Department of Educational and Counseling Psychology, University of Louisville, Louisville, KY 40292, USA.
Psychotherapy (Chic). 2012 Jun;49(2):152-62. doi: 10.1037/a0023990. Epub 2011 Oct 3.
Psychotherapy research examining the influence of psychotherapists on clients' clinical outcomes can provide valuable insights for enhancing psychotherapists' effectiveness. Psychotherapy data often have a hierarchical structure--multiple clients treated by the same psychotherapist. As such, researchers are more commonly turning to the use of advanced statistical methods, namely multilevel modeling (MLM), to address this complexity. In this article, we describe MLM for consumers of psychotherapy studies so they can better understand and evaluate studies that employ this method. We provide an example study that illustrates how traditional statistical methods may conceal meaningful findings. Also, we describe commonly utilized applications of MLM in psychotherapy research, such as variation among psychotherapist effects in outcomes, variation in the relationship between predictors and outcomes that can be attributed to psychotherapists, and sample size concerns.
研究心理治疗师对患者临床结果的影响可以为提高心理治疗师的效果提供有价值的见解。心理治疗数据通常具有层次结构——多位客户由同一位心理治疗师治疗。因此，研究人员更倾向于使用高级统计方法，即多层次建模 (MLM)，来解决这种复杂性。在本文中，我们为心理治疗研究的消费者描述了 MLM，以便他们能够更好地理解和评估使用这种方法的研究。我们提供了一个示例研究，说明了传统统计方法如何可能掩盖有意义的发现。此外，我们还描述了 MLM 在心理治疗研究中的常见应用，例如治疗效果之间的心理治疗师差异、可归因于心理治疗师的预测因素与结果之间关系的变化，以及样本量问题。
