The dissertation is devoted to the study of the functioning and development of socio-technical systems in the digital environment of organizations. The main focus is on the evaluation, analysis, and forecasting of employee productivity in IT projects. Productivity is considered as an integral characteristic of the social component of a socio-technical system that directly affects project management efficiency.
In modern conditions of digitalization, IT projects are complex socio-technical systems. In such systems, employees, information systems, development tools, and organizational processes interact. The results of work depend not only on technical aspects, but also on communication, organization of work, and management decisions.
Traditional methods of productivity evaluation are limited. They usually use simple indicators or expert opinions. They do not fully capture relationships between factors, uncertainty, or changes over time. Therefore, data analysis methods are needed for a more complete evaluation of productivity in a digital project environment.
The research uses statistical, probabilistic, regression, machine learning, and expert-based methods. Special attention is given to the interpretation of results in the form of management metrics. These include schedule deviation, cost deviation, team capacity, and risk of project delay.
The main part of the dissertation consists of an introduction, five chapters, and general conclusions.
Chapter 1 analyzes modern approaches to socio-technical systems and methods of productivity evaluation in project management. Key factors that influence efficiency in the digital environment are identified.
Chapter 2 studies the use of probabilistic methods for productivity evaluation under uncertainty and incomplete data.
Chapter 3 examines regression analysis for evaluating the influence of different factors on productivity. The results are interpreted through schedule and cost deviations.
Chapter 4 studies the use of machine learning methods, especially LSTM models, for forecasting productivity over time. These models help to evaluate team capacity changes and project risks.
Chapter 5 analyzes the use of the Bayesian Truth Serum method to identify causes of productivity changes. The method is based on differences between actual answers and expected answers of respondents.
The main results of the research are:
– systematization of methods for productivity analysis in project environments;
– formation of a system of productivity indicators based on digital project data;
– justification of the use of statistical, probabilistic, and machine learning methods for productivity evaluation and forecasting;
– interpretation of analytical results through management metrics;
– identification of the role of social factors in productivity changes based on expert data.
The scientific novelty consists in forming an analytical framework for evaluating, forecasting, and interpreting productivity in a digital project environment with a focus on management decision support and the interaction between social and technical components.
The practical significance of the results lies in their possible use for management decision support. This includes planning, resource management, team coordination, and risk evaluation. The results can be used in IT companies and project teams that use modern project management systems.
The practical value is confirmed by testing the results in an IT company and by presenting them at scientific events.
The results can be used in further research related to the development of analytical methods for managing socio-technical systems in the digital economy.
