An Empirical Study Investigating the Predictors of Software Metric Correlation in Application Code and Test Code.

Abstract

On non-trivial software, the large test code base needs adequate maintenance similarly to the application code. Using complexity metrics on large software reveals that Application code is more complex than test code but not necessarily so much more, Test code is not as simple as it should be and may therefore be very complex to maintain, and open source software are not adequately tested. While a number of authors hypothesize and experimentally confirm that CC has a very strong correlation with LOC, justifying the use of LOC in place of CC (and Halstead Effort), this strong correlation is prevalent only in production code as results indicate a very weak correlation in test code between LOC, CC and Halstead Effort. The kind of code, the kind of software and the kind of metric determines the extent of monotonicity software metrics and would be inappropriate to substitute one metric with another.