Evaluation techniques

Combining experimental study with user study

Our evaluation consists of two parts: (1) an experimental study that assesses whether the search-based technique can improve the expert’s manual assignments, and (2) a user study in which we conducted interviews with experts (GSD project managers) to get feedbacks and assess the feasibility of the approach.

Specifically, we investigated the following research questions:

• RQ1. How cost-effective is a GA-based task scheduling technique, in the context of GSD? - RQ1.1. Can it improve the expert’s manual assignments, in terms of total project duration?
  • RQ1.2. How costly is it to run the GA-based approach?
• RQ2. How valid and feasible are the solutions provided by a GA-based task scheduling technique, from the human expert’s perspective?

...

Source: An empirical study of search-based task scheduling in global software development; Proceedings - 2017 IEEE/ACM 39th International Conference on Software Engineering: Software Engineering in Practice Track, ICSE-SEIP 2017