Replication Data for: On the role of ecological validity in language and speech research

Dataset abstract This dataset contains the results from 40 language and speech researchers, who completed a survey. In the first part of the survey, respondents were asked to complete a demographic (e.g., age, gender, first language) and professional background questionnaire (e.g., current academic position, research interests). In addition, they were asked several open-ended questions about their familiarity with and understanding of the term ‘ecological validity’ (e.g., which words come to mind when you hear this term, how to measure the ecological validity of a study, how does ecological validity apply to your area of research). In the second part of the survey, respondents were presented with 24 short speech excerpts, representing 12 different stimulus types. They were asked to rate each speech excerpt on its degree of casualness (i.e. spontaneity) and naturalness, and how likely they are to encounter each excerpt in everyday listening situations. Article abstract This paper explores how researchers in the field of language and speech sciences understand and apply the concept of ecological validity. It also assesses the ecological validity of various stimulus materials, ranging from isolated word productions to sentences taken from authentic interviews. Forty researchers participated in a survey, which contained (i) a demographic and professional background questionnaire with open-ended questions about the definition, feasibility and desirability of ecological validity, and (ii) a speech rating task. In the rating task, respondents evaluated 24 speech excerpts, representing 12 types of stimulus materials, on their casualness, naturalness, and likelihood of occurrence in real-life contexts. The results showed that while most researchers acknowledge the importance of ecological validity, defining the necessary and sufficient criteria for evaluating or achieving it remains challenging. Regarding stimulus types, unscripted sentences from interviews and Map Task dialogues were rated as the most casual and natural. In contrast, carefully read sentences and digitally modified stimuli were viewed as the least casual and natural, although individual differences in rating were noticeable. Similarly, ratings for the likelihood of occurrence in everyday listening situations were highest for various types of extemporaneous speech. The survey responses not only enhance our theoretical understanding of ecological validity, but also raise awareness about the implications of methodological choices, such as the selection of tasks and stimulus materials, on the ecological validity of a study.数据集摘要 本数据集收录了40名语言与语音研究者完成的一项调研的全部结果。调研分为两个部分：第一部分要求受访者填写人口统计（如年龄、性别、母语）与职业背景问卷（如当前学术职位、研究兴趣）；此外，调研还设置了若干开放式问题，用以了解受访者对“生态效度（ecological validity）”这一术语的熟悉程度与理解情况，例如“听到该术语时你会联想到哪些词汇”“如何衡量一项研究的生态效度”“生态效度如何应用于你的研究领域”。调研的第二部分中，受访者会聆听24段简短语音片段，涵盖12种不同的刺激类型。研究者要求受访者对每段语音片段的口语化程度（即自发性）、自然度，以及在日常收听场景中遇到该片段的可能性进行评分。 论文摘要 本论文探讨了语言与语音科学领域的研究者如何理解并应用生态效度这一概念，同时评估了各类刺激材料的生态效度，材料范围涵盖从孤立单词发音到真实访谈语句的多种形式。共有40名研究者参与本次调研，调研内容包含两部分：一是人口统计与职业背景问卷，附带若干关于生态效度的定义、可行性与可取性的开放式问题；二是语音评分任务。在语音评分任务中，受访者针对24段涵盖12种刺激材料类型的语音片段，从口语化程度、自然度以及在真实场景中出现的可能性三个维度进行评估。研究结果显示，尽管多数研究者认可生态效度的重要性，但界定用于评估或达成生态效度的充要条件仍颇具挑战。就刺激类型而言，访谈中的无脚本语句与地图任务（Map Task）对话被评为最具口语化与自然度的材料；与之相对，刻意朗读语句与数字修改后的刺激材料则被认为最不具备口语化与自然度，尽管评分存在显著的个体差异。同样，在日常收听场景中出现的可能性评分中，各类即兴语音的得分最高。本次调研的结果不仅加深了我们对生态效度的理论认知，也提升了学界对研究方法选择（如任务与刺激材料的选取）对研究生态效度的影响的重视程度。