Sociology: Choice of method and the research process


Positivism versus interpretivism

Positivists argue that there is a measurable, objective social reality that exists ‘out there’, just like the physical world. They see our behaviour as the result of social forces shaping what we do, and the aim of research is to discover the underlying causes or our behaviour

They use standardised methods of research, such as questionnaires, structured interviews, structures observation and official statistics. This enables them to obtain reliable and representative quantitative data

They use this data to identify general patterns and trends in behaviour, from which they produce cause-and-effect explanations like those in the natural sciences 

1 of 26

Positivism versus interpretivism 2

Interpretivists claim that there is no objective social reality, just the subjective meanings that social actors give to event. Therefore, the aim of research is to uncover actors’ meanings or worldview

For interpretivists, this means using open-ended research methods that produce valid, qualitative data, such as unstructured interviews, participant observation and personal documents

Such methods enable the sociologists to gain understanding by experiencing the group’s lifestyle for themselves, or by allowing individuals to explain their worldview in their own words, without the sociologist imposing their own views on the research subjects

2 of 26


The relationship between sociological theory and research methods is very important. The sociologist’s position affects their choice of research method, the kind of questions asked, the type of data collected, and whether the reliability and representativeness of data, or its validity, is seen as more important 

3 of 26

Three key concepts

Sociologists use three key concepts to judge the usefulness of a research method. These are reliability, representativeness and validity

4 of 26


For a method to be reliable, it must be reliable, i.e. exactly repeatable to obtain the same results, regardless of who actually carries out the research. Reliability also means using standardised forms of measurement. A reliable method creates data that can be used to systematically re-test hypotheses about social behaviour

Positivists favour a scientific approach emphasising the need for reliability and therefore they use structured research methods that can be repeated, such as experiments, questionnaires and structured interviews 

Positivists regard participant observation and unstructured interviews as unreliable because they cannot be repeated and do not use a standardised system of measurement

5 of 26


Sociologists cannot usually study every member of the group they are interested in because generally there are simply too many of them. Therefore researchers may choose to study only a sample- a smaller sub-group drawn from the wider target group

To be representative, the characteristics of the sample need to be the same as those of the wider group. This allows the researcher to be more confident that what is true for the sample is probably also true for the whole group 

This means they will be able to make generalisations on the basis of evidence from the sample 

Positivists emphasises the importance of representativeness, because they wish to discover general pattern and make general cause-and-effect statements about social behaviour

6 of 26


Validity refers to how authentic and true  the data is. The aim of any research is to ‘get close of the reality’ of a social situation

Interpretivists emphasise the need to use methods such as participant observations or unstructured interviews which reveal the meanings people hold. surveys, experiments and other structured research techniques are rejected because they do not reveal what social actors really think or how they act   

7 of 26

Primary and secondary data

Primary data is evidence collected by sociologists themselves for their own sociological purposes 

thi s is material collected firsthand by researchers using methods such as questionnaires, observation and interviews 

Secondary data is any information that has already been collected by someone else for their own purposes and that may then be used by the sociologist 

Official statistics business records, media reports, diaries and personal documents are common forms or secondary data- as are the findings from existing sociological research that a later sociologist may go on to use    

8 of 26

Choice of research method

The main factors influence the sociologist’s choice of research method: the methodological preference of the sociologist, practical aspects or research, and ethical concern

9 of 26


Right at the start of an answer on choice of method, point out that methodological issues are the first consideration. Practical and ethical issues then tend to limit the sociologist’s choice

10 of 26

Methodological preference

Positivists prefer quantitative data. In their view, the aim of research is to reveal cause-and-effect relationships. This requires quantitative data to identify patterns and trends in behaviour

Positivists thus prefer to use structured research techniques, e.g. questionnaires, which generate reliable, representative data

Interpretivists prefer qualitative data. In their view, the aim of research is to uncover the meanings people hold. The only way to do this is to allow them to act or speak in the ways they feel are appropriate 

This is best achieved by unstructured research methods such as participant observation and unstructured interviews, which produce valid, qualitative data

11 of 26

Practical factors

Practical factors restrict the sociologist’s choice of method. These include:

Time Some methods usually take more time, e.g. participant observation studies and unstructured interviews take more time than social surveys

Finance The finance available affects the number of researchers, respondents and the amount of research time 

Some methods are cheaper than others, e.g. postal questionnaires are cheap because researchers do not have to spend long periods of time talking to respondents or observing their behaviour 

Source of funding research sponsored by government, businesses, voluntary organisations etc. reflects the concerns of these funding bodies. It is often easier to get funding for quantitative research  

12 of 26

Practical factors 2

Personal factors Researchers have careers, family commitments etc, so they may not be able to do lengthy research in the field

Research subjects Some groups are less open. E.g. criminals, so structured research methods are not appropriate 

Research opportunity if a research opportunity suddenly appears, the researcher may have no time to prepare lengthy questionnaires or interview schedules

Personal danger Methods involving direct contact with a research group increase the possibility of danger to researchers   

13 of 26

Ethical factors

Sociological research does not take place in a moral vacuum; sociologists have to think about the possible effects their research might have on people’s lives 

Informed consent Researchers should have informed consent of research subjects because of the ethics that the research may have on them. People should not be manipulated or misled about the research 

Confidentiality Research subjects have a right to anonymity, so they should not be identifiable when research is published. This is difficult to achieve with small groups- even changing names may not be enough for anonymity 

14 of 26

Ethical factors 2

Effects on research subjects Research findings can be used by political groups or the media in ways that may damage the research subjects 

Vulnerable groups Special care should be taken when investigating groups that are vulnerable because of age, gender, disability etc. Additional consents are often required 

Covert research there is a tension between the ethical problem of deception when withholding knowledge of the of the researcher’s real identity and the difficulty of gaining access to groups that do not wish to be investigated

15 of 26


Theoretical factors are a positive influence because they are about the kind of data the sociologists prefers to have- quantitative or qualitative 

Practical and ethical factors are more of a limitation on choice of method 

Practical, ethical and theoretical concerns are often interrelated. For example, collecting qualitative data produces practical problems such as gaining trust, access etc, while gathering quantitative data creates practical problems such as sampling frames, the geographical distribution of a sample, question design etc

Triangulation Some sociologists see advantages in both types of data. Triangulation combines quantitative and qualitative methods so that the strengths of one balance the weakness of the other

16 of 26

Choice of topic

The sociologist’s choice of topic is affected by several factors, including:

Practical factors Some topics may not easily be studied, e.g. high-level political decision-making may be inaccessible to the sociologist

Funding bodies These have enormous influence because they will only fund studies of topics that they consider to be important

17 of 26

Choice of topic 2

Governments are much more likely to fund research that links to their policies 

Society’s values change and the interest in particular topics and issues moves with them 

The sociologist’s theoretical perspective may affect whether or not they will become involved in studying a particular topic- e.g. feminists are likely to study gender issues 

Chance Sometimes, sociologists find themselves in a potential research situation by pure chance- e.g. hospitalisation as a result of illness gave one researcher the opportunity to do a study of a hospital ward 

18 of 26

The process of research

Once we have chosen a topic and a research method, we need to formulate an aim or hypothesis 

Aims Most studies either have a general aim or seek to test a specific hypothesis. An aim is a statement that identifies what a sociologist intends to study. Often the aim is simply to collect data on a particular topic, for example, people’s leisure patterns 

Hypotheses Other studies seek to test one or more hypotheses. A hypothesis is more specific than an aim. It is a possible explanation that can be tested by collecting evidence to prove it true or false, e.g. that educational achievement is affected by gender  

19 of 26

The process of research 2

Operationalising concepts Before research can begin, the researcher needs to define their sociological concepts or ideas in ways that can be measured 

This process of converting a concept into something measurable is called ‘operationalisation’. For example, ‘educational achievement’ might be defined as having passed five GCSEs at grades A*-C. However, when different sociologists operationalise the same concept differently, it becomes harder to compare their findings 

The pilot study If the researcher is using a survey method, the next stage is to produce a draft questionnaire or interview schedule and to give this a trial run 

This is called a pilot study and its aim is to iron out any problems, refine or clarify questions and their wording, give interviewers practice etc

20 of 26


It is often impractical to study all the members of the research population or ‘target group’ that we are interested in, e.g. because of its sheer size, lack of time, money or other resources 

Therefore researchers may only be able to a study a sample of it. A sample is a smaller part of the whole research population that the sociologist selects for study 

Sampling frames To select a sample of the research population, the sociologist first finds or creates a sampling frame- a list of all the members of the research population from which the sample can be chosen. For example, the electoral register is a list of all those people who are entitled to vote

21 of 26

Samples and representativeness

  • To be representative, a sample should have the same characteristics, in the same proportions, as the wider research population. It should be a cross-section of the whole group 

  • If the sample is a representative cross-section, then what is true of the sample is likely to be true of the whole group 

  • Representativeness is important to positivists because they want to make generalisation and discover general laws of social behaviour                         

22 of 26

Are all samples representative?

  • Small samples are less likely to be representative of large populations 

  • Interpretivists are more interested in the meanings held by social actors. Because they are not trying to establish ‘laws’ of social behaviour that might apply to large social groups, they feel it is less important to use representative samples

  • If the researcher does not have a sampling frame that includes all members of the research population, they cannot create a representative sample 

23 of 26

Types of sampling

Sociologists use several types of sampling:

Random sampling This is where every member of the sampling frame has an equal chance of being selected. This eliminates bias in simple selection 

A large enough random sample should reflect the characteristics of the whole research population. However, not all random samples are large enough to ensure this happens 

Systematic or quasi-random sampling Some sociologists introduce an element of structure into sampling by selecting every nth person in the sampling frame- e.g. every tenth name in the list 

Thus can reduce the chance of a skewed (biased) sample being randomly selected

24 of 26

Types of sampling 2

Strained sampling The researcher first stratifies (breaks down) the population by age, class, gender etc. The sample is then created in the same proportions, e.g. if 20% of the population are under 16, then 20% if the sample also have to be under 16 

Quota sampling The population is stratified as above, and then each interviewer is given a quota of say, twenty females, ten of whom are aged 60 or over etc, which they have to fill with respondents who fit these characteristics. This interviewer keeps at this task until their quota is filled 

25 of 26


With some social groups, e.g. professional criminals, it may be difficult to generate a sample. Sociologists therefore sometimes use a snowball sample, where one member of the group puts them in touch with another member and so on. however , such samples are unlikely to representative

26 of 26


No comments have yet been made

Similar Sociology resources:

See all Sociology resources »See all Sociological research methods resources »