Introduction to research methods

You are here

Introduction to research methods

'Be a good craftsman; avoid any rigid set of procedures. Above all seek to develop and use the sociological imagination. Avoid fetishism of method and technique...Let every man be his own methodologist.'

(C.W. Mills)

Clearly method has, to some extent, to be tailored to the requirements and constraints of the research. For example, a sociologist examining important changes over time is forced to use whatever historical material that could be of use. Indeed it should be emphasized that there is no such thing as a bad method, but there is certainly inappropriate use of particular methods. Methods should not be forced into uses for which they were not designed or intended.

The more general a statement the sociologists wishes to make the more s/he will be forced into using a large sample, or statistics gathered from a large group. Data is then more likely to be collected and presented in a more structured way. If however the attempt to understand the meaning attached to situations by the participants is the chief concern there will be more informal methods used and the sample will be smaller.


  • Should you collect data yourself, or use existing data?
  • Should you try to experience the phenomenon under investigation directly, or use oral or written accounts from actors concerned?
  • To what extent should you structure the situation under investigation, both by your own contribution to it, or in the possible responses of your subjects?


1. Research that relies on other peoples' evidence. The Documentary method (census).

old man

2. Research, which seeks answers to particular questions from fairly large groups of people with fairly short, often very short, periods of contact. The survey method (for example, The symmetrical family) .

3. Research which involves the sociologists working very closely among those s/he studies, often for quite long periods of time. This is the observation method. Sometimes the researcher participates in the lives of those observed. This then becomes Participant observation (for example, the Glasgow Gang) .

4. Research which involves carefully chosen groups of people who are placed in different situations to see what happens. The experimental method (for example, the Hawthorne Experiment).

Whatever the approach, the following qualifications apply:

  1. Good sociology is based on good (reliable/valid) evidence.
  2. We need to be able to distinguish between good and bad evidence (evaluation).
  3. Differences of opinion (argument) should be based on evidence.
  4. Evidence has to be collected, this requires empirical research and can involve a variety of methods.

The following concepts, if correctly identified and used, are virtually guaranteed to get you several marks in any question concerning methods:

Reliability: Could anybody, using the same method, come up with the same results? Some methods are seen as unreliable (less objective, more subjective).

Validity: Does the material give a true picture? Beliefs and actions (saying and doing) are not always consistent. For example, in surveys, peoples' answers are answers to specific questions, but what if the questions are misunderstood, or ask relatively unimportant questions? Similarly, in a lab experiment how can we be sure that the behaviour identified is typical?

Representativeness: Is the situation typical? If so, generalisations are possible.

Directly collected: Oral: Verbal:
Pre-collected: Official statistics Radio Official
T.V. Diaries
Collected by researcher: Participant observation Talks Diaries kept for researcher
Systematic observation Open-ended interviews Open-ended questionnaires
Experiments Pre-coded structured interviews questionnaires

Scientific sociologists will tend to favour more structured situations for data collection and prefer written and oral responses that are more controllable. They will favour the bottom right of the diagram. Sociologists using an action framework will tend to stress more informal methods of data collection, the top left of the diagram - concerned with data collected by the researcher.

We can also relate the type of method likely to be involved in research by asking two questions:

How many people are being investigated?

How much personal involvement from the researcher is required?

These two points work against each other, for example, the more people involved in research the less likely it is that a researcher can know each person well.

  1. Numbers involved
  2. Many
  3. Social surveys
  4. Structured interviews
  5. Unstructured interviews
  6. Observation
  7. Participant observation
  8. Few
  9. Low High
  10. Personal involvement of researcher

Source: Worsley (1977:89)

Some research aims only to describe, some wants to explain. Be wary, however, this is a very blurred distinction since explanation requires description, and description is, in itself, explanation. Try for yourself to make a distinction between describing a door and explaining a door!


Some research is called action research. This is usually undertaken to monitor some reform, or policy change, once it has been introduced. This is done to see if the hoped for results have in fact been achieved. For example, The Plowden Report 1967 recommended positive discrimination in primary schools. Educational Priority Areas were set up which received extra funding, and they were monitored.

There are innumerable reasons for doing research, some less noble than others. Here are some of the main reasons:

  1. The interests and values of the researcher - for example, Townsend (Poverty, 1957-79).
  2. Current debates - for example, Goldthorpe (embourgoisement 1969) Gavron (feminist research 1966).
  3. Attracting funding. Generally it is easier to get funding for explanatory (policy research) than for pure academic research, and easier for statistical than for qualitative research.
  4. Access. Some groups can resist intrusion into their affairs, hence we know more about the poor than we do about the rich.
  5. To get academic acclamation and a rise in status.

Finally, a caution! Written research has generally had, certainly until fairly recently, all the messy bits taken out. Research is rarely as straightforward and unproblematic as the final report makes it seem.

Quite apart from theoretical considerations, there are practical issues that can affect the method of research chosen. Among the more important of these are: time; labour; money; topic; access.