Generalization failure

Up to date, there is no progress in generalization empirical data in software engineering. Why?

It is necessary to have an effective idea to perform efficient synthesis of empirical data.

This idea may rank the data.

It becomes possible to determine what data are of primary importance, as some are minor, what data are useful and which are useless.

The idea allows you to add empirical data as a puzzle in the scientific picture of the study area, or the phenomenon under study.

Such kind of idea is needed in software engineering, there is no one. Meanwhile software engineering has accumulated a large number of observations of different facts. Though the scientific community has gained great experience, it is not put in order.

There is no coherent picture of the processes that occur in the software engineering. Since there is no coherent picture, there is no generalization. Since there are no generalizations, there are no essential laws and principles.

The idea needed to be summarized can be found in the theory the of decomposition schemes. It is a scheme of decomposition. There is also a universal model of the algorithm (the canonical algorithm) in the theory of decomposition schemes. The scheme of decomposition and the canonical algorithm are very close in a concept plan.

Scheme of decomposition and canonical algorithm are abstract models, but in the research process they are filled by specific properties.

Decomposition scheme can effectively rank the empirical data, dividing them into primary and secondary factors. None of the factors studied are lost. They only find their place in the whole picture of the processes occurring in software engineering.