The selection of variables in a given experiment is crucial, since it is the theoretical foundation that guides how data should be collected and analyzed. However, selecting variables is an intricate activity, especially considering areas such as Software Engineering and Education, whose studies should also consider human-related variables in the design. In this scenario, we aim to investigate how a support mechanism helps in the variables selection activity of the experiment process. To do so, we conducted a preliminary study on the use of an experimental framework composed of a catalog of variables. We explored the domain of the integration of software testing into programming education. Participants were divided into two groups (ad hoc and framework support) and asked to select variables for a given experiment goal. We analyzed the results by identifying threats to validity in their experimental design drafts. Results show a significant number of threats of type inadequate explication of constructs for both groups. Nonetheless, the framework helped to increase the clarity of concepts selected as variables. The cause of most raised threats, even with the framework support, was an inaccuracy in selecting the values of such variables (i.e. treatments and fixed values).

中文翻译：

---

使用实验框架支持变量选择：一项探索性研究

给定实验中变量的选择至关重要，因为它是指导如何收集和分析数据的理论基础。然而，选择变量是一项复杂的活动，特别是考虑到软件工程和教育等领域，其研究还应在设计中考虑与人类相关的变量。在这种情况下，我们旨在研究支持机制如何帮助实验过程的变量选择活动。为此，我们对使用由变量目录组成的实验框架进行了初步研究。我们探索了将软件测试集成到编程教育中的领域。参与者被分为两组（临时和框架支持），并要求为给定的实验目标选择变量。我们通过在他们的实验设计草案中识别对有效性的威胁来分析结果。结果显示，对于这两个群体，大量类型的威胁解释不充分。尽管如此，该框架有助于提高被选为变量的概念的清晰度。大多数提出威胁的原因，即使有框架支持，也是在选择这些变量的值（即处理和固定值）时不准确。