A Causal Paradox for Hypothesis

Abstract

As it is well-known any respond to a research question which intends to give an explanation has the form of a hypothesis. Such an explanation in essence aims to represent a structure which contains three components: the term/s used to explain something (explanan/s), the term used for the event to be explained (explanandum) and the term express the causal relation between them (explicandum). But if you want to put such a hypothesis into an empirical testing procedure, you have to change its causal structure by substituting with a statistical correlation structure. Otherwise you could not realize an empirical testing procedure. Here the paradox situation in essence consists in the contradiction between our aim to control empirically if the explanan/s is/are really explanatory/is factor/s (i.e. causal term/s) and the real impossibility to do empirically this control. For resolving such contradiction we need to change our aim and to substitute the control of causal relation (explicandum) with the control of statistical association between the terms included in the hypothesis. So doing, in fact we have substitute the hypothesis as explanatory structure with its “quantitative twin”. The “qualitative twin” is substituted with its “quantitative twin” (as its operationally or parametrically transformed variant) for the necessity to realize measurements as part of an empirical testing process. Logically this is not legitimated, but methodologically we have not other choice to do.

DOI: 10.5901/mjss.2014.v5n7p577