Technical, Relational, & Conditional Process Models of MI Efficacy: Meta-Analysis

描述（由申请人提供）：此R21申请响应PAR-14-053。对治疗开发者提出的理论模型进行检验的研究，对于增强现有循证行为成瘾治疗的影响是一种有价值的方法（Magill & Longabuagh, 2013）。目的是通过经验推导出对治疗效果负责的因果过程，结果是确定：1)关键治疗行为，2)技能熟练程度阈值，3)关键近端客户结果，以及4)这些原则最有效的条件（即条件过程模型）。动机访谈（MI）可以说是进行这类研究的最重要的行为治疗方法。最近，Magill及其同事（Magill et al., 2013; MS under revision /resubmission）对Miller和Rose（2009）提出的MI疗效技术模型的具体途径进行了meta分析。虽然这项研究为MI如何影响行为改变的新兴图景做出了贡献，但还需要一个额外的、更全面的元分析研究来测试完整的理论框架，以最终确定哪些因果途径得到支持，在什么条件下，以及如果有的话，哪些需要进行重大的理论修订。拟议的项目将实现这一目标，不仅有更大的初级研究样本，而且还有一个由原始数据而不是已发表和可用数据组成的数据库。该项目将首先编制一个全面的二元相关性数据集，以实现所有科学目标。接下来，它将测试技术、关系和一系列条件过程模型（例如，通过客户和治疗实施因素）在总体、元分析水平上的MI疗效。最后，这个大规模的、初级的研究数据集将使一个探索性的、心理测量性的评估成为可能，该评估包括观察评定的、会话内总结得分以及个体治疗师和客户行为代码。本研究结果的广度将为心肌梗死证据基础的以下三个关键领域提供信息：a)心肌梗死实施的临床指南，b)心肌梗死因果理论的知识，以及c)未来心肌梗死过程研究的经验衍生测量指南。

DESCRIPTION (provided by applicant): This R21 application responds to PAR-14-053. Research examining the theoretical models proposed by treatment developers is a valuable approach to enhancing the impact of existing evidence-based behavioral addictions treatments (Magill & Longabuagh, 2013). The intent is to empirically-derive the causal processes that are responsible for treatment benefit, and the result is identification of: 1) key therapeutic behavior, 2) skill proficiency thresholds, 3) key proximal client outcomes, and 4) the conditions under which these principles are most operative (i.e., conditional process models). Motivational interviewing (MI) is arguably the most important behavioral treatment upon which to conduct this kind of research. Recently, Magill and colleagues (Magill et al., 2013; MS under revise/resubmission) conducted a meta-analysis of studies testing specific pathways of the Technical Model of MI Efficacy as proposed by Miller and Rose (2009). While this research contributes to an emerging picture of how MI affects behavior change, an additional, more comprehensive meta-analytic study is needed to test the full theoretical framework, to identify conclusively which causal pathways have support, under what conditions, and which, if any, require major theoretical revision. The proposed project will address this aim not only with a larger sample of primary studies, but also with a database comprised of raw, rather than published and available, data. The Project will first compile a comprehensive dataset of bivariate correlations for fulfillment of all scientific aims. Next, it will test the Technical, Relational, nd a series of conditional process models (e.g., by client and treatment implementation factors) of MI efficacy at an aggregate, meta-analytic level. Finally, this large-scale, primary study dataset wil enable an exploratory, psychometric assessment of observationally-rated, within-session summary scores as well as individual therapist and client behavior codes. The breadth of findings from this study will inform the following three key areas of the MI evidence base: a) clinical guidelines for MI implementation, b) knowledge on MI causal theory, and c) empirically-derived measurement guidelines for future MI process research.