THE NEED FOR A LONGITUDINAL APPROACH

Longitudinal studies are vital for studying individual development and cannot normally for this purpose be replaced by other kinds of studies. This has been asserted many times (see Chapter 1 in this volume; Magnusson, 1988; Mednick, Harway & Finello, 1984). The main reason for using a longitudinal approach is really common-sense: If you want to understand how an individual develops over time, you have to study the individual over time. All other approaches are more indirect and have to make questionable assumptions for the results to be applicable for statements about individual development. For overviews of longitudinal versus other designs, the reader is referred to Bakes (1968) and Bakes and Nesselroade (1979).

The above should not be interpreted to mean that a longitudinal study always is necessary when development or change is to be studied. For instance, within epidemiology, a case-control study can be faster, cheaper, and more efficient than a longitudinal study for the (preliminary) study of the aetiology of a certain decease.

THE RESEARCH PROCESS

The research process can be described as having the following components:

1. 1 The process starts from the specific aspect of reality that is at the focus of enquiry.

2. 2 By some kind of data collection procedure, the raw data are produced.

3. 3 The raw data are often further processed to give the variable values used in the analyses (transformed, summed to form indices and so on).

4. 4 Statistical analyses are performed.

5. 5 The results are interpreted in relation to the theoretical formulations.

6. […]