Do the Test Smells Assertion Roulette and Eager Test Impact Students' Troubleshooting and Debugging Capabilities?

TL;DR

This study investigates how two common test smells, Assertion Roulette and Eager Test, negatively affect undergraduate students' debugging efficiency and troubleshooting skills in software testing.

Contribution

It provides empirical evidence on the impact of specific test smells on students' debugging performance, highlighting the need for better test case quality education.

Findings

Students take longer to fix errors with test smells.

Assertion Roulette has a greater negative impact than Eager Test.

Test smells hinder effective troubleshooting and debugging.

Abstract

To ensure the quality of a software system, developers perform an activity known as unit testing, where they write code (known as test cases) that verifies the individual software units that make up the system. Like production code, test cases are subject to bad programming practices, known as test smells, that hurt maintenance activities. An essential part of most maintenance activities is program comprehension which involves developers reading the code to understand its behavior to fix issues or update features. In this study, we conduct a controlled experiment with 96 undergraduate computer science students to investigate the impact of two common types of test smells, namely Assertion Roulette and Eager Test, on a student's ability to debug and troubleshoot test case failures. Our findings show that students take longer to correct errors in production code when smells are present in their associated test cases, especially Assertion Roulette. We envision our findings supporting academia in better equipping students with the knowledge and resources in writing and maintaining high-quality test cases. Our experimental materials are available online: https://wajdialjedaani.github.io/testsmellstd/