Introduction to research methods in psychology/IRMP102/Longitudinal Research
Similar to cross‐sectional studies, longitudinal methods are often used in developmental psychology when the subject‐variable age is studied. Unlike cross‐sectional studies, however, longitudinal studies involve following a group of individuals over time and repeatedly measuring their behaviour. For example, in one study the peak performance of individual athletes was followed over several decades (Ericsson, 1990). The best performance each year was studied for three world‐class athletes who competed in three different events: long distance running, hammer throw, and javelin. The results showed that the performance of all three athletes remained impressive over decades (there was no precipitous drop in performance as the athletes aged). On the surface, you might think this is surprisingly good news; as we age our physical capabilities diminish relatively little. However, once again the variable of age is confounded with several other factors. The age of the athletes was not the only difference between the different measurement times. Over the decades, training techniques (such as coaching, video analyses, strength conditioning, knowledge), equipment (clothing, shoes, track surface), and nutrition have also improved. In addition to these confounds, longitudinal studies may start with many participants and gradually have fewer available at each measurement time. People might move away, die, or refuse to continue to participate. This loss of participants is often not random. For example, a longitudinal study involving sprinters might lose runners who were injured more frequently and who ran much more slowly as they aged. These athletes would be more likely to stop competing. Therefore, athletes who can compete effectively over time are more likely to continue to participate. This non‐random loss of participants over time is referred to as selective attrition.
Advantages of longitudinal designs
- At each measurement time, the same participants are used so that individual differences are minimized.
- Longitudinal studies, like cross‐sectional studies, are often interesting and suggest relationships between variables that can stimulate theories and experiments to help clarify the nature of these relationships.
- These studies can help provide support for, or disconfirm, theories developed in more artificial settings, such as a laboratory.
- They allow us to study variables, such as age, that cannot be manipulated experimentally.
Limitations of longitudinal designs
- Longitudinal studies are relatively expensive in terms of time, money, and effort. The researcher has to wait until the participants have aged and then locate them to reassess them.
- Selective attrition can bias samples.
- These designs are always confounded. Therefore, cause‐and‐effect relationships can never be determined.
- The measurement times always differ in multiple ways (they are confounded), and this may artificially minimize any differences between times.
- In longitudinal studies, the participant is always repeatedly tested. This repeated testing can influence results due to the participant's performance changing purely as a result of practicing a task. This is called a Practice Effect. Imagine that you are a participant in a longitudinal study of IQ. As you repeat the IQ test, your performance may increase as you become less nervous and remember answers from previous testing sessions.
