Why study the life course? . Life-Course research and the longitudinal method

WHY STUDY THE LIFE COURSE?

For researchers, money matters, and life-course research is usually expensive. It takes a lot of time, meaning more work and more research hours, and often creates (even) more methodological issues than most other forms of research in social science. It is therefore highly reasonable to pose the question and demand an answer from its advocates: why do this form of research at all, and why has it become so prominent over the last three decades?

Because, using this methodology, we can answer important research questions which other methods have trouble addressing. The foremost questions in this regard concern the issue of continuity and change in behaviour over time.

To contrast life-course research, consider so-called cross-section methodology. Here, researchers study a sample or a population at a single, specific moment in time. The resulting data gives us the possibility to study individuals or groups of individuals when it comes to a whole range of variables (such as crime, peer rela- tions, health, employment, and so on), and how these are related to each other. We commonly call these between-individual similarities and differences. Using longitu- dinal methodology, we study the same individual at repeated points in time. Since we can study the relationship between the different variables as we go through time, we can also see whether the relationship between them is constant or changes over time. It is likely, for example, that as long as the individual engages in repeated, serious crime, he or she will have a relatively unstable form of conventional work (if he or she has a job at all), have turbulent relationships with other people, have increasing problems with mental and physical health, etc. Whether or not this is the case becomes impossible to study using cross-sectional research design – it would have to be inferred by theory.

We can also study what happens when change occurs in one of the variables: is

there a change in a person’s criminal offending when he or she gets a more stable job, starts a romantic relationship or becomes ill? Crucially, we can also – sometimes – study which change occurs before the other, i. e. does a person’s criminal offending change as a result of them getting a job etc., as Sampson and Laub (1993) argue, or is it the other way around – a person begins to decrease their criminal offending and then gets a job or starts a romantic relationship? This suggests two different causal processes, and thus two different answers to the question of what makes people cease their criminal offending, with important implications for policy and practice.

Thus, we can begin to approach the difficult question of cause and effect, that is, the question of causality.

According to de Lange (2005: 18), four criteria in research in the social sciences should be fulfilled if we are to be able to draw causal conclusions:

This research should i) demonstrate that the cause variable precedes the outcome vari- able in time, ii) show a significant statistical relationship between the presumed cause and outcome, iii) exclude possible alternative explanations, and iv) provide a profes- sional theoretical interpretation of the relationship(s) under study.

As you can imagine, criteria ii, iii and iv usually give rise to a number of challenges. But unless you are using longitudinal methodology, it can be extremely hard – in fact, often impossible – to have control over the factor of time, i. e. the first criteria. Life-course research and the longitudinal method provide you with this possibility – when done right.

LIFE-COURSE RESEARCH AND THE LONGITUDINAL METHOD

Life-course research and longitudinal research are not necessarily synonyms. To state it simply, life course suggests a distinct research perspective, whereas longitudinal suggests a specific methodology which follows from the life-course perspective.

The aim of life-course studies is to study, understand and explain one or more individuals’ lives over a short or long time span; some may focus on a specific stage of the life-course (e. g. ‘What effect does the transition to adulthood have on the indi- vidual’s future criminal career? ’), whereas others take a more holistic approach (‘How does criminal offending wax and wane across the life course? ’). The life-course per- spective assumes that what happens at one time in life, is contingent on – but not necessarily determined by – what happened at an earlier point in life. This theoretical assumption leads to methodological consequences, where life-course research must be carried out using longitudinal methodology.