Doing Social Research – From Research Design to Data Collection, Analysis and Publication

There are several stages of the research process in social research, and the actual data-collection phase is often only a small part of this process. Preparation for data collection and analysis of data post-data collection often take up considerably more time than the actual gathering of data itself.

The process of data collection will of course vary depending on the topic being studied, and the situation of the researcher, but the following stages of research are common to many research studies.

  1. Deciding on a topic to research and narrowing down a field of study
  2. Doing an extensive literature review
  3. Devising research questions and (if desirable) operationalising concepts
  4. Selecting a sample of the population to be studied
  5. Conducting a pilot study
  6. Carrying out the research (gathering data)
  7. Interpreting and analysing one’s findings – thick description versus correlation and causation
  8. Publication, publicity and follow up tasks
  9. Using one’s research data – developing theories and making an impact

The discussion below compares a Positivist and Interpretivist approach to conducting research through these nine stages:

Deciding on a topic area to research

There are many broad topics within Sociology, and many sub-topics within those topics. In a two year A level Sociology course we cover the sociology of the family, education, crime and deviance and global development, and the range of topics under investigation becomes even broader when you get up to university level.

As a general rule Sociologists tend to focus on just one broad subject area and within that topic area they specialise in just one sub-topic – For example Becky Francis has tended to specialise in researching the relationship between social class, gender and identity within education, while Tony Sewell has tended to focus on the experience of black boys in the education system.

What are the factors which influence a sociologist’s choice of research topic?

  • The personal interests and values of the researchers themselves. A Sociologist is obviously going to be more motivated to study something they are interested in – and nothing motivates quite like personal experience – Tony Sewell and Marc McCormack are two examples of Sociologists who studied groups with whom they shared personal characteristics.
  • Theoretical perspective/ political beliefs. Whether one is a Feminist, Marxist or Functionalist/ New Right Thinker/ or Post-Modernist can influence what one studies. Feminists emphasise the importance of focussing on issues of gender inequality, so might choose to research issues such as domestic violence or the impact of the Beauty Myth, while Marxists focus on researching the impact of wealth inequalities, so might research things such as class inequalities in education. All of this raises the question of whether Sociology can remain value-free (unbiased)
  • Opportunity also matters when it comes to research topic – Mac An Ghaill wanted to study the experiences of Irish students but he couldn’t study, so instead he focussed on the black and Asian students in his own college.
  • Funding – Sociologists are professionals and need get funding for their research, so funding bodies can influence topics of research.
  • Society – Societies change, and so new topics of study will emerge with social changes. For example, sociologists have studied things such as rave culture, and virtual gaming communities as these have emerged, which overlaps with the first point above!

Doing an extensive literature review of existing research

Before undertaking research it is customary to do an extensive review of existing research relevant to one’s topic. There are several reasons for doing this:

  • To make sure no one else has already done what you’re about to do
  • So you can locate your research in wider research and develop already existing knowledge and theories
  • To find ideas about questions to ask respondents
  • To find ideas about how to go about conducting the research
  • To uncover any potential problems you may encounter during the research process.

This links to secondary qualitative and quantitative data.

Devising research questions and (if desirable) operationalising concepts

A Positivist approach to devising research questions

Broadly speaking Positivists aim to conduct social research using the methods of the natural sciences – which means using methods such as social surveys so that data can be easily quantified and correlations between variables uncovered.

Positivists will typically start off with a relatively narrow research agenda, seeking to find how a handful of independent variables effect a dependent variable. For example, when researching differential educational achievement, positivists might be interested in seeking to find what aspects of home background are most closely correlated with educational success or failure.

This means that Positivists need to think up their questions in advance, In order to successfully address the above question, you need a whole range of sub-questions to make up the survey – you need to ask about educational achievement, social characteristics such as class, gender and ethnicity and your need to ask about a whole range of things which may have affected achievement – level of parental interest, income of parents and so on…. If you don’t put it in the questionnaire, then you can’t find out about it! In Positivist research operationalising concepts in advance of data collection is crucial.

Operationalising concepts

Operationalising a concept means translating abstract concepts into specific questions that can be understood and measured in practice.

In the above example, almost everything has to be operationalised – you can’t ask about class, for exmaple, you need to ask about occupation, and you can’t ask about ‘parental interest’, you have to think of other questions that would measure this.

A Interpretivist approach to designing research questions

Given that Interpretivists are interested in qualitative data collection, and want to here respondents talk in their own words, all they need to do in advance of the research is have an idea of the kind of questions they want to ask, and some sensitive ways of prompting for further information. Essentially, Interpretivists don’t start off with specific research questions, they start off with a general aim and a few starter questions and then ask further questions as the research evolves.

Selecting a Sample

Sampling is the process of selection a section of the population to take part in social research. Key terms associated with sampling include:

The Target Population – All people who could potentially be studied as part of the research

The sampling Frame – A list from which the sample will be drawn

The research sample – The actual population selected for the research – also known as the respondents.

A sample is said to be representative if the characteristics of the sample reflect the characteristics of the target population.

There are many different types of sampling technique which you need to know about, including Random sampling

  • Random sampling
  • Systematic sampling
  • Stratified random sampling
  • Quota sampling
  • Multistage sampling
  • Snowball sampling

Doing a Pilot Study

A pilot study is a test study done in advance of the actual study. Some of the advantages of doing a pilot study include:

  • To find out any practical difficulties you might come across before the actual research.
  • To see if the questions you are asking make sense to respondents
  • To see if response rates differ between different groups (in which case you might need a ‘booster’ sample of under-represented groups
  • To familiarise yourself with respondents so you feel more at ease when doing the actual research.

Pilot studies are easier with more quantitative methods, and may not be possible with more in-depth, qualitative research.

Conducting the research/ gathering data

Once the researcher has gone through all of the above stages, they are finally ready to collect data, where all of the practical, ethical and theoretical issues discussed in the previous pages apply.

For positivist inspired quantitative researchers this is a relatively easy phase of the research process. With some survey based research, for example, researchers don’t need to interact with respondents, and all you need do is to keep an eye on response rates, and maybe prompt certain groups to respond as necessary.

Obviously this is going to more difficult with qualitative research where the researcher is more involved with the respondents – here the researcher needs to think about how to record data – interviews need to be taped for example, and with observational studies, a field diary is often used to keep track of observations.

Analysing the Data

With Positivist research where researchers have used closed questions such as yes/ no answers and Likehert scales which are be pre-coded, the questionnaires can be fed straight into a computer which will read and analyse the data, turning it into statistics automatically . This can then be presented in the form of a bar chart or graph so we can easily see the relationship between such things as class and educational achievement.

Interpretivist research which is qualitative, and may amount to thousands of sides of notes may take much longer to analyse after data has been collected and cannot so easily be translated into statistics.

Stages 8 and 9: Publication and following up research

During the final write up of the research it is usual for Sociologists to comment on whether their data supports or refutes existing theories, hence relating their research back to wider theoretical debates (structure/ agency etc.)

Following research, researchers also need to think about how much effort they are going to put into feeding back to respondents and taking their views on board, which could be challenging in Interpretivist research where some of the respondents may not agree with the information the sociologist has selected for publication.

There is also the issue of how the data is to be used – Should Sociologists publicize their findings to the broader public, or should they leave this to their employers and the media? Should Sociologists get involved in government policy, or leave this to public officials

Find out more

If you like this sort of thing then you can find out more at my page on research methods.

%d bloggers like this: