| Summary: | Statistics defines itself as a methodological discipline providing a rigorous, formal framework for scientific empirism based on a mapping of contingent observable phenomena to (real) numbers that can be dealt with, or analysed, computationally. Application of the statistical methodology of data reduction, in turn, requires some representation of the problem context. Most of the time, this amounts to encoding (a part of) the problem context of observation data into another layer of data – called metadata. Based on metadata, procedures of data analysis might be enhanced to encompass also
the analysis and transformation of metadata alongside the accompanied data itself. The paper sketches the outline of a systematic approach to statistical “meta-computing” as a dual-mode proposal of statistical data processing.
|
|---|
