Abstract To ensure valid measurement in large-scale assessments, avoiding incorporating construct-irrelevant aspects is crucial. We investigated a potential source of construct-irrelevant variance, i.e. the presentation mode of essays (handwritten vs. computer-typed) and its influence on scoring. Further, we investigated whether the presentation-mode effect is moderated by text quality and legibility, as well as by orthographic, grammatical, and punctuation erroneousness, and whether the presentation mode differs over text genres. Rater-teams scored two versions (originally handwritten vs. computer-typed transcripts) of 430 essays written by students at the end of lower secondary education in Germany on content quality and style quality, as well as the considered moderators. Statistical analyses showed a main effect for presentation mode: Computer-typed transcripts were scored higher than handwritten essays. This effect was moderated by text quality and grammatical erroneousness. Furthermore, the strength of the presentation-mode effect differed by text genre. We argued that the strength as well as the direction of the presentation mode are guided by these and further factors.

中文翻译：

---

大规模写作评估中的呈现模式效应

摘要 为了确保大规模评估中的有效测量，避免纳入与结构无关的方面至关重要。我们调查了与结构无关的差异的潜在来源，即论文的呈现方式（手写与计算机打字）及其对评分的影响。此外，我们调查了演示模式效果是否受文本质量和易读性以及正字法、语法和标点错误的影响，以及演示模式是否因文本类型而异。评分团队对德国初中教育结束时学生撰写的 430 篇文章的内容质量和风格质量以及考虑的版主的两个版本（最初是手写的还是计算机打字的成绩单）进行评分。统计分析显示了演示模式的主要影响：计算机打字的成绩单比手写的论文得分更高。这种影响受到文本质量和语法错误的影响。此外，演示模式效果的强度因文本类型而异。我们认为，演示模式的强度和方向受这些和其他因素的指导。