Lindström Elisabeth, Lundström Cecilia, Sand Anders, Samson Ineke
Department of Clinical Science, Intervention and Technology, Division of Speech and Language Pathology, Karolinska Institutet, Stockholm, Sweden.
Faculty of Arts, Psychology and Theology, Logopedics, Åbo Akademi University, Turku, Finland.
Logoped Phoniatr Vocol. 2025 Apr;50(1):28-36. doi: 10.1080/14015439.2024.2338084. Epub 2024 Apr 21.
Stuttering is a communication disorder that involves both manifest speech disfluencies and associated symptoms. The purpose of the present study was to introduce an easily administered and ecologically valid assessment tool designed for perceptual evaluation of stuttered speech, FreDESS (frequency of stuttering events, duration of events, effort, secondary behaviours, and severity). More specifically, we wanted to study its reliability and validity.
Video recordings of conversations with 38 people who stutter (PWS), 19 females and 19 males aged 13-25, were assessed by three speech language pathologists (SLP). Inter- and intrajudge reliability was estimated with intraclass correlation, standard error of measurement, and agreement between listeners. Internal consistency for the FreDESS parameters was estimated with Cronbach's alpha (α). To test the validity of FreDESS, the relationships between the average estimated parameters of frequency, duration, secondary behaviours, and severity of FreDESS and the Stuttering Severity Instrument (SSI-3), were analysed using intraclass correlation.
The interjudge reliability was good, especially for the frequency, duration, and severity parameters (90 per cent + agreement given 1 scale point difference). All parameters of the FreDESS had strong intrajudge reliability (ICC = 0.86-0.94) and the overall internal consistency was high (α = 0.98). The average ratings on the two assessment scales were in line with each other ( = 0.90-0.96), indicating high concurrent validity.
The FreDESS scale for the assessment of stuttered speech may be a valuable tool in clinical and research contexts. It is a valid and more time-efficient assessment instrument than the more commonly used SSI.
口吃是一种涉及明显言语不流畅及相关症状的交流障碍。本研究的目的是介绍一种易于实施且具有生态效度的评估工具，用于对口吃言语进行感知评估，即口吃事件频率、事件持续时间、努力程度、伴随行为及严重程度评估量表（FreDESS）。更具体地说，我们想研究其信度和效度。
由三名言语语言病理学家（SLP）对38名口吃者（PWS）（19名女性和19名男性，年龄在13至25岁之间）的对话视频记录进行评估。通过组内相关系数、测量标准误差和听众间一致性来估计评判者间和评判者内信度。使用克朗巴哈α系数（α）估计FreDESS参数的内部一致性。为了检验FreDESS的效度，使用组内相关系数分析FreDESS的频率、持续时间、伴随行为和严重程度的平均估计参数与口吃严重程度量表（SSI - 3）之间的关系。
评判者间信度良好，尤其是频率、持续时间和严重程度参数（给定1个量表点差异时一致性达90%以上）。FreDESS的所有参数均具有较高的评判者内信度（ICC = 0.86 - 0.94），总体内部一致性较高（α = 0.98）。两个评估量表的平均评分相互一致（ = 0.90 - 0.96），表明具有较高的同时效度。
用于评估口吃言语的FreDESS量表在临床和研究环境中可能是一种有价值的工具。它是一种比更常用的SSI更有效且更省时的评估工具。
