Abstract:
Object: the purpose of the presented scientific article is to conduct an economic and statistical analysis
of the innovative activity of the enterprises in Kazakhstan; the object is the enterprises of Kazakhstan, characterized
by innovative activity, introducing and using in their activities the results of intellectual labor, represented
by new technologies, technical objects, patents for inventions, utility models, industrial designs, etc.
Methods: to achieve the goal of the research, general scientific methods were widely used, in particular,
the method of analysis, which made it possible to determine the entire set of parameters that characterize the
innovative activity of enterprises in Kazakhstan; a generalization method aimed at establishing existing relationships
between the considered economic objects and phenomena; method of graphic interpretation, which
made it possible to visualize the results obtained; economic and mathematical methods, represented by correlation
and regression analysis, methods of checking the constructed model based on the Student's, Fisher's
test, the coefficient of determination and the Darbin-Watson test; forecasting methods based on the constructed
multiple regression.
Results: within the framework of the research, the authors built a multiple regression model for the innovation
activity of the enterprises in Kazakhstan, the adequacy and correctness of which was verified using
the Student's t-test, Fisher's test, the coefficient of determination and the Darbin-Watson test. At the same
time, using the Gauss method, the normal system of equations was solved, which made it possible to obtain a
standardized regression equation, taking into account the calculated coefficients. It is also important to note
that the article provides a comprehensive analysis of the development of the theory of innovation and key
approaches to the interpretation of the concept of “innovative activity of an enterprise” based on the research
of various scientific materials, including those widely represented in scientometric databases - Scopus,
Clarivate Analytics, Google Scholar and RSCI.
Conclusions: using of correlation-regression analysis in terms of constructing multiple regression models
is the optimal method that allows not only to effectively and comprehensively describe the innovative
activities of the enterprises in Kazakhstan, but also to use it for the subsequent forecasting of the analyzed
base indicator (dependent variable). The most important aspect remains the choice of parameters, which
should be built on a deep and comprehensive understanding of the economic phenomenon under consideration,
while their selection should be based on their strict economic interpretation, in particular by constructing
matrices of pair correlation coefficients and taking into account multicollinearity.