Work

Development of a multilevel classification model for selecting the type of hybrid work for information technology projects

Year: 2026
Type: article
Abstract: This study investigates the process of choosing the type of hybrid work for projects in the field of information technology (IT). The task addressed relates to the lack of a substantiated toolkit for selecting the type of hybrid work based on factors of different levels. Most studies analyze individual groups of factors only. This work proposes a methodological approach to building a classification model for choosing the type of hybrid work, which includes five stages. At the first stage, a factor base is built based on literature analysis using the Term Frequency-Inverse Document Frequency (TF-IDF) method and cosine similarity. Next, factors are prioritized using the analysis of hierarchies method and empirical data are collected (n = 179). After that, significant factors are statistically selected using the Least Absolute Shrinkage and Selection Operator (Lasso); the Classification and Regression Trees (CART) model is then built. The multilevel structure of factors has been justified on the basis of ecological theory, life field theory, as well as the competency model. As a result, 58 factors were formed, of which 15 statistically significant ones were selected. A classification tree was constructed with an accuracy of 74.1%, containing 6 rules for choosing the type of hybrid work. The results resolve the task set by combining a multilevel approach, automated selection of factors, and the formation of interpreted rules. It has been shown that the type of hybrid work is determined not by individual factors but by their configuration – the balance of flexibility and structure. The above is attributed to the nonlinear interactions among factors and the concentration of their influence in several subgroups. The approach could be used by IT project managers and HR specialists to choose the type of hybrid work provided data on the actual state of the factors are available
Language: English
Open Access status: gold
Estimated APC paid: $242
OpenAlex ID: W7204994477
FWCI:
Cites: 22
Cited by:
Related to: 10