Sociological Research Methods

Representative of whom

Sociologists cannot study everybody, so they study some people and then say something about many more. Everything that can go wrong in research is hiding in that move. You will often read that a sample was representative, or that it was not. Taken on its own the word means nothing. ⭐ A sample is never representative full stop. It is representative OF a group that somebody has to name. A hundred people can be a fine sample for one question and a useless sample for another, without a single thing about them changing. ⚠️ And the size of a sample cannot rescue it. If the list you drew from already left a group out, asking more people from that same list makes you precisely wrong instead of roughly wrong.

Two places data comes from

Before the sampling question bites, one simple split. Sociologists either go and get the data themselves or use data somebody else already collected. Both kinds are used in every part of this course, and both raise the same question about who is in them.

Order the design decisions

A sociologist plans a study in this order. Arrange the five decisions.

• Decide the research question
• Name the exact group the findings are meant to describe
• Find or build a list of that group that can actually be reached
• Choose the method and the sample from that list
• Report the findings together with who was not reached

The supermarket survey

A researcher wants to know what adults in a town think about local bus services. She surveys two thousand adults leaving one supermarket on a Tuesday morning. What is the main problem with her study?

• Everybody at work on a Tuesday morning had no chance of being asked, so the findings describe a narrower group than the one she named
• Two thousand people is too small a sample for a whole town
• She should have asked more questions about the buses
• She collected primary data when official statistics would have been more accurate

Words for judging a study

Five words that let you say exactly what is wrong with a piece of research, instead of saying it was unfair.

Find the sentence about who is missing

Here is a short description of an invented study. Select the ONE sentence that tells you who the findings cannot describe.

• A sociologist wanted to know how people in a city use public libraries.
• She wrote a questionnaire of twelve questions and checked that it was clear.
• She handed it to people as they came out of three libraries on a weekday afternoon.
• Most of those who filled it in said the libraries were important to them.
• She concluded that people in the city value their libraries.

What threatens representativeness

Three of these five make a sample less representative of the named population. Choose the three.

• The list the sample was drawn from leaves out part of the group being studied
• Many of those chosen did not reply, and the ones who did had strong feelings about the topic
• Everybody was approached in one place at one time of day
• The questionnaire took twenty minutes rather than five to complete
• One question was worded in a way that pushed people towards agreeing

Two questions that evaluate anything

Whenever you are given a study, a survey or a single figure, ask two questions in this order. ⭐ Representative of whom? ⭐ And who is missing? Those two questions will earn you more marks than any amount of writing that the sample was too small. They work on official statistics as well: a figure counted by an official body still only counts the people who came to its attention.

Match the study to the group it misses

The queue at the returns desk

Forget sociology for a moment. A shop wants to know what its customers think of a new coat it has started selling. The manager has an idea that costs nothing. Every afternoon there is a queue at the returns desk. Those people have bought the coat, they are standing right there, and they are very willing to talk. ⚠️ So he asks them, and the answers are dreadful. Poor stitching. Wrong size. Came apart at the shoulder. ⚠️⚠️ He pulls the coat from the shelves, and the shop loses its best seller of the season. ⭐ What went wrong is not that he asked too few people. He could have asked four hundred and got the same answer, because standing in that queue is the one thing every person he spoke to had in common, and it is precisely the thing that made them unhappy. Everybody who wore the coat all winter and loved it had no reason to be in the shop at all. ⭐⭐ The people you can most easily reach are almost never a fair picture of the people you want to know about, and the easier they were to reach, the more suspicious you should be about what they have in common. That is the whole of this module, in a shop.

Complete the research methods facts

Before collecting anything, a sociologist names the _____, the whole group the findings are meant to describe. The sample is then drawn from a list that can actually be reached, and if that list leaves a group out, the sample cannot be _____ of the group that was named, however large it is. A study is _____ if repeating it the same way would give the same results, and it is _____ if it really measures what it claims to measure. Two things then decide who is genuinely in the study: the _____ rate, meaning how many of those approached took part, and the _____, meaning the duty to gain informed consent, which some people will refuse.

Build the evaluation sentence

Complete the evaluation of an invented study in which questionnaires were left in one doctor waiting room for a week.

Find the research methods words

Six words from this module are hidden in the grid. Each clue describes one of them.

Methods quick-fire

Five quick decisions. Three lives.

Design a study that can be defended

You have been asked to find out why some pupils at a large school stop attending after-school clubs. Work through the decisions.

• First, name the group your findings should describe. Which is the right one?
• You want to know WHY they stopped, in their own words. Which method fits, and why?
• You ask club leaders to pass your request to pupils who left. Half of those asked do not reply. What do you do?

Evaluate a piece of research

A researcher stood outside one railway station on a Monday morning and asked four hundred people whether they were satisfied with transport in their area. She reported that people in the area are satisfied with transport. Evaluate her study.

• Name the group her conclusion claims to describe, and the group she actually reached
• Explain why four hundred is not the point, using the idea of the list she drew from
• Name one group of people who could not have been asked, and say why
• Use at least two of these words correctly: representativeness, reliability, validity
• Finish by saying one thing she could change, and what it would fix