Psychology in school and for school: an "experimental" reflection on the use of tests.

Abstract

Research in psychology is used to dealing with large numbers. Researchers take great care to validate their hypotheses on very broad samples, so as to guarantee the generalizability of the results obtained. This is also shown by the space given in psychology research manuals to the concept of external validity and to the factors that threaten it. External validity means the extent to which the results of an experiment can be generalized beyond the experimental conditions (Kazdin, 1996). Good external validity is obtained, on the one hand, by constructing a representative sample, through a careful analysis of the socio-demographic characteristics of the people making it up; on the other, by ensuring the standardization of the experimental situation to prevent the environment where the research takes place from invalidating the results obtained (Kazdin, 1996). Most experiments use a large number of people due to the desire to generalize the results to as broad a slice of the population as possible. In their studies, therefore, psychologists come into contact with many people, at times with many groups of people, but, being driven essentially by the need to ensure external validity for their research, they are more interested in the number of people contacted and in their individual characteristics than in the relationship that can be established between them and towards the subject of the research. This means that the unit of analysis with which the data is interpreted is the single individual. Despite the large sample of people available, these are all studies as single cases. As Kazdin underlines when talking about the importance of external validity, the explicit reference point of these experimental designs is medical research, which, by its mission, takes the single person as the object of its intervention: “the concerns about the characteristics of the sample and the implications arising from the need to extend the results to other subjects are well illustrated in medical research in which the intervention (for instance, consumption of soft drinks or of a particular food) is administered to the subjects (for instancelaboratory rats) with the intention of showing that it causes cancer. Undoubtedly non-laboratory rats would be glad to know if these results can also be applied to them and to their daily diet” (Kazdin, 1996, pp. 43-44).
The only difference lies in marking out the portion of the individual that is of interest: the body in medicine; personality traits, the emotions (or other things depending on the school of orientation) in psychology. This brings to mind another practice, the custom in school which “strangely [...] has used from its institutional beginnings, the “class”  as place of learning [...] but the fact remains that the teacher, at school, has made minimal use, and only in sporadic cases, of the resource that the “class-group” embodies; today he or she still usually pursues individual learning, assesses the learning of the single individual, reasons according to strictly individualistic models when he/she has to think about his/her own educational performance” (Carli, 2001, p. 66).