INTERVIEWER SEVERITY RATINGS AND COMPOSITE SCORES OF THE ASI - A FURTHER LOOK
INTERVIEWER SEVERITY RATINGS AND COMPOSITE SCORES OF THE ASI - A FURTHER LOOK
复制标题
DOI：
10.1016/0376-8716(94)90157-0
复制
发表时间：
1994-02-01
影响因子：
4.2
通讯作者：
MCKAY, JR
中科院分区：
文献类型：
--
作者：
ALTERMAN, AI;BROWN, LS;MCKAY, JR
对新培训的访谈者进行成瘾严重程度指数（ASI）、访谈者严重程度评分（ISRs）和综合评分（CSs）的心理测量特征检测。对9名美沙酮维持（MM）试点受试者进行7个评分者的互译信度测定。我们发现这些基因对CSs非常好，但对isr只有中等水平。对407例美沙酮维持患者的数据进行回归分析，将各区域的所有变量作为自变量，ISR作为因变量。这些分析表明，平均而言，55-60%的isr差异得到了解释。然而，法律量表63%的方差得到了解释，而药物量表只有38%的方差得到了解释。受试者对治疗的需要或问题的严重性的评价是最大的差异，除了药物和酒精领域。然后使用Cronbach's标准化alpha统计量对该样本的综合分数的内部一致性进行检验。研究发现，从药物量表的0.62到酒精和精神量表的0.87，这些分数大体上是令人满意的。药物量表的平均项目间相关系数为0.11，表明项目同质性相对较低。最后，计算每个尺度的CSs和isr之间的相关性。除就业区相关系数仅为0.08外，其余均为中高相关（0.53-0.78）。研究结果在ISRs和CSs的性质和局限性方面进行了讨论。
The psychometric characteristics of the Addiction Severity Index (ASI) interviewer severity ratings (ISRs) and composite scores (CSs) were examined for a newly trained group of interviewers. The interrater reliabilities of seven raters for 9 methadone maintenance (MM) pilot subjects were determined. These were found to be excellent for the CSs, but only moderate for the ISRs. Regression analyses were performed on the data of 407 methadone maintenance patients entering all of the variables in each area as the independent variables and the ISR as the dependent variable. These analyses indicated that, on average, 55-60% of the variance in ISRs was explained. However, while 63% of the variance was explained for the legal scale only 38% of the variance of the drug scale was explained. The subject's rating of either the need for treatment or the seriousness of the problem accounted for the most variance, with the exception of the drug and alcohol areas. The internal consistency of the composite scores was then examined for this sample using Cronbach's standardized alpha statistic. These were found to be generally satisfactory ranging from 0.62 for the drug scale to 0.87 for the alcohol and psychiatric scales. A mean interitem correlation of 0.11 for the drug scale was obtained suggesting relatively low item homogeneity. Finally, correlations between the CSs and ISRs were calculated for each scale. Moderate to high relationships were found (0.53-0.78) with the exception of the correlation for the employment area which was only 0.08. The findings are discussed in terms of the nature of and limitations of the ISRs and CSs.