Theory Use In Software Engineering: A Systematic Mapping Study

Context: The use of theories in Software Engineering (SE) is not as common or explicit as in other fields, with most studies focusing on technical aspects. However, establishing a more robust theoretical foundation could significantly contribute to the maturation of SE as a science.

Objective: This study investigates the use of theory in SE by providing a longitudinal mapping of 174 primary studies published between 1989 and 2025. It analyzes the extent of theory use, disciplinary origins, types, and roles.

Method: We conducted a systematic mapping study that combined an automated search with structured snowball sampling. We classified evidence based on Gregor’s theory types, Hannay’s theory roles, and SWEBOK knowledge areas.

Results: We identified 159 distinct theories; however, the field relies heavily on a small core, as 85% of theories are imported (mainly from Psychology and Information Systems), and only 18 theories appear in three or more studies. While "Theory" is the dominant label (75%), terminology remains inconsistent. In terms of usage, 53% of studies use theory for Design (framing research), while only 13% explicitly test theoretical propositions.

Conclusion: Despite an increase in SE-native proposals after 2016, the discipline primarily uses theory to justify or motivate research rather than to accumulate knowledge through testing and refinement. There is a critical need to move from ad-hoc borrowing to the development and operationalization of SE-specific frameworks, particularly in technical knowledge areas.