Rounsaville B J, Kleber H D, Wilber C, Rosenberger D, Rosenberger P
Am J Drug Alcohol Abuse. 1981;8(1):51-69. doi: 10.3109/00952998109016918.
This study compares self-reports by a sample of 50 opiate addicts with reports of 50 "significant other" informants, on past and current functioning. Moderate to high levels of agreement are found using the Intraclass Correlation Coefficient (ICC) in reporting current opiate use, current employment, and current legal pressures. This finding is consistent with previous studies which found good agreement between addicts and informants in these areas. However, much poorer agreement between addicts and informants is found in assessing the severity of addict's social, employment, and psychological problems, and in assessing the more detailed aspects of drug use. A comparison of the percent agreement data with ICC data used in this study provides empirical support for the argument that percent agreement data represents an inflated estimate of concordance between raters. These findings suggest that addicts and their families have rather divergent perceptions of the addict's problems and, in addition, indicate the value of including family members or significant others in the process of evaluating and treating opiate addicts. The results of the statistical analysis underscore the need in reliability studies of rigorous statistics that take into account chance agreement.
本研究比较了50名阿片类成瘾者样本的自我报告与50名“重要他人” informant的报告，内容涉及过去和当前的功能状况。在报告当前阿片类药物使用情况、当前就业情况和当前法律压力方面，使用组内相关系数（ICC）发现了中度到高度的一致性。这一发现与之前的研究一致，即在这些领域成瘾者和 informant之间存在良好的一致性。然而，在评估成瘾者的社会、就业和心理问题的严重程度以及评估药物使用的更详细方面时，成瘾者和 informant之间的一致性要差得多。将本研究中使用的百分比一致性数据与ICC数据进行比较，为百分比一致性数据高估了评分者之间的一致性这一论点提供了实证支持。这些发现表明，成瘾者及其家人对成瘾者问题的看法存在相当大的差异，此外，还表明在评估和治疗阿片类成瘾者的过程中纳入家庭成员或重要他人的价值。统计分析结果强调了在可靠性研究中需要采用考虑到偶然一致性的严格统计方法。
