A cross-sectional study gathers data from one point in time, but if your research questions relate to changes over time you may wish to consider designing a longitudinal study. This will enable you to look at the same people or situations at key points in time and to con- sider how the changes over time have affected different groups of people. Data is gathered on at least two occasions separated in time.
Many longitudinal studies are large scale and are set up as longitudinal studies from the beginning (prospective longitudinal studies). There are a number of well-known examples of longitudinal studies which look at babies born at a particular time (cohort studies) and then gather data about them at set points throughout their early years – or indeed their life- time. A popular example of this is the television documentary ‘7 Up’ which follows the lives of 14 children, selected from different socio-economic backgrounds, who were 7 years old in 1964 when the programme was first shown. Every seven years a documentary is made about them in the World in Action series, the last one being screened in 2005 show- ing them as 49 year olds. Longitudinal studies are major research undertakings and involve numerous, and often changing, social researchers.
The research design
A large-scale representative survey was used to enable the researchers to identify the smaller sub-groups in which they were interested. A range of data had been col- lected about each couple, and this enabled the researchers to look for relationships between different variables.
The original research had collected data at the level that the researchers were in- terested in – the household. So, in this case, the unit of analysis was the household.
Your research
Research quality check: cross-sectional research designs
Will you be able to use your data to look for possible causal associations between variables?
Are you able to select a sample that reflects a cross-section of the population you are interested in? Are any groups likely to be excluded by your research design?
Could your research design be replicated by another researcher?
Does the research design enable you to claim that your research findings can be generalised to a wider population or different setting? This will depend on the way you have selected your research participants.
Does the research design enable you to collect data that ‘stands in’ for social reality – data that reflects the social reality of natural social settings?
Does the research design enable you to collect data in a consistent and reliable way?
What are the ethical implications of your research design for your research participants?
prospective longitudinal studies Research studies that are initially designed to be longitudinal. cohort studies A type of longitudinal study which looks at a group of people of the same age and then gathers data about them at set points throughout their lives.
However, it is also possible to design smaller-scale studies over quite short time periods. Interviewing people before, during and after a particular event could be regarded as a lon- gitudinal design, if the focus of the research is to track how each individual research partic- ipant anticipated, experienced and reflected on the event.
Real research
The Millennium Cohort Study
Following two major cohort studies during the twentieth century – the 1958 National Child Development Study and 1970 British Cohort Study (both of these studies are still gathering data from their cohorts) – the Centre for Longitudinal Studies (www.cls.ioe.ac.uk/) set up the Millennium Cohort Study involving children born in the year 2000, for which 18,818 children were selected using a stratified sample. By 2008, data had been gathered from the families of the children on four occasions.
The research design for the Millennium Cohort Study (MCS) was based on the following five principles:
1. The MCS should provide data about children living and growing up in each of the four countries of the UK.
2. The MCS should provide useable data for sub-groups of children, in particular those living in advantaged and disadvantaged circumstances, and for children of ethnic minorities and those living in Scotland, Wales and Northern Ireland.
3. As well as data about children, the MCS should provide data about their family circumstances and the broader socio-economic context in which the children grow up.
4. The MCS should include children born throughout a single 12-month period.
5. All children born as members of the MCS population should have a known and non-zero probability of being included in the selected sample.
Source: Center for Longitudinal Studies. www.cls.ioe.ac.uk/studies.asp?section=0001000200010010 (accessed 12 August 2008).
Example B3.7
Students and alcohol: longitudinal study
A longitudinal study could be designed to look at the way students’ drinking pat- terns changed over the course of their university programme. This could include collecting the same data from a sample of students at the key points during their three-year degree programme.
A retrospective longitudinal study may be designed which asks final-year students to recall their drinking patterns at key times during their time at university (for example, during freshers’ week, at exam time and during the vacation), and to identify changes over the three-year period. But, as with all retrospective studies, people’s memories are not always accurate and may be coloured by their current situation and perceptions. We may gain more accurate data if we choose to include students from each of the three years of a degree programme at one particular point in time, and compare the different year groups.
For practical reasons some longitudinal studies cannot involve the same participants on each of the data-gathering occasions. This may be because the participants have died or moved and cannot be contacted or because they no longer want to take part. A study may therefore include different samples of people but using similar criteria for selection. Simi- larly, while a longitudinal study is usually planned as such, there are some situations where data is available from the past (a retrospective longitudinal study) which can be used: for example, where participants can be asked to provide data about their past experiences; or where records may be available – for example, medical records or social work case notes. A
life history approachto gathering data, where participants are asked to tell their life story focusing on common events, could be considered a longitudinal research design if partici- pants are asked to report and reflect on the same key points in their life.
While many large-scale longitudinal studies depend on the gathering of quantitative data from large numbers of participants, a longitudinal study may also be based on gathering in-depth data from a small number of selected participants on a number of occasions.
Real research
Longitudinal and cross-sectional research
B. Milton, S. E. Woods, L. Dugdill, L. Porcellato and R. J. Springett (2008) Starting young? Children’s experiences of trying smoking during pre-adolescence, Health Education Research, 23(2): 298–309
Why do children try smoking when the risks to health are now well known? The researchers wanted to hear from children their own accounts of their first experi- ences of smoking.
What the researchers did
The study is part of a prospective longitudinal study of smoking in a large city in the UK. The study is following a cohort of children from six primary schools in areas selected to include variation in terms of health and socio-economic status. This study gathered data from children when they were aged 9, 10 and 11. The data was collected using questionnaires, interviews and focus groups.
Children were asked about any experiences of smoking they had had, family smoking habits and their opinions of smoking.
Findings
By the age of 11, 27 per cent of the cohort had tried smoking: curiosity and peer influence were important factors in prompting experimentation. The researchers emphasised that approaches to preventing children from smoking must take account of social factors (including friends and family) which play a significant role in whether a child tries smoking. Account must also be taken of the different ages at which pre-adolescent children tried smoking, as different approaches to prevention needed to be taken.
The research design
The study was designed to include a cross-section of children born in the same spe- cific year, ensuring that children from areas of different social and economic back- grounds were included.
As data was gathered each year between the ages of 9 and 11, children were asked about their first experiences of smoking near in time to the event. The same children were involved on each occasion, enabling researchers to look at the changes and processes within children’s lives during pre-adolescence.
retrospective longitudinal study A type of longitudinal study where data is available from the past, for example, where participants can be asked to provide data about their past experi- ences, or where records may be available. life history approach A data collection method where participants are asked to tell their life story focusing on common events.
Real research
Longitudinal and qualitative research
A. Corden and K. Nice (2007) Qualitative longitudinal analysis for policy: Incapacity Benefits recipi- ents taking part in Pathways to Work, Social Policy and Society, 6(4): 557–69
When a new policy is introduced, there is naturally some interest in how it is put into practice and how people experience the changes. The Pathways to Work Pilot was a new approach to helping people who were receiving Incapacity Benefit to take up paid employment. Policy-makers wanted to hear about service users’ experiences of using the new project.
What the researchers did
The study focused on the service users’ experiences of using the service over a pe- riod of time, while also gathering data about changes in people’s lives, their em- ployment, financial situation and so on. Data was collected initially through an in-depth interview and then by telephone after three and then six months. There were three cohorts of research participants, starting at six-month intervals. This ensured that people were included who joined the project when it had been running for some time as well as those involved at the very beginning. A total of 105 people were involved.
Findings
Though people’s experience of the service varied, this longitudinal study was able to show how different groups of people experienced the service over time and also gave some people the opportunity to reflect on the experience after a year. The researchers were able to give feedback to the policy officers throughout the process, and to demonstrate at the end of the study that changes through time are complex, with changing family, financial and employment situations.
Research design
The use of a longitudinal design enabled the researchers to look at the changes and processes involved in the introduction of the project and the experience of the serv- ice users. The use of qualitative interviews to gather data enabled service users to discuss and reflect on their experience at different stages in the process.
The use of three cohorts of research participants starting at different stages in the initial year of the project enabled the researchers to study any changes that were introduced as the project settled down, and to compare the experiences of each cohort.
Your research
Research quality check: longitudinal research designs
Will you be able to collect data on the same variables on each occasion?
Will you be able to gather data from the same people/cases on each occasion? What will you do about people who drop out, cannot be found?