论文部分内容阅读
目的对2005年至今发表的有关人际需求问卷(INQ)的文献进行信度概化分析,为INQ在自杀预防研究中的应用提供描述性信度水平及影响因素的参考。方法在Embase等数据库检索到以INQ为测量工具发表的文献173篇,对最终符合筛选标准的137篇文献进行分析。采用α系数效果量法检测INQ两个维度的信度水平,并以被试数量等8个因素作为预测变量,通过加权回归分析探究影响因素。结果 INQ两个维度的描述性信度水平均在0.86以上;题目数量、被试国籍、被试数量和性别比例能显著预测TB维度的α系数(β=-0.301、0.524、-0.426,P<0.01;β=-0.151,P<0.05),被试数量能显著预测PB维度的α系数(β=-0.310,P<0.01)。结论 INQ能满足自杀预防研究对测量稳定性的要求;建议在使用时具体报告两个维度的α系数,慎选题目过少的简化版本,选取足量被试(女性比例不能过低),同时考虑不同国籍对信度的影响。
OBJECTIVE To analyze the reliability of literature about INQ published from 2005 to now, and to provide a reference for the reliability and the reliability of INQ in suicide prevention research. Methods A total of 173 articles published in the Embase database were retrieved using INQ as a measuring tool, and 137 articles that ultimately met the screening criteria were analyzed. The α-factor method was used to test the reliability of two dimensions of INQ. Eight factors including the number of subjects were used as predictors, and the influencing factors were explored by weighted regression analysis. Results The descriptive reliability of INQ in both dimensions was above 0.86. The number of subjects, the nationality of subjects, the number of subjects and the sex ratio significantly predicted α-factor of TB dimension (β = -0.301, 0.524, -0.426, P < 0.01; β = -0.151, P <0.05). The number of test subjects could significantly predict the α coefficient of PB dimension (β = -0.310, P <0.01). Conclusions INQ can meet the requirements of suicide prevention research on measurement stability. It is suggested that α-coefficients of two dimensions should be reported in detail. Prudent versions with too few titles should be carefully selected, and sufficient numbers of subjects should be selected (female ratio should not be too low) Consider the impact of different nationalities on the reliability.