Types of data and evidence

Last time

Before we start: last time, in How sociologists investigate society, you learned these points. Here they are again on one card. Keep them in mind, because today builds on them.

Numbers and words

Sociologists use two kinds of data: numbers, such as how many marriages end in divorce, and words, such as how a divorced person felt. Each kind can be collected first-hand, or found in sources others have made. This lesson covers the types of data and how to judge them.

Primary and secondary data

Primary data is collected by the researcher themselves, through questionnaires, interviews or observation. Secondary data already exists, collected by others: official statistics, newspapers, diaries, letters, and earlier research. Secondary data saves time and money, but may not fit the researcher's exact question.

Quantitative and qualitative

Quantitative data is numerical: statistics, percentages, tables. It shows patterns and trends. Qualitative data is descriptive: words, feelings, experiences. It shows meanings and reasons. Positivists prefer quantitative data; interpretivists prefer qualitative.

Official statistics

Official statistics are collected by governments: crime rates, exam results, births, marriages and divorces. Strengths: large, national, cheap to use, and good for comparing over time. Problems: they can be inaccurate or incomplete, such as unreported crime, and definitions can change over time.

Documents

Personal documents, such as diaries and letters, give rich, qualitative insight into people's lives. Public documents include reports from schools, courts and governments. Documents may be biased, written for an audience, or hard to check.

Judging evidence

Reliability: would the same research give the same results if repeated? Validity: does it give a true picture of what it claims to measure? Representativeness: is the sample typical of the population? Generalisability: can the findings apply to the wider population?

Bias

Bias means the evidence is one-sided or distorted. It can come from the researcher's values, leading questions, an unrepresentative sample, or a source written to persuade. Good research reduces bias by careful sampling, neutral questions and triangulation.

Using a source

Source questions ask what the data shows, or the problems with it. Look at who collected it, when, how large the sample was, and what it leaves out. For example, data on gender inequality from one country, or one year, may not represent the wider picture.

Data types compared

The main kinds of data, with a strength and a weakness each.

Mistakes that lose marks

Mistake one: calling official statistics primary data for a sociologist. They are secondary. Mistake two: saying data is unreliable when you mean it is not valid. Mistake three: in a source question, giving general points not linked to the source. The fix: the right term, and evidence from the source.

The full-marks tip

The full-marks tip: for any data, judge it with reliability, validity, representativeness and bias. In a source question, quote or refer to the source for each point.

On one card

So, on one card: data and evidence. Keep it for revision.