Why is longitudinal study so important?

To mark the publishing of Insights 2014, Dr Jon Burton, Survey Manager for Understanding Study explains why longitudinal studies are so important for society.

Our society is under constant change, through technological innovations, financial and economic crises, environmental upheavals and globalisation. Longitudinal studies are our best tool to understand how we are affected by these changes.

Most surveys ask a set of people some questions, and then after some time they ask a different set of people the same questions. An example of these surveys are the political opinion polls often reported in the media. The information we get from these is what proportion of the population of voters favour one party or another. Comparing two of these surveys we can see how the population has changed across the two time points – one party may have increased their support by one or two percentage points.

Only by asking the same people at different points in time can we see how things are really changing, or staying the same.

Surveys with a difference

Longitudinal surveys are different. One set of people are asked the questions, and then after some time, the same people are asked those questions again. This allows us to look at individual level of change, not just change in the population. We may find that whilst overall one party has increased their support by a couple of percentage points, this actually masks a lot of ‘churn’ as individuals shift their support from one party to another, from preferring a party to undecided, or from undecided to a particular party.

Only by asking the same people at different points in time can we see how things are really changing, or staying the same. This can profoundly affect how we think about important issues. For example, in the early-1990s it appeared that there was a portion of the population who were persistently in poverty. The poor, it seemed, are always poor. Using a longitudinal survey – in this case, the British Household Panel Survey, the forerunner to Understanding Society – researchers found that in actual fact there was a dynamic picture of change. Some families would be exiting poverty, as other households entered poverty. Half of those in poverty in one year were not found to be in poverty the next year. The group of households who were persistently poor was much smaller than had previously been thought. Poverty was something which could be escaped, but it was also something that affected more people at one time or another.

Measuring change over time

Another strength of the longitudinal design is that it shows how events and actions at one point in time can affect outcomes later in life. If we only interview people once, we cannot tell whether, for example, their ill health caused them to lose their job or whether losing their job led to their poor health. We can only say that ill health and unemployment are associated with each other. By talking to the same people every year, researchers can observe which event happened first, and so can draw out the causal relationship between ill health and unemployment. This example can be extended to a wide range of other areas; how health affects educational attainment, or how education affects employment prospects.

We are all products of our past experiences. Longitudinal surveys allow researchers and research users to see how those experiences have impacted on our outcomes today and in the future.

This article is an extract from Insights 2014, follow the #insights2014 conversation on Twitter.

More about the author

Jon Burton is the survey manager for Understanding Society. He co-ordinates survey operations and is responsible for the delivery of the survey – how it looks and the questions it contains – together with our fieldwork agencies.

Related links