Stanley M A, Beck J G, Zebb B J
University of Texas Health Science Center, Houston 77030-3497, USA.
Behav Res Ther. 1996 Oct;34(10):827-38. doi: 10.1016/0005-7967(96)00064-2.
Despite relatively high prevalence rates of anxiety disorders in older adults, little attention has been paid to the establishment of psychometrically sound measures for this population. The current study addresses this issue by examining the psychometric properties of four self-report anxiety measures: the Spielberger State-Trait Anxiety Inventory (STAI), Worry Scale (WS), Fear Questionnaire (FQ), and Padua Inventory (PI). Two older adult community subsamples were assessed, one with Generalized Anxiety Disorder (GAD; n = 50) and the other with no anxiety complaints (Normal Controls: NC; n = 94). Descriptive data revealed that mean scores in the GAD sample were similar to those reported in studies of younger GAD patients. Mean scores in the NC sample, however, were lower than those reported in studies of younger control samples. Internal consistency for all measures generally was adequate in both the GAD and NC subsamples, although alpha coefficients for two of the FQ subscales were marginal. Test-retest reliability over a 2-4 week interval (assessed in a subgroup of NC subjects) was mixed, with some measures apparently assessing stable, trait-like dimensions of fear and anxiety, and others estimating more state-like clinical features. Intercorrelations among subscales for all four measures in both the GAD and NC groups generally were high and consistent with findings from prior research. Finally, some support for the convergent validity of the four anxiety measures was obtained, particularly in the NC sample. Results are discussed in terms of the utility of these instruments for future investigations of the psychopathology and treatment of anxiety disorders in the elderly.
尽管焦虑症在老年人中的患病率相对较高,但针对这一人群建立心理测量学上可靠的测量方法却很少受到关注。当前的研究通过检验四种自我报告焦虑测量方法的心理测量特性来解决这一问题:斯皮尔伯格状态-特质焦虑量表(STAI)、担忧量表(WS)、恐惧问卷(FQ)和帕多瓦量表(PI)。对两个老年社区子样本进行了评估,一个样本患有广泛性焦虑症(GAD;n = 50),另一个样本没有焦虑症状(正常对照组:NC;n = 94)。描述性数据显示,GAD样本中的平均得分与年轻GAD患者研究中报告的得分相似。然而,NC样本中的平均得分低于年轻对照组样本研究中报告的得分。在GAD和NC子样本中,所有测量方法的内部一致性总体上是足够的,尽管FQ两个分量表的α系数处于临界值。在2 - 4周的间隔内进行的重测信度(在NC受试者的一个亚组中进行评估)参差不齐,一些测量方法显然评估的是恐惧和焦虑的稳定的、特质性维度,而其他方法估计的是更具状态性的临床特征。GAD组和NC组中所有四种测量方法的分量表之间的相互关联通常都很高,并且与先前研究的结果一致。最后,获得了对这四种焦虑测量方法收敛效度的一些支持,特别是在NC样本中。根据这些工具在未来对老年人焦虑症心理病理学和治疗的研究中的效用对结果进行了讨论。