|
Description: |
In the usage of present-day statistics ‘statistical inference’ is a profoundly ambiguous expression. In some literature a statistical inference is a ‘decision made under risk’, in other literature it is ‘a conclusion drawn from given data’, and most of the literature displays no awareness that the two meanings might be different.
This book concerns the problem of drawing conclusions from given data, in which respect we have to ask: Does there exist a need for the term ‘statistical inference’? If so, does there also exist a corresponding need for every other science? If so, how does, for example, agronomy then manage to reason in terms of botanical inference, soil scientific inference, meteorological inference, biochemical inference, molecular biological inference, entomological inference, plant pathological inference, etc. without incoherence or self-contradiction? Consider the possibility that agronomy does not reason in terms of such a motley of special kinds of inference. Consider the possibility that, apart from subject matter, botany, soil science, entomology, etc. all employ the same kind of reasoning. If so, must we then believe that statistics, alone among all the sciences, is the only one that requires its own special kind of inference? |
