In psychology, data has been collected for years using methods that focus on variation between test subjects. This variation is called inter-individual variation. Statistical measures (such as the average and standard deviation) are calculated using the inter-individual variation. All statistical data is extracted from the population and all types of research look at interindividual variation. It does not matter whether longitudinal or cross-cultural research is being carried out, the analyzes that are used focus on interindividual variation.

According to some researchers, this is sometimes a downside. When the findings obtained from a total population are applied to the individual, the level of interindividual variation shifts to the level of intraindividual variation. With the help of ergodic theories one can find under which circumstances findings based on interindividual variation can be applied to individuals. The conditions are very strict and it will therefore not often happen that the findings can be shifted from the inter-individual to the intra-individual level.

Conditions for ergodicity

Ergodicity is about when (under what conditions) analyzes of interindividual variation yield the same results as analyzes of intraindividual variation. To be able to do this, people must be seen as a dynamic system consisting of behavior, emotional, cognitive and other psychological processes that change over time.

According to Cattell, all these different psychological dimensions can be measured with time as a dimension, psychological variables as a different dimension and different test subjects as the third dimension. This is called Cattell’s data box. An analysis of the intraindividual variation is called the P technique and the analysis of the interindividual variation is called the R technique. These two (and therefore the condition for ergodicity) can only be equal if there is homogeneity in the population and if the data is stationary.

Homogeneity

Every test subject must act according to the same statistical model if homogeneity is to be achieved. An IQ test can measure general intelligence in one person, while it measures verbal intelligence in another person. In this case there is no homogeneity. If the strength for items differs from person to person, then there is no homogeneity either.

Stationary

The other condition for ergodicity is stationary. This means that statistical characteristics remain the same over time. This cannot happen if development processes are involved. The IQ of a four-year-old changes as he or she gets older and this is therefore not a stationary statistical characteristic (and therefore ergodicity is not met).

The writers of this article believe that psychological processes differ from person to person and that, therefore, one should actually look at the intra-individual variation if one wants to study these processes. This also has the advantage that these processes can be guided to an optimal level per person. According to them, doctors should look more at the person when they prescribe a certain dose of a certain medicine and less at the average of people (the dose that is usually written out).

The intra-individual approach is fairly new and the writers should pay more attention to it. According to the writers, this approach should actually be necessary.