Longitudinal Studies

Longitudinal studies: meaning

A longitudinal study is a research method commonly used by sociologists. Research that takes the form of a longitudinal study can also be called a longitudinal study design.

There is no set period of time for a study to be called a longitudinal study as long as the participants are observed repeatedly throughout. However, it is common for studies to last for months or even several years.

Fig. 1 - Longitudinal studies look for changes over time

Differences between longitudinal and cross-sectional studies

To fully understand how longitudinal studies work, it may be useful to consider the differences between longitudinal studies and cross-sectional studies.

As mentioned above, longitudinal studies track changes over a certain period of time. Cross-sectional studies are the opposite; they take a 'snapshot' of how a certain population (or a group within that population) looks at a specific moment.

Cross-sectional studies are also useful tools and can work with longitudinal studies as they offer different advantages:

• They are quick, simple, and cheaper to carry out.

• They can help to uncover correlations between two variables, which may then be studied longitudinally. For example, it is not worth carrying out a longitudinal study on the relationship between education level and happiness if the cross-sectional study has not shown any correlations between the two.

• They tell us how a population or certain group of society is doing at any specific moment.

Carrying out a longitudinal study

As a researcher, before carrying out a longitudinal study you must consider where the data will come from. If you wish, you can use data that has already been collected by someone else - secondary data. You may be able to do this if the sources are trustworthy, such as statistics produced by the government.

However, if you choose to collect your own data, this will be primary data that you obtain through your own research. You can carry out either a retrospective or prospective longitudinal study.

Fig. 2 - There are several examples of long and well-known longitudinal studies

Retrospective longitudinal studies

This type of study collects data on events that have happened already. As an example, you may choose to analyse the educational attainment of girls from the years 1980 to 2000.

Prospective longitudinal studies

These studies collect data on a chosen group and follow it in real-time to track events or variables that come up in the timeframe. For example, you may choose to follow a small ethnic community and study their development every year for the next five years.

Longitudinal study example

Let's take a look at how you might go about conducting a longitudinal study on the relationship between health and socioeconomic status.

As a researcher, you decide to look at the relationship between health and socioeconomic status in children. However, you are not sure how to conduct this research.

At first, you start with a cross-sectional study to examine whether there is a link between health and socioeconomic status (the variables) in the first place. You look at 50 working-class and 50 middle-class children, all of whom are five years old. You discover that the physical and mental health of working-class children is weaker than their middle-class counterparts.

Now that you have found a relationship between the variables, you opt for a longitudinal study design, tracking the physical and mental health of both groups of children at the ages of 5, 10, and 15. At the end of the study, you have extensive data about the health of 100 children who have been studied for 10 years. You can now draw conclusions about any links between health and socioeconomic status.

Famous longitudinal studies

Let's look at three prominent longitudinal studies.

The Harvard Study of Adult Development

The Harvard Study of Adult Development is one of the longest known longitudinal studies, with over 80 years of data collected. Also known as the Grant Study, researchers started following a group of 268 male Harvard students in 1939, and tracked their health and happiness during their lifespans. Although very few of the original participants are still alive, the study later expanded to include their offspring.

In the 1970s, researchers expanded the study again to include a control group of 456 participants (inner-city residents, all men), called the Glueck Study. The project aimed to identify predictors of healthy ageing.

The Millennium Cohort Study

A well-known and recent longitudinal study is the Millennium Cohort Study, which tracked the lives of approximately 19,000 people born across the UK in 2000-02. It ran from the year 2000 to 2011 and tracked children until they were around 11 years old.

The study looked at the impact of early socialisation on health and educational development.

1970 British Cohort Study

The 1970 British Cohort Study was a longitudinal study that followed the lives of 17,000 people born in a single week in England, Scotland, and Wales in 1970. At first, the data collection focused on medical health; however, it then broadened to examine:

• overall health
• physical, social, and educational development
• economic factors

This study was carried out by the UK Centre for Longitudinal Studies (CLS) and the data is freely available to the public.

Advantages of longitudinal studies

There are several reasons why researchers would opt to use longitudinal studies in their research.

• They allow researchers to analyse how variables change over time. For instance, does social mobility improve through generations of the same family?

• Researchers can collect quantitative data which can then help identify trends and patterns using questionnaires and structured interviews.

• Longitudinal studies can use triangulation, which is the combination of two or more research methods. This can increase the validity of the data collected.

Disadvantages of longitudinal studies

Not all research would make sense or achieve its outcomes using a longitudinal study. The table below looks at the advantages and disadvantages of this method.

• It is hard to keep the sample intact over time if participants are not committed to the project and choose to withdraw - this is called the attrition (drop-out) rate.

• It can be extremely expensive - researchers need to procure long-term funding.

• Participants may act differently if they know they are being studied - displaying demand characteristics.

• It can be hard to process qualitative data (such as from unstructured interviews and observations) due to the often large sample sizes.