Controlling for Endogeneity Biases in Estimating AA Effects on Drinking Outcomes

描述（由申请人提供）：互助的随机对照试验（rct）极难进行。现有的，主要是相关的，匿名戒酒协会（AA）有效性的研究容易受到两种主要人为因素的影响，这是由于变量的潜在内生性偏差，定义为当被视为外源性的变量实际上可能是内源性的。也就是说，他们假定忽略了需要在因果模型中明确表示的原因；这也被称为规范错误。其中一个人工产物是自我选择偏差，即不同类型的人选择参加或不参加AA。这方面的一个例子是参加嗜酒者互诫会和饮酒之间的虚假关系，因为那些有戒酒动机的人更有可能参加嗜酒者互诫会并戒酒；也就是说，高动机可以很好地“解释”观察到的AA参与和饮酒行为之间的相关性。第二种可能的人为因素是反向因果关系，即，当减少或停止饮酒导致增加的AA参与。通过控制这些可能的内生性偏差，确定参加AA和饮酒行为之间关系的真实影响程度和因果方向是至关重要的。本研究的总体目标是使用三种统计技术对项目MATCH酒精中毒治疗数据库中的参加AA和饮酒行为进行二次分析，这些技术旨在控制相关数据中潜在的内生性偏差：倾向得分匹配（PSM），工具变量（IV）分析和结构方程建模（SEM）的面板数据与交叉滞后效应。模拟建模也将进行，以测试分析的鲁棒性与违反假设。研究结果可以极大地加强匿名戒酒会和12步促进疗法的证据基础，并为康复社区提供额外的证据，有助于鼓励消费者继续参与互助。

DESCRIPTION (provided by applicant): Randomized controlled trials (RCTs) of mutual aid are extremely difficult to conduct. The existing, mainly correlational, studies of Alcoholics Anonymous (AA) effectiveness are susceptible to two main kinds of artifacts due to potential endogeneity bias in variables, defined as when variables treated as exogenous may actually be endogenous. That is, they have presumed omitted causes that need to be explicitly represented in a causal model; this is also termed specification error. One such artifact is self-selection bias, i.e., where different types of people choose to participate or not participate in AA. An example of this would be a spurious relation between AA participation and drinking due to the possibility that individuals who develop motivation to stop drinking are more likely both to attend AA and to stop drinking; that is, high motivation could well "explain" the observed correlation between AA participation and drinking behavior. The second possible type of artifact is reverse causation, i.e., when reducing or stopping drinking leads to increased AA participation. It is essential to determine the true magnitude of effect and causal direction in the relation between AA participation and drinking behavior by controlling for these possible endogeneity biases, The overall objective of the study is to conduct a secondary analysis of AA participation and drinking behavior in the Project MATCH alcoholism treatment database using three statistical techniques which are designed to control for potential endogeneity biases in correlational data: propensity score matching (PSM), instrumental variable (IV) analysis, and structural equation modeling (SEM) of panel data with cross-lagged effects. Simulation modeling will also be conducted to test the robustness of the analyses with respect to violation of assumptions. The results could greatly strengthen the evidence base for Alcoholics Anonymous and 12 step facilitation therapy, as well as provide the recovery community with additional evidence useful for encouraging consumers to stay involved in mutual aid. PUBLIC HEALTH RELEVANCE: The results of this study could increase the utilization of 12 step facilitation therapies and 12 step mutual aid programs such as Alcoholics Anonymous to reduce public health and social problems attributable to excessive and hazardous alcohol consumption.