Department of Management, University of Antwerp, Antwerp, Belgium.
Department of Training and Education Sciences, University of Antwerp, Antwerp, Belgium.
Behav Res Methods. 2022 Aug;54(4):1976-1988. doi: 10.3758/s13428-021-01707-1. Epub 2021 Nov 23.
Researchers often decide on the number of trials included in an experiment without adhering to an empirical method or framework. This might compromise generalizability and unnecessarily increase participant burden. In this article we want to put forward generalizability theory as a guide for task reduction. We will use a sentence production task to demonstrate how a generalizability and a decision study can help researchers to estimate the minimum number of trials and of items per trial that are necessary to generalize over trials. We obtained writing process data for 116 participants. Each of them completed a sentence production task that had 40 trials. Pause times between and within all words, target nouns and target verbs were logged with the keystroke logging tool ScriptLog. Results demonstrate that generalizability theory can serve as an empirical framework to ensure generalizable measurements on the one hand, and reduce participant burden to a minimum on the other. This finding is particularly valuable for studies with vulnerable target groups, such as participants suffering from aphasia, dyslexia or Alzheimer's disease.
研究人员在进行实验时，经常会决定纳入的试验数量，而不遵循经验方法或框架。这可能会影响推广性，并不必要地增加参与者的负担。在本文中，我们希望提出可推广性理论作为减少任务的指南。我们将使用句子生成任务来演示如何进行可推广性和决策研究，以帮助研究人员估计在试验之间进行推广所需的最小试验次数和每个试验的项目数。我们获得了 116 名参与者的写作过程数据。他们每个人都完成了一个有 40 个试验的句子生成任务。使用按键记录工具 ScriptLog 记录了所有单词、目标名词和目标动词之间和之内的停顿时间。结果表明，可推广性理论可以作为一个经验框架，一方面确保可推广的测量，另一方面将参与者的负担降至最低。这一发现对于研究脆弱的目标群体（如患有失语症、阅读障碍或阿尔茨海默病的参与者）的研究特别有价值。
