Complementarity in Software Code Complexity Metrics

Hao Gao

doi:10.1016/j.jss.2025.112679

Complementarity in Software Code Complexity Metrics

2025-11-01

0

PRE

AI

DOI：10.1016/j.jss.2025.112679

Original

Original request for help

Save

Abstract

En 中文

• Evaluates how well static code metrics reflect EEG-measured cognitive load • Finds Halstead Effort and Difficulty correlate better with mental effort than V(g) • Shows that no single metric fully explains perceived code complexity • Combines metrics via Gaussian Process Regression to improve prediction (R 2 = 0.87) • Supports a hybrid approach for holistic code understandability assessment

Complementarity in Software Code Complexity Metrics

Abstract

Journal

Organization