, organized into a particular order and wording. However, the range and form of possible responses can vary from providing only a limited number of possible responses such as ‘yes’, ‘no’ and ‘don’t know’ or ticking one of a list of statements, or they can be more open-ended. The range of responses affects the extent to which data can be coded and analysed. Using the example of the question on voting given below, it would be straightforward to code

Case study
Asking questions about sex
We may think that members of the public would keep their sexual behaviour to themselves, but a series of surveys have gained a range of data.
Structured closed-ended questions through interview or by questionnaire are the conventional techniques: ●

●

Kinsey Reports (1940s –1950s) – Almost 20,000 interviews of volunteers (approximately 90 minutes per interview). Described as a ‘scientific voyage to explore the unknown world (of) the sex life of human beings’, these revealed some surprising sexual antics in conservative America.
1949 mass observation survey of 2000 men and women in UK (also known as ‘Little Kinsey’) –
Regarded as ‘too explosive to publish’.

126

●

Hite Report (1974) – Thousands of questionnaires to gain data on sexual experiences of women challenged many myths about female sexuality.

●

Scully (1990) Understanding Sexual Violence – Highly structured (and very long) interviews with 114 rapists.

●

British National Survey of Sexual Attitudes (1992) –
Almost 19,000 men and women in a random-sample survey. Published in spite of government withdrawl of funding. ●

The Global Sex Survey (2005) – More than 317,000 people from 41 countries took part in the world’s largest sex survey.

Visit the website www.durex.com/cm/gss2005results.asp to discover some interesting cross-cultural differences; for example, Greeks come top of the global sex chart, with an average 138 bonks per year, while the Japanese can manage only a weedy 45.

Chapter 3 Sociological research

A closer look
Question styles
In a questionnaire or structured interview that aimed to provide information on voting, a closed-ended question might be:

How did you vote in the 2005 general election? (please tick)

Why did you vote the way you did in the 2005 election?

Conservative
Labour
Liberal Democrat
Other

‘Conservative’, ‘Labour’, ‘Liberal Democrat’ and ‘other’ as 1, 2, 3 and 4: such coding would make the statistical analysis of the data easier.
As questionnaires and structured interviews are considered to provide data that are both reliable and quantitative, their use is generally advocated by positivist research in an attempt to provide a scientific basis for sociology. However, these methods do raise some awkward questions concerning the reliability and validity of sociological research.
How can we be sure that the people we want to question will agree? If they do, can we be sure that they are giving honest answers? Not only will the researcher have little control over these problems – it is not feasible to force people to be involved in research – but there is also the difficulty of ensuring that the way in which the question is asked does not influence the way that it is answered. It is important to be aware of the danger of leading questions and loaded words in the design of questionnaires and interviews.

Order effect
Another factor that can influence people’s responses to questionnaires and structured interviews is the way in which questions are ordered. In research into order effect in survey questionnaires, Schuman et al. (1985) noted a marked difference in the responses to the same question. They asked first: ‘Do you think the United
States should let Communist newspaper reporters from other countries in here and send back to their papers the news as they see it?’ They followed this with ‘Do you think a Communist country should let American newspaper reporters come in and send back to their papers the news as they see it?’ In that order, 44 per cent of Americans asked said ‘yes’ to the first question. Using a split sample technique, where half the sample was asked the second question first, the numbers agreeing

An open-ended question might ask:

with space left for a lengthy response. that communist reporters should be allowed to visit the
USA rose markedly, to 70 per cent. It would seem that an initial antagonism to foreign, communist reporters was significantly modified once people considered how they would feel about limitations on access to other countries for their ‘own’ reporters.

Bias
Survey research, with the use of structured methods of collecting data, is seen as the most effective way to provide an objective science of society. As well as the doubts raised concerning the influence of the wording and ordering of questions on the objectivity of such research, the extent to which bias is eliminated from these methods has been challenged. The way in which people respond to questions may be influenced by prestige bias, in that answers that might be felt to undermine or threaten prestige may be avoided.
People tend to claim that they read more than they do, for instance, or that they engage in more ‘cultural’ activities than they do. Answers to certain questions can reflect unfavourably on an individual’s lifestyle: negative answers to questions such as ‘Are you satisfied with your job?’ could be seen as being too self-critical, admitting one’s life to be a bit of a failure. Research into deviant behaviour is often confronted with this problem, as respondents may exaggerate or play down levels of alcohol consumption, drug abuse or criminal behaviour. A classic, widely reported example of respondents reacting to a word concerns the different responses to the terms ‘working class’ and ‘lower class’. A Gallup poll survey in the USA in 1933 found that 88 per cent of a sample of the population described themselves as middle class, while only 1 per cent said they were lower class. Members of the sample had been offered a choice of three alternatives – upper, middle and lower class.
127

Part 1 PT Introduction to the sociological imagination

A closer look
Leading questions and loaded words
Leading questions are worded so that they are not neutral: they either suggest an answer or indicate the questioner’s point of view. They are called leading questions because they lead the

respondent towards one answer over another or to one response only. Kinsey has been accused of this in his sex surveys, when he assumed that all respondents were sexually active and phrased his questions accordingly, for example
‘How often do you masturbate?’

A similar survey was repeated shortly afterwards, with the term ‘lower class’ replaced by ‘working class’; this time, 51 per cent of the sample described themselves as working class.
Finally, there is also the danger of interviewer bias.
Even with tightly structured questionnaires and interviews, the respondent might still react to the interviewer – to their age, gender or race, for example – and provide answers that it is felt the interviewer is looking for. In his research into the failure of black children in the US educational system, Labov (1969) found that black children responded differently to white and black interviewers; with the white interviewer, there seemed to be a sense of hostility that limited the responses from the children. Of course, race is not the only factor that influences interview responses.
As Lawson (1986) puts it, ‘the interviewee may be antagonistic towards interviewers for no other reason than a dislike of the clothes they are wearing’.

Stop and think
➤ You want to find out about attitudes to (a) unemployment and (b) juvenile delinquency.
Give examples of ‘loaded words’ that you should avoid in framing your questions.
➤ Write down an example of a leading question for both areas of investigation.
➤ Write down a ‘neutral’ version of those questions. ➤ How might prestige bias and interviewer bias affect the findings from surveys?

128

Loaded words excite emotions in the respondent that will be likely to suggest automatic feelings of approval or disapproval: the respondent reacts to the particular word or phrase rather than to the question itself.

This discussion of survey research has not covered the range of research methods that could be put under the heading of ‘quantitative’, and we shall look at some of the other methods used by sociologists later in the chapter. In essence, quantitative research attempts to follow the scientific method of positivism. The research should be reliable and replicable: the data should be collected systematically and be standardized so that, regardless of who collects the data, the same findings will always emerge. These findings should be generalizable, to a wider population, allowing laws to be established on their basis.
In the next section, qualitative research is examined.
Here, the interest is in the smaller-scale research; the focus is on ‘meanings’ and ‘experiences’. An attempt is made to understand the lives of those being studied; as well as the use of less structured, informal interviews, qualitative research has emphasized the importance of observation. At the risk of being overly repetitive, it should be stressed again that sociological research is not simply an either/or choice and that the trend in recent research has been to use a number of different methods, both quantitative and qualitative.

Qualitative research
There is considerable variety in what we term
‘qualitative research’, but we can outline some basic features. In doing so, it is useful to consider Mason’s question: ‘What is qualitative about qualitative research?’ (2002: 3 – 4). Qualitative research is:
●

Based on an interpretivist position – Qualitative researchers are interested in exploring the ways in which participants interpret and experience the world around them.

Chapter 3 Sociological research

●

●

Based on flexible methods of data collection –
Flexibility is central to qualitative research because there is a recognized need be able to respond to context within which research is taking place. This means that highly structured, standardized methods are not considered suitable.
Focused on developing explanations that take into account the complexity of the social world and the lives of the participants in the research – Detail and context are not seen as problems that may lead to unreliability – they are the very essence of the social world to be studied and of the qualitative methodology.

These features of qualitative research enable researchers to explore a wide array of dimensions of the social world, including the texture and weave of everyday life, the understandings, experiences and imaginings of our research participants, the ways that social processes, institutions, discourses or relationships work, and the significance of the meanings that they generate. We can do all this qualitatively by using methodologies that celebrate richness, depth, nuance, context, multi-dimensionality and complexity, rather than being embarrassed or inconvenienced by them.
(Mason 2002: 1).

Qualitative interviews
Interviews are one of the most common methods used within qualitative research, though they are very

different from the structured interview method we discussed earlier. Qualitative interviews can be loosely structured around a set of questions; be based on very little preconsidered structure other than some relevant topics; or be centred on the participant’s life history. A fundamental difference between these types of interview is the extent to which the interview is led by the researcher or the participant. All qualitative interviews intend to allow the participant to ‘tell their story’, but in some the researcher will gently lead the participant towards a discussion of certain topics that they consider to be relevant to their research questions, while in others the participant leads the direction of the interview. Despite this difference, there are features common to all qualitative interviews:
●

●

●

In depth – Qualitative interviews are intended to look at the in-depth opinions and responses of interviewees, rather than yes/no responses.
Interactive – Kvale (1996: 2) noted that ‘an interview is literally an inter view, an interchange of views between two persons conversing about a theme of mutual interest’.
Generative – Create new knowledge rather than simply excavate existing knowledge. As Gubrium and Holstein
(1997: 114) argue, both the interviewee and the interviewer are participants in the social process that is the interview: ‘Respondents are not so much repositories of knowledge, treasuries of information awaiting excavation, so to speak, as they are constructors of knowledge in collaboration with the interviewers’.

A closer look

●

Semi-structured interviews –
Topics or questions are developed for use in the interview (interview guide).
The interview guide is intended to initiate discussion from the participants, which will then lead to many further unplanned questions arising from their particular initial responses. This

on the person’s life story.
While all qualitative interviewing stresses the need for context, it is more explicit in this approach. flexibility also means that the questions will not necessarily be asked in a preset order.

Types of qualitative interview ●

●

Unstructured interviews – At times a very loose set of topics is used, or perhaps a single question is used to start the interview and thereafter the interviewee leads the direction of the topics covered.
Life history/biographical method
– The focus of the interview is

●

Focus groups – Small groups are asked to discuss social issues or services. The direction the discussion takes may be structured by the researcher, or the group may raise its own questions. 129

Part 1 PT Introduction to the sociological imagination

●

Informal style – The aim of good qualitative interviewing is to make the interview seem like a conversation rather than an interview. Despite having the appearance of being a conversation, considerable planning is required beforehand.

●
●

●
●

Planning an interview
It was indicated above that the interview may appear like a conversation while in fact it requires a lot of planning. Researchers will have developed research questions so will have some topics that they are interested in, and many will develop an interview guide as a starting point to explore these topics. The interview guide helps to clarify the researcher’s thinking about the research; it helps them to remember, especially in the first few interviews, what the key topics are. Some interview guides are many pages long and quite detailed; others are very short, with brief headings.
However, it should be remembered that the interview guide is not a preset list of questions to be read out, as in the case of a structured interview. The interview guide may be worded as questions, but these will not be all the questions the researcher will ask – far from it – and the interviewer may well not ask the questions in that particular order. In qualitative interviewing, the emphasis is on responding to what the participant is saying: thinking on your feet. This is a challenge and, when doing so, you should try to avoid:

using terminology; using leading questions; for example, asking ‘Were you furious when he said that?’ rather than ‘How did you react when he said that?’; using emotionally charged words or concepts; asking questions that would make people defensive; for example, when asking about child-care arrangements, you might want to ask about the amount of time parents spend with their child.
Asking ‘Do you spend enough time with your child?’ may elicit a defensive response; instead, you could ask ‘Do you have any difficulties in managing your time between work and home?’

Beyond questions
Interviews do not have to be simply about asking questions. There are other techniques we can use in an interview situation in order to communicate with interviewees and to generate data. One of the most commonly used is vignettes. Finch (1987) describes them as ‘short stories about hypothetical characters in specified circumstances, to whose situation the interviewee is invited to respond’. Vignettes can be used as an icebreaker to start interviews but are often used to facilitate discussion (Barter and Renold 1999). They are often used in research with children in order to reduce any pressure of a one-to-one interview that a child may feel. Vignettes are also useful when tackling sensitive

A closer look
Vignettes
This vignette was used in a project exploring children’s and parent’s perceptions of risk. Based on a letter to the problem page of a magazine aimed at 10- to 14-yearold girls, it was hoped that it would allow them to speak about whatever they felt were risks to children; about risks and specific locations, such as parks, about

130

parental controls and about age-appropriate behaviour. More specifically, it encouraged them to think about urban and rural differences. After reading the vignette, participants were asked to comment and the discussion then developed in line with their responses. My mum and dad are really strict and won’t let me go out

anywhere. It’s like they don’t trust me. We lived in a house in the countryside for a few years and it was fine there, but since we’ve moved to the city it’s changed. I’m not allowed to go down to the local park with friends even though
I beg them. They say I’m only
13 and they want to keep an eye on me.

Chapter 3 Sociological research

topics. Commenting on a story is less personal than asking about direct experience. Vignettes provide the opportunity for participants to have greater control over the interaction by enabling them to determine at what stage they introduce their own experiences.
When writing vignettes, the researcher must ensure that the stories are plausible to the participants. They should seem real and be suitable to the topic and the group with whom you are researching. The vignettes should also contain sufficient context for respondents to have an understanding about the situation being described but be vague enough to encourage them to discuss and speculate.

Interview challenges
Carrying out interviews can be very challenging and requires a range of intellectual, social and practical skills. As the interviewer, you need to be knowledgeable about your subject area; you should be able to interact well with others, actively listening and responding with appropriate questions; and you need to be organized in arranging and recording the interview. While many people will be open to discussing the particular issues that you want to explore, others will seem very reluctant.
As a researcher, you need to try and understand the basis of this reluctance and find ways to encourage discussions. At the other extreme, some interviewees may have no problem talking but may talk about anything and everything except the topics on which you want to focus. To a certain extent, you can plan ahead for circumstances such as these, but little can replace direct practical experience in building the skills required.

Power relations in interviewing
The relationship between interviewer and interviewee is necessarily one based on power – you are the researcher and have created the relationship with the interviewee and to some extent may want to shape its outcomes, for example through the range of topics covered. Feminist research has been highly influential in debates around this issue. A wider discussion of the feminist influence on sociological research is found towards the end of this chapter. Here, we will highlight the contribution to debates on the relationship between researcher and interviewee. While researching certain experiences of women, feminist researchers found existing techniques, or the conventional ways of using these techniques, inappropriate. In her research on the transition to motherhood, Oakley (1981) exposed the limitations of

conventional sociological criteria for interviewing. After six months as an observer on a maternity ward in
London, Oakley interviewed 66 women on four occasions during their pregnancy. In this context she found that, as a feminist, she had to reassess how she had been trained as a sociologist to carry out interviewing. She argues that in many textbooks the interview is presented as distanced from normal social interaction – as a clinical research tool. In order to maximize data collection, the subject must be put at ease, and yet at the same time the interviewer must remain detached in order to avoid ‘interviewer bias’.
In order to gain cooperation, interviewers must strike up ‘rapport’ but avoid involvement. Interviewing is presented as a one-way process in which the interviewer gathers information and does not emit any information.
The relationship between interviewer and interviewee is hierarchical, and it is the body of knowledge possessed by the interviewer that allows the interview to proceed successfully. Oakley is critical of this model, which, she argues, is a product of the desire for scientific status often associated with ‘malestream’ sociology (see the case study ‘Critique of positivism: male reason versus female emotion’ on pp. 000–000).
Feminist researchers argue that bias is introduced when an interview is taken out of ordinary everyday relations and becomes a constructed and artificial relationship. The orthodox model is highly problematic for feminists whose aim is to validate the subjective experiences of women. Oakley (1981) points out that so-called correct interviewing is associated with a set of values that, in our patriarchal culture, are more readily associated with the masculine, such as objectivity, detachment, science and hierarchy. Oakley argues that interviewing, which relies on subjectivity and equality, is devalued as it does not meet the ‘masculine’ standards of social science, rationality and scientific objectivity and that it has been seen as potentially undermining the status of sociology as a science. In the social science model of interviewing, feeling, emotion and involvement are conventionally denigrated.
The hierarchical model of interviewing is not congruent with feminist principles, which challenge all relations of dominance and submission. Oakley suggests that when a feminist interviews women, it would be morally and ethically wrong to use prescribed interviewing practice. In her research, Oakley (1981) discussed personal and intimate issues with women in repeated interviews, which inevitably meant that a personal relationship evolved. Oakley built up
131

Part 1 PT Introduction to the sociological imagination

relationships and became close friends with some of these women; she answered all personal questions and questions about the research. As Oakley herself was a mother, the women would ask about her opinion and experience of young babies and she would share her experience with them. Oakley argues that she could have taken no other direction than to treat the whole research relationship as a two-way process: the relationship cannot be left in the interview room but exists beyond the interview.

Stop and think
➤ Oakley emphasizes the importance of developing close personal relationships with the women she was researching. What difficulties might she have in doing this?
➤ How might this influence the information they gathered? ➤ Do you think this matters?

interviews. She felt that she was able to gain more trust in the interviews with Caribbean women because of their common background. However, this does not mean that she considered the data from those interviews to be ‘better’ than the data from interviews with white women. Both sets of data are equally valid, though the context in which they are produced may be different; what must be recognized and included in any analysis of data is the perceived impact that factors such as ethnicity and gender have on the interview process.
These issues are also very apparent when analysing qualitative interview data.

Analysing qualitative interview data
There are many different approaches used to analyse qualitative interview data, arising from the theoretical position of the researcher. However, Coffey and
Atkinson note that what links all the approaches is a central concern with transforming and interpreting qualitative data
– in a rigorous and scholarly way – in order to capture the complexities of the social worlds we seek to explain.
(Coffey and Atkinson 1996: 3)

The relationship between the researcher and the research participants is clearly a contentious issue.
Many qualitative researchers point to the need to be reflexive about their role within the research process.
Moving away from an emphasis on objectivity leads the qualitative researcher to a more reflexive research position. Qualitative research is an active process, the practice of which involves the researcher continually questioning themselves and their actions. Reflexivity involves critical thinking and awareness about what you are doing and why you are doing it; questioning what assumptions you have about the research; and documenting how your own views and experiences shape the research you do, how you interact with people and how you interpret what they say. As human beings engaging in interactions with others, despite the fact that we consider it to be research, we need to consider the impact that we have on the research. Our gender, ethnicity or social class may shape how participants react towards us or indeed how we react to participants.
Equally, our life experiences and values may influence the way we relate to a particular topic or a particular interviewee. For example, Scott (1999) carried out interviews with Caribbean and white women. She noted the significance of her own ethnic identity in doing the
132

Most qualitative researchers try to give meaning to the data, often by developing themes or concepts that help to pull the data together and make sense of them.
One such approach is outlined by Ritchie, Spencer and
O’Connor (2003). They outline their thematic approach as involving moving through stages from the systematic sorting and ordering of the data through to more indepth interpretation by developing descriptive accounts, looking for patterns, and ultimately developing explanations derived from the data. They, like most qualitative researchers, argue that, although it is systematic, the analytical process should be flexible and researchers should move forwards and backwards between the stages rather then seeing them as steps moving only in one direction.
Having looked in some detail at qualitative interviewing, we can now go on to consider the observational method. While some of the issues raised are similar, there are also specific challenges and concerns relating to observation.

Observation
Observation can be either participant or not; however, in sociological research, participant observation or

Chapter 3 Sociological research

fieldwork (the terms can be used interchangeably in introducing these methods) has been a widely and successfully used approach. Participant observation has its roots in anthropology and the studies of non-Western societies by anthropologists such as
Bronislaw Malinowski, Edward Evans-Pritchard and
Margaret Mead in the first half of the twentieth century.
These researchers lived with the peoples they studied, learned their languages and cultures, and provided fascinating accounts of such societies. More recently, this approach to sociological research has been used to study groups and cultures within Western societies. The work of sociologists at the University of Chicago in the 1930s (led by Robert Park) applied anthropological techniques to the lifestyles they found in the city of Chicago (Park et al. 1923).
They promoted participant observation, with researchers observing the life of social groups while actually participating in them. Our introduction to participant observation will look at examples of sociological work that have used this method and will discuss some of the issues raised by it.

A closer look
Ethnography
Hammersley (1992) highlights a basic disagreement over the definition of ethnography: some see it as a specific method of research and others as a more general approach to research.
Here, we adopt the latter usage advocated by Hammersley, ethnography being the in-depth study of a specific group or culture over a lengthy period. The emphasis of such study is usually on forms of social interaction (in, for example, a school, factory or juvenile gang) and the meanings that lie behind them. An ethnographic study

would typically involve observing the behaviour of a social group and interpreting and describing that behaviour. Hammersley summarized the key assumptions that ethnography makes about the social world and how it should be studied: ●

●

An understanding of human behaviour has to be achieved by first-hand contact with it; thus, ethnographers adopt a naturalistic focus and do their research in ‘real life’ settings.
Human actions do not consist of fixed or learned responses; to explain such actions, it is

necessary to understand the cultural perspectives on which they are based.
●

Research should aim to explore the nature of social phenomena rather than be limited to the testing of hypotheses; the emphasis should be on getting at the meanings and motivations that underlie behaviour.

These assumptions indicate why ethnography is linked with qualitative research and explain why ethnography uses methods of research that are less structured and do not follow the traditional scientific model discussed earlier.

133

Part 1 PT Introduction to the sociological imagination

Stop and think
➤ Participant observation has been used particularly in the study of unusual and deviant behaviour. One reason for this is the fact that such behaviour is, by its very nature, liable to be secretive and/or illegal and thus often difficult to study by more conventional means.
Why might participant observation be particularly useful for the study of crime and deviance?
➤ Participant observation has also tended to be used when researching the less powerful groups in society. Why might this be?

Participant observation enables the researcher to gain insight into behaviour through direct experience.
This does not mean that it is an easy method to use; the observer has to remain neutral while at the same time being closely involved with those being studied.
Howard Becker (1963), in his studies of the sociology of deviance, attempted to understand such behaviour through observation and close contact with the people he was studying. He described the role of the participant observer as someone who watches the people he is studying to see what situations they ordinarily meet and how they behave in them.
He enters into conversations with some or all of the participants in these situations and discovers their interpretations of the events he has observed.
(Becker 1982: 247)

The influence of the participant observer
The extent to which the participant observer might influence the group or activity being studied has to be considered. People are likely to behave differently when they are being observed, though this will depend to some extent on whether the subjects are aware of the fact that they are being observed. This raises the important distinction between open (overt) and secret
(covert) observation. If the observation is going to be done overtly, then the researcher will need to inform the subjects of the research of his or her identity. If it is to be done covertly, then the researcher will need to observe under some sort of guise. Even if the researcher does not tell the subjects and they do not know that
134

they are being investigated, the presence of another person may still, unwittingly, affect their behaviour.
Covert participation observation, in particular, highlights ethical issues about observing people without informing them (although, incidentally, this is habitually done by journalists reporting on celebrities). The points raised above about the influence of the participant observer and the ethics of such observation can be illustrated by looking at examples of sociological research. Eileen Barker’s (1984) study of the Reverend
Moon’s Unification Church adopted several research methods, including participant observation. Barker considered the argument that, in researching religion, more information would be gained if the researcher pretends to believe in the religion being studied – that is, if observation is done covertly. This raised for her the ethical question as to whether it was morally permissible to get information through false pretences.
In rejecting the covert style of observation, Barker pointed to the psychological difficulties of pretending to hold beliefs and performing actions that go against one’s conscience.
Although this sort of dilemma is not confined to the sociology of religion, it does arise particularly in the examination of secretive religious movements unwilling to be studied. Barker argues that it is possible to carry out overt research even into fairly closed groups such as the ‘Moonies’. Although suspicious of outsiders and of publicity, the Unification Church gave her access to a great deal of information and supported her research.
Their media image was so bad that the ‘Moonies’ could not believe that someone who really tried to understand them and listen to them would come up with a worse account than that provided by the media. Barker also pointed out that covert observation can hamper the research by making it impossible to ask certain questions or to appear too curious. In contrast, the recognized, overt observer is expected to ask questions and exhibit curiosity; indeed, they might find themselves being sought out and told things that the believers want to share with a ‘stranger’ who is not part of the particular organization. Both overt and covert observers have to be careful of the extent to which they influence the behaviour they are meant to be observing.
Two well-known studies of deviant groups that used the method of participant observation and that illustrate the issues and problems attendant on this style of research are described in the following ‘World in focus’ and case study.

Chapter 3 Sociological research

World in focus
When Prophecy
Fails
When Prophecy Fails (Festinger et al. 1956) is a classic covert participant observation study of a small deviant religious movement that predicted the imminent end of the world. In Christian-based movements this has usually referred to the second coming of
Jesus Christ and the establishment of a new Heaven and Earth, accompanied by the destruction of all sinners. Movements such as the Millerites in the 1840s and the Jehovah’s Witnesses have prophesied that the world would end at a certain time (although the Jehovah’s Witnesses have now abandoned these specific date-centred predictions).
Festinger and colleagues were fortunate to find a small group who appeared to believe in a prediction of catastrophe due to occur in the near future. The group was located as a result of a story in an
American provincial paper, the
Lake City Herald. This story detailed the prophecy of a
Mrs Marian Keetch that Lake City

would be destroyed by a flood before dawn on 21 December;
Mrs Keetch had received messages sent to her by superior beings from the planet ‘Clarion’ who had visited the earth and observed fault lines in its crust. The authors called on Mrs Keetch to discover whether there was a group of believers based around her. Their initial contact with her made it clear that any research could not be conducted openly. Given this, they described their basic research problem as ‘obtaining entry for a sufficient number of observers to provide the needed coverage of members’ activities, and keeping at a minimum any influence which these observers might have on the beliefs and actions of the members of the group’ (Festinger et al.
1956: 234).
The bulk of the study describes how the group prepared for the end of the world and then how the followers came to terms with dis-confirmation of their beliefs.
Fascinating though the whole study is, our interest is in the methodology of the research. On the whole, the authors and the additional hired

As well as issues concerning access, influence and ethics, participant observation raises practical problems.
Because of the difficulties involved in gaining access and trust, it is a time-consuming and therefore expensive method. Furthermore, it may well have significant effects on the lives of the researchers as well as the observed, as the case study below demonstrates.
This case study and the ‘World in focus’ box above illustrate that research by observation is by no means straightforward. Observers, particularly when working in a covert context, have to be detectives – listening,

observers they used were welcomed into the group as new converts. It was clear, however, that the involvement of a number of new observers-cum-believers was having a definite influence on the group itself, as the following extract illustrates:
One of the most obvious kinds of pressure on observers was to get them to take various kinds of responsibilities for recommending or taking action in the group. Most blatant was the situation that one of the authors encountered on
November 23 when Marian
Keetch asked him, in fact commanded him, to lead the meeting that night. His solution was to suggest that the group meditate silently and wait for inspiration. The agonising silence that followed was broken by Bertha’s first plunge into medianship . . . an act that was undoubtedly made possible by the silence and by the author’s failure to act himself.
(Festinger et al. 1956: 241)

probing and ensuring that their ‘cover’ is not blown.
Some of the practical and ethical considerations that need to be taken into account when pursuing sociological research are examined further in Activity 1 at the end of this chapter.

Analysing data from participant observation
In concluding this review of research by participant observation, we shall refer to Howard Becker’s (1982) reflections on the theoretical problems faced by those who adopt this method. Observational research
135

Case study
Hell’s Angels
Hunter Thompson’s (1967) study of the notorious
San Francisco Hell’s Angels motorcycle gangs highlights the potential dangers of covert participant observation.
My dealings with the Angels lasted about a year, and never really ended. I came to know some of them well and most of them well enough to relax with . . .
By the middle of summer [1965] I became so involved in the outlaw scene that I was no longer sure whether I was doing research on the Hell’s
Angels or being slowly absorbed by them. I found myself spending two or three days each week in
Angel bars, in their homes, and on runs and parties.
In the beginning I kept them out of my own world, but after several months my friends grew accustomed to finding Hell’s Angels in my apartment at any hour of the day or night. Their arrivals and departures caused periodic alarm in the neighbourhood and sometimes drew crowds . . . One morning I had Terry the Tramp answer the doorbell to fend off a rent collection, but this act was cut short by the arrival of a patrol car summoned by the woman next door. She was very polite while the Angels moved their bikes off her driveway, but the next day she asked me whether
‘those boys’ were my friends. I said yes and four days later received an eviction notice.
(Thompson 1967: 52–6)

Figure 3.3 Hell’s Angels
© Getty/Time & Life Pictures

1 36

As well as losing his accommodation, Thompson’s research proved physically painful.
On Labour Day 1966 I pushed my luck a bit too far and got badly stomped by four or five Angels who seemed to feel I was taking advantage of them.
A minor disagreement suddenly became very serious . . . The first blow was launched without warning and I thought for a moment it was just one of those accidents that a man has to live with in this league. But within seconds I was clubbed from behind by ithe Angel I’d been talking to just a moment earlier. Then I was swamped in a general flail. As I went down I caught a glimpse of Tiny, standing on the rim of the action. His was the only familiar face I could see . . . and if there is any one person a non-Angel does not want to see among his attackers it is Tiny. I yelled to him for help but more out of desperation than hope.
(Thompson 1967: 283)

Questions
1. Did being a participant observer influence
Thompson’s attitude towards the Hell’s Angels and his relationships with people outside the group?
2. What other research methods could provide insights into the Hell’s Angels? Give reasons for your answer.

Chapter 3 Sociological research

typically produces vast quantities of descriptive material, which the researcher has to analyse. This analysis,
Becker suggests, needs to be carried out sequentially, while additional data are still being collected – and these additional data will take their direction from the provisional analysis. Becker distinguishes four stages involved in the analysis of data gathered from observational research:
1. The selection and definition stage – The observer looks for problems and issues that will help to provide an understanding of the topic or organization being studied. The researcher will be using available data and material to speculate about possibilities. The credibility of the informants will also have to be considered: do they have reason to lie about or conceal information? In assessing the reliability of evidence, the observer’s role in the situation has to be examined: was observation overt or covert?
2. The frequency and distribution of the data have to be checked – Are the events typical? Does every member of the group respond in the same way?
3. The data have to be incorporated into a model that will help to explain it.
4. The presentation of the evidence and ‘proof ’ of the results – Quantitative, statistical data are relatively easy to present in tables and charts. However, the qualitative data gained from observation are much more difficult to present adequately. Such data are less easy to count and categorize; the data are also generally too detailed to present in full, which raises the issue of selectivity of presentation.

Other methods of research
Our discussion of different methods of research has focused on those adopted most often in sociological investigations: questionnaires, interviews and observation. Other methods of research are also used, and this section will outline the most prominent.

Secondary data
The research methods we have discussed – questionnaires, interviews and observation – involve the researcher collecting new (primary) data. However, sociologists also use data collected by others, including other researchers and organizations. One of the most common forms of secondary data used by sociologists is statistics from surveys conducted by or for the government. These include the census (a survey of the

whole population of the UK every ten years) and smaller surveys based on samples of the population, such as the General Household Survey. Official statistics are available to the public through a range of HMSO
(Her Majesty’s Stationery Office) publications, including the Annual Abstract of Statistics and Social
Trends, which provide summaries of statistical information under specific headings, such as
‘Education’, ‘Employment’ and ‘Crime’. See, for example, www.statistics.gov.uk.
Official statistics are a useful source of material for sociologists: they are already available and provide very full information that can be compared from one year to another. However, official statistics are collected by people other than sociologists for purposes other than social scientific research. Official crime statistics, for example, are collected and published by the Home
Office and are based on police records of crime; these statistics, then, omit many activities that break the criminal law but remain unknown to the police. There is a ‘dark figure of unrecorded crime’. It is impossible to say what this figure may be, but common sense would suggest that, for each crime the police know about, there are likely to be many more that they are not aware of and so unable to record (problems with crime statistics are discussed in more detail in Chapter 14).
In addition to the amount of crime not recorded, the official statistics are also influenced by the way that the police enforce the law. Whether the police decide to arrest and recommend the prosecution of individuals or to ‘warn’ them will affect crime figures, as will the amount of police resources, including police officers.
All the way through the process of dealing with crime, decisions are made about whether individuals are cautioned, arrested, charged and convicted, all of which affect the official crime statistics.

Stop and think
➤ What kinds of problems might the sociologist face in using official statistics on
(a) unemployment; (b) religious beliefs;
(c) child abuse; and (d) homelessness?

Content analysis and discourse analysis
Content analysis concerns how people communicate and the messages conveyed when people talk or write.
It is, essentially, the analysis of the content of
137

Part 1 PT Introduction to the sociological imagination

communication and involves the classification of this content in order to understand the basic structure of communication. In practice it involves researchers creating categories that relate to the particular issues being studied and then classifying the content of the communication into these categories. Content analysis produces quantitative data that can be measured and analysed objectively, although the extent of this objectivity can be questioned in that subjective judgements have to be made in creating the categories that form the basis of the analysis.
Content analysis has been used widely to investigate how the media transmit ideas and images. If examining the position of women in society, for example, a content analysis of school textbooks would probably find that the majority of characters are males and that women are portrayed in a more limited number of roles than men. A content analysis of television programmes would determine the percentage of leading characters that are male and female. Thus, the content of books, television programmes and films can reveal aspects of the society and culture in which they are situated and so can be an important source of information for sociologists.
One of the advantages of content analysis is that it is cheap and easy to conduct, but it is limited in its application and at the end of the day merely records the amount of times a sign or message occurs – it cannot tell us what they are taken to mean or whether they have any effect. As Sumner (1979: 69) has pointed out,
‘It is not the significance of repetition that is important but the repetition of significance’ – in other words, it does not matter how often a message is repeated if the audience makes no sense of it in the first place.
Discourse analysis is also a text-based research method that looks at and analyses forms of communication, or ‘discourses’ – in particular, speech and writing. Discourse analysis has been utilized in a number of subject areas, including sociology, and these different subject disciplines (such as psychology, philosophy, history and literature) emphasize different aspects of discourse analysis. However, all approaches highlight the social nature of discourse: the fact that the meanings of words, for instance, depend on who uses them, with whom they are used and where they are used. So, words and their meanings, and hence analyses of them, vary according to the social and institutional situations within which they are used. Thus, a number of different discourses exist at any one time, and these discourses can be, and often are, in competition and conflict with one another.
138

Discourse analysis does not aim to produce quantifiable data in the way that content analysis does, but rather it adopts a thematic approach, looking at areas and issues that are not (or are not so readily) quantifiable. For instance, Foucault, a key figure in the development of discourse analysis, described and examined discourses of madness (ways of communicating and thinking about the concept of madness) in one of his early studies (Foucault 1967).
Jupp and Norris (1993) identify two broad theoretical strands of discourse analysis, each of which focuses on the relationship between power and discourse. First, the analysis of discourse and power at a macrosocietal level: an approach emphasizing the role of the state in the production of ideologies and associated with structural Marxist analyses and the work of Althusser
(see p. 000). Second, the work of Foucault and his concern with why different forms of discourse and knowledge emerge at different points of time. In
Discipline and Punish, for example, Foucault (1977) discusses how different forms of punishment became dominant during different periods of history. Foucault sees discourses as mechanisms of power and control, but rather than highlighting the role of the state he argues that the state is only one of several points of control: power and control are pervasive throughout society.

Case studies and life histories
Other qualitative approaches to sociological research include case studies and life histories. A case study investigates one or a few particular cases in some depth.
Although case studies are only illustrative (‘one-offs’), they can be used as guides for further research. McNeill
(1990) makes the point that, to a certain extent, any piece of qualitative research could be described as a case study, given that all such research focuses on a relatively small group or on one particular institution. Case studies provide a different sort of data that can supplement other methods of research. As Walters puts it: the use of case studies aids the capturing of a process of events; they provide a sequence and structure that is often omitted in surveys or interviews . . . [they are] a useful means by which to chart ideas and develop themes for analysis.
(Walters 2003: 179)

Classic examples of case studies include Goffman’s
Asylums (1968), Beynon’s Working for Ford (1975) and
Hunter Thompson’s study of the Hell’s Angels (1967).

Chapter 3 Sociological research

A life history consists of biographical material that has usually been gathered from a particular individual, perhaps from an interview or conversation. As well as relying on people’s memories, other sources of information are used to build as detailed a picture as possible of the experiences, beliefs and attitudes of that individual; these sources might include letters and newspaper articles. As with qualitative research in general, the emphasis is on the individual’s interpretation of behaviour and events. Like case studies, life histories are unreliable as a basis for generalizing about social behaviour, but they can be valuable sources of insight for the sociologist.
In particular, life histories can provide more detailed information than other methods of research about the development of beliefs and attitudes over time. And, in general terms, it could be argued that an historical outlook and analysis is important in many, if not all, areas of sociological research. In such cases, the life-history method can complement the examination of documents and written historical records. Thomas and Znanieki’s (1918) study of the Polish community in Chicago is a classic example of this technique.

Stop and think
➤ Reflect on your own life history. Consider the extent to which your personal development has been influenced by cultural factors.

A multi-method approach
In this chapter, we have introduced and discussed a range of methods by which sociologists gather new evidence and apply already existing information in order to address social issues and problems. We have tried to present these methods as complementary rather than mutually exclusive ways of pursuing research. The complementary nature of different research methods is shown when two or more methods are combined in one research project; this process is given many names but is most often called triangulation. For example, Walters’
(2003) use of different styles of interview in his research into the politics of criminology was supplemented by a number of other research techniques. He used case studies so as to ‘bring to life’ research questions and

themes identified through his interviews; he examined academic literature; and he conducted an extensive documentary analysis of official information, including records of parliamentary debates, the League of
Nations, the United Nations and reports to the US
Congress.
Triangulation can be understood in many ways, often depending on the theoretical position of the researcher (Moran-Ellis et al. 2006). From a positivist stance, triangulation involves the idea that each method can corroborate the evidence produced by other methods, thereby producing a more valid account.
However, this approach has been criticized by interpretivist researchers because it assumes that
‘there is one, objective, and knowable social reality, and all that social researchers have to do, is to work out which are the most appropriate triangulation points to measure it by’ (Mason 2002: 190). As we have discussed, this assumption is not something with which many interpretivist, qualitative researchers would agree. Instead, triangulation is understood to be the combining of methods to enable researchers to explore different aspects of the same phenomenon. This illustrates the point made throughout this chapter: decisions around methods are explicitly connected to the researcher’s beliefs about what we can and should study in sociology and about what knowledge sociological research can produce. We end this chapter by looking in some detail at feminist research, not only because it illustrates these connections, but also because feminists have made significant contributions to key debates within sociological research as a whole.

Feminist research
Feminists have been concerned with the techniques used in carrying out research, the way research is practised and, more fundamentally, the processes via which sociological knowledge is formulated. They have been concerned with methods, methodology and epistemology. Feminists have produced a scathing critique of orthodox sociological research methodology
(Abbott and Wallace 1990).
As research generates the raw material of sociological theory and knowledge, feminists have in turn challenged how sociological knowledge is produced
(Spender 1981). They have recognized that knowledge is socially constructed, the product of social and cultural relations. All human beings may generate explanations of the world; not all of them become legitimized and
139

Part 1 PT Introduction to the sociological imagination

Case study
●

Researching new religions movements Following a bad press in which the Unification Church was accused of corruption, kidnapping and the brainwashing of its members, Eileen Barker set out to discover whether there was any truth to these allegations.

Why should – how could – anyone become a Moonie?
What possible explanation is there for the fact that men and women will sacrifice their family, their friends, their careers in order to sell tracts, flowers or candy on the streets for 16, 18 or even
20 hours a day? How can well-educated adults be persuaded to abrogate the right to decide whom they will marry, whether they can live with their spouse and whether they can bring up their own children? . . . When asking the question ‘Why do
Moonies do what they do?’ part of the answer is to be found through an understanding of how they came to be Moonies in the first place.
(Barker 1984:8)

Methodology
In order to find some answers, Barker decided to analyse the question on three levels:
Personal

●

She also kept a diary and met with parents and anti-cult groups.

●

Moonies attend residential courses and were therefore subject to ‘a carefully controlled situation which can exert a considerable influence on some of the guests’.

●

Moonies may target individuals in different ways to influence them – to manipulate what ‘resonates’ with each one.

●

Distorted versions of Moonie history and activity may be given to enhance its image.

●

She found no evidence of coercion.

●

Converts were ‘normal’ (no special characteristics).
Individuals eventually decide – ‘conversion to the movement is the result of a (limited) number of individual experiences; it is not the result of mass-induced hypnosis’.

Consequently, she concluded:
I have not been persuaded that they are brainwashed zombies. Interpersonal

●

●

Barker’s findings

The research question

●

Impersonal (social factors) – Questionnaires.
A pilot study (sample of 20), followed by a full study of all Moonies in the UK (about 500).
High response rate.

Impersonal (social).

She then chose to use triangulation and adopt a different technique to investigate each level;
●

●

Personal – In-depth interviews (36 at the start of the research) between 2 and 12 hours.
Interpersonal – Overt participant observation. Over six years Barker lived with Moonies at various centres.
She made no attempt to deceive them.

140

(Barker 1984: 233)

Questions
1. Contrast Barker’s approach with Wallis’ initial attempt to infiltrate the Church of Scientology.
2. Do you think the behaviour of the Moonies would be affected by Barker’s presence?
3. Do you think Barker’s findings may have been affected by her participation in the group?

Chapter 3 Sociological research

accepted explanations. Women have been excluded as producers of knowledge and as subjects of sociological knowledge. Stop and think
➤ Name as many sociology books or studies as you can that have been written by (a) women;
(b) men; and (c) both.
➤ What topics were covered by the books and studies in lists (a), (b) and (c)?
➤ Is there a pattern in the topics researched and written about by female and male sociologists?

Feminists examined the sociological research community itself. This includes academic institutions, departments, funding bodies and publishing houses.
Collectively these constitute what Liz Stanley (1990) has called the academic mode of production. She argues that the structure of this has contributed to the production of partial or limited knowledge. Certain individuals and groups have greater control over who can carry out research. Those individuals and groups with greater control include heads of departments, professors, referees and editors; women are underrepresented in these senior positions (Abbott
1991). Abbott and Wallace (1990) have described orthodox sociological research as ‘malestream’; they have described several levels on which sociological research has been male-centred.
Research has also been androcentric because it has been based on male experiences. Women have often been absent from research samples. For many years sociological research was carried out by male researchers and on male samples (Abbott and Wallace
1990). The majority of studies during the 1960s and
1970s in the sociology of work were of male paid labour. Findings and theories from these samples were often generalized to the whole population, including women. Goldthorpe and Lockwood’s famous affluent worker study suggested that manual workers were becoming increasingly affluent and were increasingly adopting the values and lifestyles of the middle class
(Goldthorpe et al. 1968). Yet they were describing only the experiences of male workers; women have a different relationship to the labour market (see pp. 000–000).

Men set the research agenda in sociology for many years and, as a result, areas and issues relevant to women were neglected or marginalized. Topics such as sexual violence, domestic labour, childbirth and contraception received very little attention until the feminist impact on sociology. Feminists believe that there must be feminist research that addresses issues and social problems that affect women. A key aim of feminist research is to make women visible, to observe, listen and record the experiences of women, and to write women back into sociological research.
(Our central argument) has been that sociology has ignored, distorted or marginalized women. We have also suggested that this is a result of the systematic biases and inadequacies in malestream theories, not just an omission of women from samples.
Malestream theories do not ask questions or do research in areas of concern to women, and frequently women are excluded from the samples; when they are included they are viewed from a position that sees males as the norm . . . Feminist sociologists have argued that it is necessary to develop feminist theories: theories that explain the world from the position of women, theories that enable us to rethink the sexual division of labour and to conceptualise reality in a way that reflects women’s interests and values, drawing on women’s own interpretations of their of their own experiences.
(Abbott and Wallace 1997: 283)

Feminism has had an impact more broadly on sociology. Mary Maynard (1990) argues that feminism has begun to reshape sociology as many non-feminist sociologists are beginning to consider women and gender in their research and analysis. Feminists have stressed that men are influenced by gender relations, hence gender should be examined not only in research on women but also in all research. They have moved beyond criticism and suggested principles that should underlie feminist research. Just as feminist theory is not a unified body of thought, similarly there is no one feminist methodology (Reinharz 1993).

Feminist principles
Stanley and Wise (among others) have argued that no one research method or set of methods should be seen as distinctively feminist: ‘Feminists should use any and every means available for investigating the condition of women in sexist society’ (Stanley and Wise 1983).
141

Part 1 PT Introduction to the sociological imagination

Feminists have considered how the logistics of particular methods engage with feminist aims, yet it is the principles that underlie selected methods that distinguish a range of emerging feminist praxes (Stanley 1990). Feminist research involves a commitment to a particular way of practising research, and this may shape how the specific techniques are utilized. Liz Kelly (1988) has suggested that ‘feminist practice’ would be a more appropriate term than
‘feminist methodology’, in order to avoid the assumption that particular methods are feminist.
Not all feminists share a common view of research methodology: there is debate about what is and what should constitute a feminist methodology (Harding
1987). However, a set of recurring themes and principles does emerge, four of which have been selected here.

They are of particular importance to feminist ‘practice’
(Kelly 1988):
●
●
●
●

The centrality of women’s experience
Research for women
The rejection of hierarchy and empowerment
The critical reflexivity.

Centrality of women’s experience
Feminists in their research draw on new empirical resources, the most significant being women’s experiences. Feminism moves away from the androcentric position, which sees male experiences as central, and places women’s experiences as the foundation of social knowledge. The task of feminist research is to explore how women see themselves and the social world.

Case study
Critique of positivism: male reason versus female emotion
Stanley and Wise (1983, 1993) have argued that positivism is problematic for feminists for several reasons.
First, it is based on a series of dichotomies: science versus nature, objectivity versus subjectivity, reason versus emotion and male versus female. The problem is that positivism elevates science, objectivity, rationality and the masculine, hence denigrating nature, subjectivity, emotion and the feminine. Feminists have argued that in
Western scientific thought and culture the masculine is associated with reason, science and objectivity, and the feminine with nature, emotion and subjectivity.
Second, Stanley and Wise argue that the positivist emphasis on objectivity divorces sociological knowledge from the social conditions in which that knowledge is produced. Feminists do not see research as orderly; they are suspicious of ‘hygienic’ research: hygienic research in which no problems occur, no emotions are involved, is ‘research as it is described’ and not ‘research as it is experienced’.
(Stanley and Wise 1983)

Feminists argue that researchers are always part of the social relations that produce particular findings. Their beliefs and values will shape the research. The private and public spheres, the emotional and the rational,
142

subjectivity and objectivity cannot be separated.
Feminists have argued that personal subjective experience is political and important and should be recognized as such in research.
Third, feminists argue that the scientific approach produces a division and an imbalance of power between social science researchers and those people whose lives they research. Social scientists are seen to have special knowledge and skills, they control the research process, and they come along and do their research on people.
Feminists have been keen to involve women in the research process itself in an attempt to reduce the imbalance of power and hierarchical relations.
In challenging the concept of objectivity, feminists have challenged the view that research should be value-free and apolitical. Indeed, they stress that feminist research not only must be of intellectual interest but also should further the political interests of women. Feminists argue that research should raise consciousness, empower women and bring about change.
(adapted from Harding 1987)

Questions
1. What is ‘hygienic research’ and why are some feminists critical of it?
2. How might a researcher’s values influence the way they study (a) crime and (b) poverty?

Chapter 3 Sociological research

McRobbie (1991) argues that the most important achievement of feminist research is the revealing of women’s hidden experiences, both past and present.
Underlying all feminist research is the goal of correcting both the invisibility and the distortion of women’s experiences by providing a vehicle for women to speak through. Women’s experience must provide the raw material for theory construction. They stress the need to ground theory in research and stress the two-way relationship between experience and theory. Indeed, many feminists reject any divide between theory and research. Research for women
Ramazanoglu (1991a) notes that there has been a distinct shift in feminist methodology from an earlier position that defined feminist research as research of and by women to research that has a political commitment to be for women. Duelli-Klein (1983) differentiates between research on women, which merely records aspects of women’s lives, and research for women, which ‘tries to take women’s needs, interests and experiences into account and aims at being instrumental in improving women’s lives in one way or another’.
Feminist research is committed to improving women’s position in society. Several feminists have stressed consciousness raising as a key role for feminist research in the emancipation process. Pollert (1981), in her participation observation of women factory workers, did not take a neutral stance but challenged both male managers and female workers about the sexist assumptions they made. She treated the situation as a consciousness-raising process for herself and the women in the factory.

Rejection of hierarchy and empowerment
A basic principle of feminist methodology has been not only to challenge relationships typical of traditional research based on hierarchy and power but also to aim to democratize the research process. Stanley (1990) argues that feminists should be committed to a belief that research and theorizing are not the result of the thinking of a group of experts different from those of
‘mere people’. Stanley and Wise (1983) warn of the danger of the emergence of an academic elite of feminist researchers who distance their activities from the mass of women. Feminists have attempted to restore their subjects as active participants in the research process and to ensure that knowledge and skill are

shared equally between researcher and subject.
There has been a range of feminist research aimed at empowering and actively involving the women involved (Lather 1988). The aim to change as well as to understand the world means that some feminists build conscious empowerment into the research design.
Lather (1988) coordinated a group called the
Women’s Economic Development Project in South
Carolina. Low-income women were trained to research their own economic circumstances, in order to understand and change them. This participatory research design involved 11 low-income women working as community researchers on a one-year study of the economic circumstances of 3000 low-income women. Information was gathered in order to bring action as a catalyst for change, to raise the consciousness of women regarding the sources of their economic circumstances, to promote community-based leadership, and to set up an active network of low-income rural women in the state to support new legislation concerning women’s work and educational opportunities.
Abbott and Wallace (1990) note that the logic of the feminist position on research seems to demand nonindividual cooperative research, where the researcher helps the women involved to undertake their own research. The researcher acts as an enabler. Subject and researcher decide together how the findings are to be used, although in practice this is difficult to achieve.

Stop and think
➤ Suggest how Abbott and Wallace’s notion of cooperative research could be applied in the following areas:
(a) male and female roles in the catering industry; (b) the sexual harassment of girls at school;
(c) homelessness;
(d) the role of women in the police force.
➤ What are the arguments for and against personal involvement in research?

Critical reflexivity
In the section on qualitative interviewing we looked at the importance of reflexivity in qualitative research, highlighting the work of feminists. As Mauthner notes,
143

Part 1 PT Introduction to the sociological imagination

Reflexivity is a central tenet of a feminist methodology whereby the researcher documents the production of knowledge and locates herself in this process.
(Mauthner 2000: 299)

She described how, in her study of sister relationships; her own experiences shaped the research:
Clearly, my own sistering experiences coloured my interpretation of the data: they influenced the way that I listened to the narratives, the patterns that I noticed in the data, and the themes that I pursued.
(Mauthner 2000: 300)

For Oakley (1981) it is essential that the relationship between researcher and participant is reciprocal – that is, it involves the mutual exchange of information, ideas and personal experiences. In doing so, the need to be reflexive becomes a central part of the research process.
However, there are also difficulties for feminist researchers in establishing a fully reciprocal relationship.
Mauthner describes how she engaged with the participants and shared her own sistering experiences during interviews but noted that, when analysing and writing up the data, she stepped back from this reciprocity. In part this was to preserve her privacy; researchers are not afforded the confidentiality and anonymity of research participants. The notion of a mutual exchange was also limited at the analysis stage by the use of sociological theory in making sense of the data. While theory is an essential part of qualitative data analysis it also ‘distances the participants from the research product’ because they do not have access to that body of knowledge (Mauthner 2000: 301).
Similarly, McRobbie (1991) argues that feminists must recognize that there is an unequal distribution of privilege. Feminist researchers often represent powerful educational establishments. They must acknowledge that this may be one reason why women are willing to participate in research. For example, she criticizes
Oakley (whose work we considered earlier), who she claims fails to consider the imbalance of power between herself as a researcher and the young mothers:
She does not concern herself with the fact that pregnant, in hospital . . . the women were delighted to find a friendly articulate knowledgeable woman to talk to [sic] their experiences about . . . their extreme involvement in the research could also be interpreted as yet another index of their powerlessness. (McRobbie 1991: 79)
144

There is always the danger that some women will use the experiences of women’s oppression to further their careers, with women’s suffering becoming a commodity. Angela McRobbie speaks honestly about the fact that doing research sometimes feels like
‘holidaying on other people’s misery’. She describes an interview with a 19-year-old woman who had been brought up in care:
I was almost enjoying the interview, pleased it was going well and that Carol was relaxed and talkative.
Yet there was Carol with her eyes filling up with tears as she recounted her life and how her mother had died.
(McRobbie 1991: 77)

McRobbie also challenges the assumption that the feminist researcher will necessarily understand the women because of their ‘shared’ oppression. Women have a multiplicity of experiences. Feminists may have valuable personal experience, but they cannot assume that this will be the same as those they are researching.
While feminism attempts to foster sisterhood, it cannot naively assume that women are bound together purely on the grounds of gender.
Therefore, although reflexivity is a key strand in feminist ‘practice’, it raises many challenging questions concerning the relationship between feminist researchers and the women involved in research.
Although the principles enshrined in feminist research cannot eliminate the possibility of exploitation, they can serve as a check against it. As part of the conscious reflexivity, feminists attempt to be explicit and open about power relations as they operate in the research process. Quantitative or qualitative methods?
It is often assumed that feminist research involves only qualitative methods. Some feminist sociologists have been critical of quantitative methods. They have argued that questionnaires and structured interviews precode experience, producing a false body of data that distorts the actors’ meanings. Graham (1984b) claimed that survey methods and structured interviews ‘fracture women’s experiences’. Barbara Smith (1987) argues that there are aspects of women’s lives that cannot be preknown or predefined in such a way. Some feminists have stressed that the female subject gets lost in social science survey research. Oppression is such that it cannot be ‘neatly encapsulated in the categories of

Chapter 3 Sociological research

Case study
Women’s Leisure, What Leisure?
Green et al. (1990) used a social survey in combination with unstructured interviews and discussion groups in what is the most comprehensive study of women’s leisure to date in Britain. They wanted to collect both general information about the types and levels of women’s leisure participation and more detailed knowledge about women’s perceptions and attitudes to leisure. They argued that, as feminists are concerned both with understanding patriarchal structures that oppress women and with seeking to change them, they must utilize the strengths of quantitative evidence. They argued that using a survey enabled them to generalize from their results to the larger female population and to exert greater political influence. They wanted to provide a statistical body of research that could actually form the

survey research’ (Graham 1984b). On many of these points feminist arguments overlap with ethnographic researchers. Stop and think
➤ On the whole, feminists have favoured qualitative methods, claiming that they fit more comfortably with feminist principles as well as being more appropriate and sensitive to women’s experiences of oppression. Why do you think feminist research has favoured qualitative methods?

Although feminist research has tended to be defined in terms of qualitative research, a growing body of feminists have attempted to break down the distinction between quantitative and qualitative research. They have stressed that there is nothing inherently sexist with quantitative research methods and techniques such as surveys (Jayaratne 1993; Kelly et al. 1992; Pugh 1990).
If used sensitively, they can complement broader feminist research aims.

basis for more informed policy decisions. They pointed out that policymaking bodies were more impressed with statistical data, and it was crucial that such bodies examine and take note of their findings. They were fully aware of the limitations of the survey method in the context of women’s lives and hence used qualitative research to complement the quantitative data. Hence the shift towards triangulation (see pp. 000–000) is also reflected in feminist research. Feminists will use any appropriate method to expose and oppose women’s oppression. Question
1. What ‘limitations of the survey method’ might there be in pursuing research into women’s leisure activities? Many feminists have also proposed that the use of qualitative methods does not necessarily overcome some of the problems identified with quantitative methods. Stacey (1988) has pointed out that, although ethnographic methods seem ideally suited to research in that they involve empathy and allow for an egalitarian, reciprocal relationship, they may expose the research subjects ‘to greater risk of exploitation, betrayal and abandonment by the researcher than does much positivistic research’ (Stacey 1988: 21).
There is a great deal of debate about what a feminist methodology should contain. It is important to remember that there is not one easily identifiable feminist research methodology (Reinharz 1993).
Yet there is some consensus about principles that feminists should consider when carrying out research.
Kelly notes this ambivalent position:
There is not, as yet, a distinctive ‘feminist methodology’. Many of the methods used by feminist researchers are not original. What is new are the questions we have asked, the way we locate ourselves within our questions and the purposes of our work.
(Kelly 1988: 5–6)
145

Part 1 PT Introduction to the sociological imagination

Summary
● Sociological research involves the gathering of relevant material and data and interpreting and analysing

them.
● In undertaking research, the sociologist can use a variety of techniques. The method chosen will be

influenced by the researcher’s views about the social world and by the theoretical approach favoured by the researcher.
● One issue that has been central to the style of research adopted has been the extent to which sociological

research should follow the methodological approach of the natural sciences. A by-product of this issue has been the debate over whether sociology can and should be value-free and the extent to which the data gathered by sociologists are reliable and valid.
● Quantitative research is associated most closely with the conventional scientific methodology. Surveys

involve collecting data from large numbers of people, usually from questionnaires or interviews that ask people about their behaviour and attitudes. Surveys are typically based on a sample of respondents drawn from a specific population. The use of official statistics and content analysis are other examples of quantitative research.
● Qualitative research gathers more detailed information from a smaller number of respondents. The focus is

on the experiences of people and the meaning given to such experiences. Observation, in-depth interviews and case studies are qualitative methods.
● Feminist research has criticized orthodox sociological research methodology for being ‘malestream’, i.e.

centred on men. Although feminist research is not a unified body of research, a major focus has been on women’s experiences and how an understanding of them can help to explain and improve women’s position in society.

Links
As with sociological theory, sociological research also underpins all of the topics and areas of sociology covered in the rest of this book. Rather than indicate all of the links, we would urge you to think about the research processes involved in

the various studies and examples of sociology you come across in the remaining chapters and to consider the sort of research issues that those studies would raise.

Further reading
Becker, H.S. (1963) Outsiders: Studies in the Sociology of Deviance, New York: Free Press.
As well as ‘methodology’ texts, it is good practice to look at the methodological sections of specific sociological studies. Such sections may be found in the introductions to studies or as appendices. Chapter 9 in Becker’s classic study of deviants, Outsiders, for instance, discusses some
146

of the key issues attendant on studying deviant behaviour. Bryman, A. (2008) Social Research Methods, 2nd edn,
Oxford: Oxford University Press.
Although focusing on social research rather than sociology, this general text provides a very accessible but detailed introduction to all the topics covered in this chapter.

Chapter 3 Sociological research

Durkheim, E. (1964) The Rules of Sociological Method,
New York: Free Press.
With so many secondary sources available, there is a tendency not to consider reading the originals. However,
Durkheim’s key work on the studying of social phenomena is no more difficult to read than some of the commentaries on it. It is a short study that sets out the requirements that rigorous sociological research needs to follow.
Gilbert, N. (ed.) (2001) Researching Social Life, 2nd edn, London: Sage.
This is a collection of papers that outlines the main ways in which sociologists gather data and describes a range of ways of analysing such data. The connections between quantitative and qualitative research methods and their theoretical bases are emphasized throughout. The papers are written by experienced social researchers who reflect on their own research experiences.
Hammersley, M. (ed.) (1993) Social Research:
Philosophy, Politics and Practice, London: Sage.
This collection contains some particularly useful material on the politics of the research process with articles on, for instance, issues of race, gender and power in social research. Ritchie, J. and Lewis, J. (2003) Qualitative Research
Practice, London: Sage
A very practical book giving advice on ‘doing’ qualitative research and framed within an interesting discussion of the main debates. There are plenty of examples to illustrate the processes they describe.
Salkind, N.J. (2000) Statistics for People Who (Think
They) Hate Statistics, London: Sage.
The title says it all – a book for those who want to learn the basics of statistics for social research but who are worried about their abilities in this area.
Silverman, D. (2001) Interpreting Qualitative Data, 2nd edn, London: Sage.
The second edition of this popular introductory text takes account of the expansion in qualitative research since the
1990s.
Stanley, L. and Wise, S. (1993) Breaking Out Again:
Feminist Ontology and Epistemology, London:
Routledge.
This updates and reviews the main arguments in feminist thinking and research since the original Breaking Out – an important feminist text, first published in 1983, that challenged conventional positivist practices in sociological research. Websites www.statistics.gov.uk www.asanet.org/student/archive/data.html

The UK National Statistics Gateway
Provides access to a range of links to government statistics.

American Sociological Association

www.soc.surrey.ac.uk/sru

www.cessda.org
Council of European Social Science Data Archives

University of Surrey, Sociology Department
Here you can find articles on research published by the
Sociology Department at the University of Surrey.

Activities
Activity 1
Problems in researching a deviant religious movement This extract illustrates some of the difficulties of research into deviant religious movements. It is taken from the methodological appendix to When Prophecy Fails (Festinger et al. 1956), the classic study of a small group that believed that the world was about to be destroyed and which we referred to earlier in the chapter (p. 000).

In our investigation of the group which gathered about Dr Armstrong and Marian Keetch, our observers posed as ordinary members who believed as the others did. In short, our investigation was conducted without either the knowledge or the consent of the group members. This situation presented a number of problems that merit detailed discussion . . .
Both of our ‘local’ observers were under pressure at various times in mid-December to quit their jobs and spend all their time with the group . . . Their evasion of these requests and their failure to quit
147

Part 1 PT Introduction to the sociological imagination

their jobs at once were not only embarrassing to them and threatening to their rapport with the group, but also may have had the effect of making the members who had quit their jobs less sure that they had done the right thing. In short, as members, the observers could not be neutral – any action had consequence . . .
Observing, in this study, was exhausting work.
In addition to the strain created by having to play an accepting, passive role vis-à-vis an ideology that aroused constant incredulity, which had to be concealed, observers frequently had to stay in the group for long hours without having an opportunity to record what they had learned . . .
The circumstances of observation made it impossible to make notes openly except on a single occasion, the meeting of November 23, when the Creator ordered that notes be taken. Apart from this, it was difficult to make notes privately or secretly, for the observers were rarely left alone inside the house and it was necessary to be ingenious enough to find excuses for leaving the group temporarily. One device used occasionally was to make notes in the bathroom. This was not entirely satisfactory, however, since too frequent trips there would probably arouse curiosity if not suspicion . . .
Our observers had their daily lives to care for as well as the job, and were subject to occasional bouts of illness or fatigue from lack of sleep. The job was frequently irritating because of the irrelevancies (from the point of view of our main interest) that occupied vast quantities of time during the all-night meetings.
(Festinger et al. 1956: 234–46)

Questions
1. You are a sociologist just starting a research project seeking information on ritualistic abuse. A friend tells you they know of a Satanic cult meeting regularly in the area. Most of their activities are harmless but there is a rumour that an animal is to be sacrificed at Hallowe’en. Your friend tells you they know of a member looking for new recruits. You feel this is a rare opportunity of getting access to the group that may not come up again. How would you continue with your research? (Consider the advantages and disadvantages of alternative methods of research that might be used.)
2. Suggest the practical and ethical consequences of the methods of research that you propose to follow.
148

Problems in researching female factory workers This extract is taken from a study of female workers in an electronics factory in Malaysia. It focuses on the problems faced by these workers, who were termed ‘Minah Karan’
(meaning, roughly, ‘loose women’). Daud’s research was carried out in a number of stages and involved several research methods: participant observation, in-depth interviews, official documents and surveys (based on questionnaires and including a follow-up survey five years after the initial research).
I started my fieldwork as a participant observer, by becoming a factory worker. It began in October 1976 and continued until June 1977. During the first five months I worked in the Variable Resistance Section
(V.R.) and the Electrolytic Capacitor Section, working two-and-a-half months in each section. Then I carried out my in-depth interviews with 100 workers over a period of one month. At the end of the last two months of fieldwork I conducted a survey of 111 respondents to see the changes that had occurred between 1977 and 1982 . . .
Although work in the factory numbed the mind and tired the body, my research was strengthened by real knowledge of the long hours and tough working conditions. I discovered many advantages in becoming a worker in order to understand the real problems and situations of factory life. I could feel for myself the tiredness, depression, tension and other physical hazards; I discovered a lot of truth about the nature of a worker’s life. For instance, cohabitation and illegitimate pregnancies were usually subjects for gossip. However, the workers talked about this only within their own groups. I believe that if I had not undertaken participant observation by becoming a factory worker, such private and personal matters would be very difficult to uncover . . .
During the first few days of my work I had asked them whether it was true, as people claimed, that factory workers cohabit. They replied that it was nonsense. But when they had accepted me as one of them, they told the truth . . .
There were drawbacks in becoming a worker and doing this ‘undercover’ work. Some of them are discussed below: The chief methodological problem was to keep my notes up to date while working normally with others. Each worker sat near another in the ‘line’ and I found there was a general interest in what I did and how I behaved since I was a newcomer. Only when there was a break could

Chapter 3 Sociological research

I write my notes in any detail. By becoming a worker
I had to follow work regulations. Workers were not allowed to talk during work . . .
A few seniors especially the ‘line leaders’ became suspicious and were unhappy with me. They feared that I might be appointed to the supervision post after my ‘training’. They viewed this as unfair since they were senior to me. They resented me and refused to be friendly. It is undeniable that becoming a factory worker was a very good method of carrying out my fieldwork. In not disclosing my real identity, I always had to be careful and be on the alert for rumours and suspicion among the workers . . .
The major advantage of using structured questionnaires is that it produces systematic data on information obtained during observation and in-depth interviews. But, there are also a few disavantages in using this method, namely: I noticed great differences in the quality of information obtained through the methods of participation and descriptive survey . . .
For example, when I asked about their attitudes towards the management, during my period of anonymity, almost all the workers condemned the management and said they hated it. But when questioned during the descriptive survey, half the sample said the managment was ‘good’. I think this is because they were trying to be careful when answering the questionnaires administered by someone who was no longer their fellow-worker.
(Adapted from Daud 1985: 134–41)

Questions
1. What sort of information on factory work would each of the methods adopted by Daud be likely to provide? 2. To what extent can these different methods be seen as (a) valid sources of information; and (b) reliable sources of information?

Activity 2
How sociological is your imagination?
Take out your notebook, sharpen your pencil, discard your most cherished cultural assumptions and suspend good old common sense. You are an outsider and your job is to observe and describe what you see around you.

Choose one type of activity (preferably one that you are not familiar with) and simply record your observations. This may be as simple as observing people in a café or pub (Who are they?
Where do they sit? What do they do?) or it may involve you in an anthropological pursuit of the exotic and the bizarre: a night at the wrestling, an evening at the opera or an afternoon spent watching football or playing bingo (Who are the punters? What do they wear? How do they behave?).
Using your newly found sociological imagination, how do you interpret your observations? Do any patterns emerge? Do any hypotheses suggest themselves? What problems have you encountered in your search for new sociological truths?

Notes for students
This activity is intended to help you to understand the process all observers go through in making sense of the world around them. We all use information previously learned to make sense of situations, and we pick up clues from those we are watching or interacting with. No two people will therefore make sense in exactly the same way. The point of this exercise is to make you aware of what sociologists do when they ‘explain behaviour’, to raise the issues of subjectivity, value and interpretation.
Whether or not you have studied sociology previously, it is hoped that you will be stimulated into thinking differently about the world around you, questioning behaviour that might have been taken for granted.
1. First you must select your observation. Think through some possible choices. Religious and cultural events lend themselves well to this activity, and you may find that local newspapers contain some helpful ideas.
2. You need to decide how you are going to record your observation. Will you keep notes in a book (obviously or unobtrusively)? Will you trust to your memory and write up your impression as soon as possible? Will you use a tape recorder/still camera/video camera? There is no right way in this activity: think about the pros and cons of each recording method.
3. After the observation, note the following. How long did the observation last? Did you know whether the participants were typical? (How would you decide this?) Did anyone explain what they were doing? Were there written instructions (e.g. church prayer book – how much sense did that make?)? Did you make your own sense of what you saw?
4. Let your imagination generate explanations: you are allowed to present competing explanations for what you think you saw and should select the most likely account from your standpoint, explaining in terms of further logical arguments or theories.
149

Part 1 PT Introduction to the sociological imagination

Outline
A suggested outline for the written part of the assignment is as follows: 1. Introduction – Observation chosen: Where? When? Why?
2. Method – How the observation was recorded – pros and cons. Did you remain ‘non-participant’ or did you participate? What were the effects of this? In retrospect, would you change the way you did the activity?
3. Explanations – This section should describe your observation – clues picked up both orally and visually from the participants should be highlighted. You should go beyond pure description and attempt imaginative explanations in order to generate hypotheses about the social behaviour you observed. Factors that may help you in considering explanations and gaining a sociological understanding of what you are observing might include the following: Does there appear to be a shared set of values and rules among the group? Is there a certain social

150

etiquette that must be maintained? What social characteristics strike you about the group? Are they all male? All female?
Predominantly older people? Why? What do the participants appear to get out of the interaction? Escape, pleasure, social contact, sense of identity, etc.?
4. Reflexive account – In this section you reflect on your own role in the observation: if you had been a different type of person, would you have made another interpretation?
It is in this section that you raise issues of objectivity/ subjectivity, value and interpretation. You need to account for your own role in the process of understanding – that is, the interpretation you made of the situation. Are you male/female, white/black, disabled/able-bodied, young/mature, middle class/working class? Do you hold certain religious, moral or political beliefs? How did these factors shape your interpretation and shape the behaviour of the people involved in the interaction?
5. Conclusion – How useful did you find this exercise?

Links: Bryman, A. (2008) Social Research Methods, 2nd edn, Oxford: Oxford University Press. Gilbert, N. (ed.) (2001) Researching Social Life, 2nd edn, London: Sage. Hammersley, M. (ed.) (1993) Social Research: Philosophy, Politics and Practice, London: Sage. Ritchie, J. and Lewis, J. (2003) Qualitative Research Practice, London: Sage Salkind, N.J. (2000) Statistics for People Who (Think They) Hate Statistics, London: Sage. Silverman, D. (2001) Interpreting Qualitative Data, 2nd edn, London: Sage. Stanley, L. and Wise, S. (1993) Breaking Out Again: Feminist Ontology and Epistemology, London: