This review investigated whether ceiling effects on client-reported working alliance measures represent measurement artifacts or valid information related to the formation of the working alliance. Using data from two previously published meta-analyses, a total of 92 estimates of ceiling effects were calculated based on 37 studies with 6,439 participants. Analyses examined the size of ceiling effects, relation with demographic variables, type of alliance measure, and ceiling effect stability across sessions. Moderate to large ceiling effects appeared across score distributions of multiple measures of client-rated alliance, across time administered, and across different sample characteristics such as gender, age, and ethnicity. When examined with the Session Rating Scale (SRS), analyses indicated ceiling effects had a moderate correlation with session number. When SRS ceiling effects were examined in a single study with a large sample of complete cases (N = 2,990) across seven sessions, large initial ceiling effects continued to increase slightly in size across sessions. Higher ceiling effects were also observed with the Working Alliance Inventory. Given the prevalence and relative stability of ceiling effects on score distributions, working alliance scores do not exhibit the characteristics of a normally distributed continuous variable. While the working alliance has typically been defined in terms of theoretical content such as tasks, goals, and bond, study findings suggest another key element may be a threshold structure where clients shift to an experience of the therapeutic relationship as established. Discussion focuses on directions for alliance research and clinical practice as well as study limitations. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

中文翻译：

---

天花板效应表明工作联盟可能存在门槛结构。

本次审查调查了客户报告的工作联盟措施的上限效应是否代表了测量工件或与工作联盟的形成相关的有效信息。使用先前发表的两项荟萃分析的数据，根据 37 项研究（涉及 6,439 名参与者）总共计算出 92 项上限效应估计值。分析检查了上限效应的大小、与人口变量的关系、联盟措施的类型以及跨会议的上限效应稳定性。客户评价联盟的多种衡量标准的分数分布、管理时间以及不同样本特征（例如性别、年龄和种族）都出现了中等到大的天花板效应。当使用会话评级量表（SRS）进行检查时，分析表明上限效应与会话数具有中等相关性。当在一项研究中检查 SRS 上限效应时，在七个会话中使用大量完整病例样本 (N = 2,990)，发现较大的初始上限效应在整个会话中继续略有增加。工作联盟库存也观察到了更高的上限效应。鉴于分数分布上限效应的普遍性和相对稳定性，工作联盟分数并不表现出正态分布连续变量的特征。虽然工作联盟通常是根据任务、目标和联系等理论内容来定义的，但研究结果表明，另一个关键要素可能是一个阈值结构，在这个阈值结构中，客户转向建立治疗关系的体验。讨论重点是联盟研究和临床实践的方向以及研究局限性。（PsycInfo 数据库记录 (c) 2022 APA，保留所有权利）。