Generalization is an important issue in scientific research in general. When we conduct experiments we want to find patterns using a small sample of our universe of cases that can be generalized to the population. The above question seems to emerge from the fact that certain research methods in Psychology (qualitative methods: QM), are not about generalization, because it is argued, either they cannot produce universal laws of behaviour, or the aim is not to generalize, but rather investigate about other aspects of individuals, often neglected, it is said, by quantitative methods (Howitt, 2013). As Howitt mentions, there is no single characteristic that defines qualitative research, but in general they tend to be concerned with the richness of description or the capturing of the individual’s perspective. There is also a rejection of positivism and a use of post-modern perspectives. At the same time, these methods tend to focus on the examination of the constraints of everyday life. This would not be the case with quantitative research, which is more concern with numbers and statistics and falls within a more positivist view of reality and scientific inquiry.
There seems to be a trend of questioning the importance or validity of generalization to give more relevance to QM, specially in social sciences, since if generalization is questioned and found not to be a necessary condition for scientific research in psychology, then QM can gain validity as a way to do scientific research in psychology. There has been a general attitude of opposition to the classical–positivist paradigm from those who have promoted qualitative research (Howitt, 2013). The author mentions the “freedom” that students feel in it from the “tyranny” of numbers and statistics. I think this exemplifies that kind of negative view towards quantitative research. Our main question is related to this, but it would seem an improper one if it came from the fields of physics or chemistry, as in those areas, establishing universal laws could be regarded as the main purpose. We could ask however about the validity of generalization itself.
One could say that the quantitative vision of reality is out-dated, that “we can’t perceive reality as it is” or “we cannot know reality”. I would criticize what could be regarded as a “hard position of positivism”, where only the usual statements of the sciences (e.g. physics, chemistry, biology, etc.) can be regarded as truthful or not, leaving aside fields like aesthetics or ethics (Wittgenstein, Muñoz, & Reguera, 2003). But I find that if we say that “we cannot know reality”, the assertion includes itself, in the sense that it refers to the truth of things that includes the statement itself, which would imply that we know reality, at least to some extent.
On the other hand, generalization in scientific research might be criticized by saying that applying some rule to everyone from a few cases is absurd, because there might be a case that does not follow the rule. Newton had a brilliant answer to that issue when he said: “I frame no hypotheses” (hypothesis in this sense refers to any assertion that it is not based on empirical evidence), meaning that every assertion not based on evidence is irrelevant in science, it would be absurd therefore to question the relevance of induction on the basis of some non-existent entity, that is, saying that there could be a case out there that does not follow the rule, the important thing is that we base our claims in evidence, and that is what matters in science, the possibility that there might be certain case is no solid foundation (Peikoff, 2012).
In any case, I do not think that being successful in disproving generalization as a necessary and sufficient condition, if one can do that, gives QM a sort of validation, because one still has to provide the conditions that the methodology must meet in order for it to be valid, that is, to be considered scientific. If generalization is not a necessary or sufficient condition, which is it then? The answer is not provided.
To address our problem, I think we must ask ourselves which are the criteria we are referring to, in other words, according to what do we evaluate generalization as a condition. If we are trying to find universal laws of behaviour, then indeed I think it would be a necessary condition. But our main question can be confusing, because we are measuring generalization according to “scientific research in psychology”, and within the scientific method there are many stages or steps, like observation and hypothesis making (Nola, 2013). This author mentions the distinction in philosophy between the methods for finding, inventing, constructing or discovering hypotheses, and those for justifying, proving, validating, warranting or appraising hypotheses. We could make the case that QMfall roughly under the first category. At the same time, one has to take into account other conditions that are also important, such as parsimony, efficiency or falsifiability (Gauch Jr, 2002). In that sense, generalization is not enough to do research, and even more, it could just be considered as one of the steps of scientific research (one of the final steps, when we arrive at conclusions from our experiments). We could also say that there has to be parsimony or falsifiability in our experiments in order for the generalizations to be regarded as good ones, in other words, that they can assume a scientific standard, so generalization itself would not be a sufficient condition.
Therefore the issue is more complex than the question seems to show, and I think we can say that there is no need for an opposition between one kind of method and the other. I think one can try to generate an opposition, for example, if one wants to necessarily apply generalization to QM, which, if we analyse all the stages of the scientific method, would be inaccurate, as the use of QM could be considered as the first stage in scientific enquiry, when one generates hypotheses, and generalization would be another further stage.
There seems to be, in my opinion a true opposition between a subjective, constructivist view of reality that is often present in qualitative research, and the idea of an objective, independent existence of reality, which we can perceive and represent (quantitative). This view of the qualitative research does not affect, I think, its ability to generate knowledge about human psychology. Indeed, there are many examples of powerful insights about human behaviour that resulted from QM such as observation and interviews (Gazzaniga, Ivry, & Mangun, 2013), despite the prevailing constructivist philosophy.
Gauch Jr, H. (2002). Sientific Method in Practice.
Gazzaniga, M., Ivry, R., & Mangun, G. (2013). Cognitive neuroscience: The biology of the mind(Fourth Ed.). New York, N.Y.: W. W. Norton & Company.
Howitt, D. (2013). Introduction to qualitative methods in psychology(Second Ed.).
Nola, R. (2013). Theories of Scientific Method: An Introduction.
Peikoff, L. (2012). The DIM hypothesis: Why the lights of the West are going out.New York: New American Library.
Wittgenstein, L., Muñoz, J., & Reguera, I. (2003). Tractatus logico-philosophicus. Madrid: Alianza Editorial.