Research Design, Qualitative and Quantitative Methods

The method decides the data

Sociologists study people, which makes their research harder than it looks. You cannot put a society in a laboratory, and everything you might want to know has to be got from people who may not tell you the truth, may not know, or may behave differently because they are being studied. So a sociologist designing a study makes a series of choices, and the thing to understand about all of them is that none of them is simply right. Every choice is a trade-off, and the biggest one is the method. A questionnaire sent to a large number of people produces answers you can count, compare and turn into figures, but only about the things you thought to ask, in the form you asked them. An unstructured interview with a handful of people produces depth you could never get from a form - the reasons, the hesitations, the things you did not know to ask - but you cannot count it, and a handful of people may not represent anybody else. Neither of those is better research. They answer different questions, and that is why your specification names the mixed-methods approach, where more than one is used together. ⚠️ The consequence runs through this whole topic and it is what exam questions are really testing: the question "which method should they use?" is answered from the AIM, never from a list of strengths. Decide what you need to find out, and the method follows. ⚠️ One thing this module deliberately leaves out. Every method also raises practical questions of time, cost and access, and ethical ones about consent, confidentiality and harm to the people being studied. Those are studied in the next module on this topic, and they matter enough to deserve it: here we stay with the design and the methods themselves.

Words for designing a study

Six terms. The first two are what a study is trying to do; the last two are where the data came from.

Order the design decisions

Put the stages of designing and carrying out a piece of research into a sensible order.

• Decide the aim: what the research is trying to find out
• Turn it into a hypothesis: a specific statement the research can test
• Choose a method suited to that aim, and a sampling method for choosing participants
• Run a pilot study to find problems while they can still be fixed
• Carry out the main research and collect the data
• Analyse the data and say whether it supports the hypothesis

Trying it out first

A researcher runs a pilot study before their main research. What is it for?

• To find problems with the design while they can still be fixed, such as questions that are unclear or produce useless answers
• To get early results so the researcher can report something sooner
• To increase the sample size of the main study
• To prove the hypothesis before the main research begins

Numbers, or accounts

The division your specification builds this topic on. Read the last row of each column, because that is what actually decides which one a study needs.

Match each method to the data it gives

Which two are strengths of questionnaires

Select the TWO genuine strengths of using a questionnaire with closed questions.

• It can reach a large number of people relatively quickly
• The answers come in fixed categories, so they can be counted and compared directly
• It gives detailed reasons for why people answered as they did
• It guarantees that everybody answers honestly

Designing a study in words

A study begins with an _____, stating what the researcher wants to find out, which is turned into a _____ that the research can test. A small trial run carried out first is called a _____ study. Data the researcher collects themselves is _____ data, while data collected by somebody else for their own purpose is _____ data. Using more than one method in the same study is the mixed-methods approach.

Collected by you, or found

Your specification asks for two further distinctions, and they are about where the data came from rather than what form it takes. Primary data is collected by the researcher themselves, for their own study. Its advantage is fit: it answers exactly the question being asked, because it was gathered to do so. Its cost is that gathering it takes time, effort and access to people. Secondary data was collected by somebody else, for their own purpose, and the researcher uses it afterwards. Its advantage is that it already exists, often covering far more people and a longer period than one researcher could ever manage. Its cost is fit: it was not collected to answer your question, so it may not contain quite what you need, and you did not control how it was gathered. Within secondary data, your specification separates official from non-official statistics. Official statistics are produced by government and public bodies. They are usually large-scale, regularly collected, and available over long periods, which makes comparison across years possible. But they record what the state chose to count, in the categories it chose, and a change in how something is recorded can look like a change in the thing itself. Non-official statistics come from other organisations such as charities, businesses and researchers. They may cover things official figures ignore, which is often exactly why they exist, though the organisation collecting them may have a position of its own. ⚠️ The question to ask of any secondary source is the one that runs through this course: who collected this, and what for?

Choosing from the aim

Exam questions on this topic usually describe a piece of research and ask which method should be used, or why one was chosen. The answer always comes from the aim, and here are two aims that lead to opposite methods. Aim one: to find out how many students in a school do part-time work. This asks HOW MANY, so it needs data that can be counted across a large group. A questionnaire with closed questions suits it: it reaches every year group quickly, the answers come in fixed categories that can be totalled, and the numbers can be compared between groups. The cost is accepted openly: it will not tell you why anybody works, or what the work is like, and it only covers what was asked. Aim two: to find out what part-time work is like for the students who do it. This asks WHAT IS IT LIKE, so counting will not help. Unstructured interviews suit it: the researcher can follow whatever each student raises, including things nobody thought to ask about. The cost, again stated rather than hidden: only a few students can be interviewed, so the findings may not represent others, and the answers cannot be turned into figures. Notice the shape of both answers. Name the aim and what kind of thing it asks for. Choose the method that produces that kind of data. Then concede what the choice gives up, because a method with no weaknesses does not exist. ⚠️ And a third answer is often available: mixed methods, using a questionnaire to find out how common something is and interviews to find out what it is like. Suggesting that, with a reason, is usually the strongest response of all.

The method picked before the question

Four answers to "which method should this researcher use?". Select the ONE that chooses a method without reference to the aim.

• A questionnaire, because the aim asks how common something is and closed answers can be counted across a large group.
• Unstructured interviews are the best method because they give the most detailed and interesting data.
• Participant observation, because the aim is to find out what daily life in the group is actually like.
• Mixed methods, using a questionnaire for the scale and interviews for the reasons behind it.

Justify the method

Assemble an answer in the form these questions reward.

Designing it for a real question

A sociologist wants to study how students choose their subjects. Work through the design decisions.

• The aim is to find out which factors students say influenced them most. What comes first?
• They want figures they can compare between year groups. Which method suits that?
• The pilot study shows several questions were understood differently by different students. What now?
• They now also want to know what the decision felt like for students. What do you suggest?

Explain research design and methods

Explain how a sociologist designs a piece of research, and explain the difference between qualitative and quantitative methods and data. Use an aim of your own choosing.

• Explain the difference between an aim and a hypothesis
• Explain what a pilot study is for, and why it is done before the main research
• Explain the difference between quantitative and qualitative data
• Describe at least three methods and say what kind of data each produces
• Explain the difference between primary and secondary data
• Explain what official statistics are, and one limitation of using them
• Choose a method for your aim and justify it from the aim rather than from a list of strengths
• Concede what your chosen method gives up, and say whether mixed methods would help