Using survey data
Using survey data in a dissertation
What is available?
The amount of survey data available is amazing. For instance, in the UK, government departments and the Office for National Statistics (ONS) use representative surveys to inform policy. Academics and independent research organisations (such as NatCen) run large-scale social surveys on a wide range of topics. Data from these studies is usually archived and made available for others to use.
These large social surveys tend to be:
- Designed by experts using tested methods of data collection.
- Have representative samples that can be used to describe a population.
- Contain enough cases to allow you to compare groups in the population.
Some surveys focus on a specific topic such as health or crime while others collect more general information about people’s lives. Collectively, there is data available on a wide range of topics.
There are too many surveys to list them all, but our top surveys for student projects include:
- British Social Attitudes (BSA) – Nearly 40 years of data on the attitudes of the British public towards a wide range of social issues. A new survey comes out each year, giving topical information on public opinion. If you want to know more visit our quick start guide to the BSA.
- Crime Survey for England and Wales (CSEW) – The most widely used survey data for researching crime rates and the effects of crime in the UK. As well as experiences of crime, there’s data on public attitudes towards crime, the police and the criminal justice system.
- Health Survey for England (HSE) – An annual survey carried out since 1991. With a mix of questionnaires, physical measurements, and blood samples – there is data on most aspects of health and health related behaviour including obesity, smoking, drinking, mental health and other medical conditions. If you want to know more visit our quick start guide to the HSE.
- Labour Force Survey (LFS) – A survey of the employment circumstances of the UK population. It provides data for official measures of employment and unemployment. There’s also data on job types, satisfaction, pay and training.
- Understanding Society – An ongoing study following UK households over time. It is a very large study with questions on most social and economic topics. You can examine using data from one or more sweeps.
- Millennium Cohort Study (MCS) – Following the lives of young people born 2000-02, this study gives amazing insights into the lives of today’s young adults. Across ‘sweeps’ of the study, there’s data on varied themes such as education, health, and family networks.
There are many others covering topics such as living costs, housing, volunteering, cultural participation, sex and drugs.
How can we use survey data?
Survey data can be used for varied research projects and questions.
The datasets contain records for each case in the sample, with cases typically being individuals or perhaps households or businesses. We can use survey data to:
- Make tables and graphs to show the prevalence of social phenomena (e.g. how common something is).
- Examine associations between variables – for example, differences between young and old or how attitudes relate to behaviour.
- Select specific groups to look at such as young adults, women or the unemployed.
How does paid and voluntary work experience affect early career outcomes for English graduates?
Next Steps was ideal for this project because it contained information collected from young people during their degree about their work experience and later about their career outcomes including earnings at age 25. Using the data, the student examined how earnings varied by work experience and other characteristics.
Is screen time associated with body mass among adolescents?
The Millennium Cohort Study was ideal because the study:
- Measured participants’ body mass and body fat.
- Asked about TV, social media and gaming use.
- Is longitudinal, making it possible to look at the relationship between behaviour at 14 and outcomes at age 17.
How are attitudes towards welfare policy influenced by receiving different benefits?
The British Social Attitudes (BSA) survey was used to explore how attitudes towards government spending on different aspects of social policy (e.g. unemployment) varied among those receiving benefits. The data was ideal as the BSA regularly asks questions about support for welfare policy as well as whether people receive various types of benefits.
We can think about change over time at a collective level (e.g. has society become more fearful of crime?) or an individual level (e.g. Does growing up in a high crime area affect someone’s fear of crime?). What is possible to examine depends on how time is captured in the data, see below to explore further.
Repeated cross-sectional studies
Cross-sectional surveys collect data from a sample in a single point of time. When surveys are repeated with new samples (repeated cross-section data), we can use the data to analyse trends such as whether the prevalence of some (such as smoking, voting, or certain attitudes) has changed over time for the whole population or for certain groups over time.
Longitudinal studies collect data from the same sample over time. We can use longitudinal data to examine how individual life experiences affect outcomes such as (likelihood of smoking, voting or having certain attitudes). If you would like to learn more about longitudinal data, check out our Data Skills Modules on Introduction to longitudinal data.
Questions about the past
Surveys can ask respondents about their past. For instance, surveys might ask a respondent about their childhood to measure ‘upbringing’. Measuring retrospectively can produce more measurement error but may be best we can get.