The strengths and limitations of secondary data

Last Updated on October 11, 2026 by Karl Thompson

Secondary data is information that someone other than the researcher has already collected. In brief:

1. Secondary data can be quantitative (official statistics, existing survey datasets) or qualitative (documents such as diaries, letters, newspapers, reports and websites).
2. Its main strengths are that there is a huge amount of it, it is often free and quick to use, it may be the only way of studying the past, it allows comparisons over time and between groups, and it involves no direct contact with the people being studied.
3. Its main limitations are that it was produced for someone else’s purposes, the researcher has no control over how it was collected, definitions and measurements change over time, it may reflect the biases of whoever produced it, and documents can lack authenticity and representativeness.

Secondary data is part of the research methods topic in AQA A-level sociology. It comes up mainly in the theory and methods section of Paper 3, and also on Paper 1, in both the theory and methods question and the methods in context question. You can find more on the other methods on my research methods page.

What is secondary data?

Secondary data is data that has been collected previously, by someone other than the researcher, usually for a different purpose. A researcher who analyses it is doing secondary research.

The contrast is with primary data, which the researcher collects themselves for their own research, through methods such as questionnaires, interviews, observation or experiments.

The distinction is about who collected the data, not about what form it takes. That is why secondary data can be either quantitative or qualitative.

Is secondary data qualitative or quantitative?

It can be either. The table below shows the main types you need to know.

Type	Examples	More detail
Quantitative: official statistics	Crime statistics, exam results, the Census, marriage and divorce rates	Official statistics in sociology
Quantitative: existing survey datasets	The Crime Survey for England and Wales, British Social Attitudes	Official statistics in sociology
Qualitative: public documents	Government reports, Ofsted reports, school websites, parish records	Secondary qualitative data
Qualitative: personal documents	Diaries, letters, autobiographies, social media posts	Personal documents
Qualitative: mass media	Newspaper articles, TV news, advertising, online content	Content analysis of the mass media

Official statistics are the most common form of quantitative secondary data, and they have their own post. This post covers the strengths and limitations of secondary data in general, across both types.

Strengths of secondary data

There is a lot of it

Secondary data is the richest vein of information available to researchers in many topic areas. Governments produce statistics on almost every area of social life, and organisations, the media and ordinary people produce documents constantly. Some large datasets simply would not exist if the government didn’t collect them, because no individual researcher could afford to survey the whole population.

It is often cheap and quick to use

Many official statistics and public documents are freely available, often online. The UK Data Service, for example, gives researchers access to the datasets from major national surveys. Using data that already exists saves the researcher the time and money of designing a study, finding a sample and collecting the data. That leaves more time for analysis.

It may be the only way of studying the past

A researcher can’t interview people who are no longer alive, so documents and old statistics may be the only evidence of how people lived. For example, Peter Laslett (1965) used parish records to argue that nuclear families were common in pre-industrial England, which challenged the idea that industrialisation created the nuclear family.

It allows comparisons over time and between groups

Official statistics are collected in a similar way year after year, which makes them useful for identifying trends. The Census in England and Wales was first taken in 1801 and has been repeated every ten years since, apart from 1941 (ONS).

Large datasets also allow comparisons between social groups, such as social classes, genders, ethnic groups and regions. Many national statistics can be compared between countries, too. Durkheim’s study of suicide (1897) compared official suicide rates between countries and religious groups.

It has fewer ethical problems, and the researcher can’t influence it

With most secondary data, the researcher has no contact with the people the data is about. Nobody is put under pressure, deceived or harmed by the research. And because the data already exists, the researcher can’t affect how people behaved when it was produced. There is no interviewer bias and no Hawthorne effect.

HOWEVER, this doesn’t apply to all secondary data. Personal documents such as diaries and letters raise questions of consent and privacy, especially if the people who wrote them are still alive.

Limitations of secondary data

It was collected for someone else’s purposes

Secondary data was produced to answer someone else’s questions, not the researcher’s. The categories used may not match what the sociologist wants to measure, and key information may simply be missing. Official statistics, for example, may not break the data down by the variable the researcher is interested in, such as social class.

The researcher has no control over how it was collected

The researcher can’t check how carefully the data was gathered, or who was missed out. Even major national surveys have quality problems. The response rate for the Crime Survey for England and Wales is now 48%, against 64% before the pandemic, which raises the risk of non-response bias. British Social Attitudes moved from face-to-face interviews to an online method in 2020, and its response rate is now 11.7%.

A secondary researcher using either dataset has to accept these problems rather than fix them. Large datasets can also be complex, and a researcher who didn’t design the study may take a long time to understand how it works.

The way things are measured may change over time

Official statistics only allow valid comparisons over time if the thing being counted is defined and recorded in the same way. Often it isn’t. The rules for counting police-recorded crime have changed several times, for example, so a rise or fall in recorded crime may partly reflect a change in recording practice rather than a change in the amount of crime.

It may reflect the biases of those who produced it

Official statistics are produced by the state, and what gets measured, and how, reflects the priorities of those in power. That limits what a researcher can find out from them.

Documents are just as likely to be biased. A government report may present a policy in the best possible light, a newspaper is written to sell copies, and an autobiography is written to present its author in a particular way.

Interpretivists go further and argue that many official statistics are social constructions. They are the product of decisions made by the people who record them. J. Maxwell Atkinson (1978) argued that suicide statistics reflect how coroners decide whether a death counts as suicide, not simply how many people took their own lives.

Documents may lack authenticity and representativeness

John Scott (1990) suggested four criteria for assessing documents:

Criterion	The question to ask	Example of a problem
Authenticity	Is the document genuine, and do we know who wrote it?	Parts of an old document may be missing, or its author may be unknown
Credibility	Is it free from distortion, and is it sincere?	A politician’s diary may be written with later publication in mind
Representativeness	Is it typical of the documents of its kind?	Many documents don’t survive, and others are deliberately withheld from researchers
Meaning	Is it clear and understandable?	Older documents may use language whose meaning has changed

Representativeness is especially a problem with older documents. Many were never stored. Others deteriorate with age and become unusable. And historically, the people most likely to leave written records were the literate and the powerful, so documents tend to tell us more about them than about everyone else.

Secondary data and the theoretical perspectives

Which kind of secondary data a sociologist prefers depends largely on their theoretical perspective (see positivism and interpretivism).

Positivists favour quantitative secondary data, especially official statistics. They see it as objective, reliable and suitable for identifying patterns, correlations and trends, as Durkheim did in his study of suicide.

Interpretivists favour qualitative secondary data, especially personal documents. Documents such as diaries and letters give insight into the meanings people attach to their experiences, in their own words. HOWEVER, interpretivists are critical of official statistics, which they see as social constructions rather than social facts.

Strengths and limitations of secondary data: summary table

This table organises the main points using the usual theoretical, practical and ethical headings.

	Strengths	Limitations
Theoretical	Large datasets can be representative and allow comparisons over time and between groups (positivists). Personal documents can give valid, in-depth insight into meanings (interpretivists)	Collected for someone else’s purposes. Definitions change over time. May reflect the biases of those who produced it. Documents may lack authenticity, credibility and representativeness
Practical	Huge amount available. Often free and quick to use. May be the only way to study the past	No control over how it was collected. The data needed may not exist. Large datasets can be complex to understand
Ethical	Usually no contact with participants, so less risk of harm	Personal documents raise issues of consent and privacy

How secondary data comes up in the exam

Secondary data can come up in several ways. The most common are a 10-mark “outline and explain” question on the practical or theoretical problems or advantages of a type of secondary data, or an essay on the usefulness of documents or official statistics. In those questions, name the type of secondary data you are writing about. “Official statistics” or “personal documents” will earn more than “secondary data” in general, because the strengths and limitations differ so much between them.

Two worked examples:

• Two practical problems of using documents in social research (10)
• Outline and explain two reasons why some sociologists choose to use official statistics in their research (10)

Theory and Methods revision bundle

If you like this sort of thing, then you might like my Theory and Methods revision bundle, which contains the following:

Signposting and related posts

This post has been written primarily for students studying the research methods topic, as part of the AQA’s A-Level Sociology course.

Secondary data is usually taught alongside official statistics, the main form of quantitative secondary data, and secondary qualitative data, which looks at documents in more depth.

For the form of data rather than who collected it, see qualitative data: strengths and limitations.

Personal documents and content analysis of the mass media cover two specific types of qualitative secondary data.

For more detailed posts on research methods, please see my page on research methods.

Please click here to return to the homepage – ReviseSociology.com