Energy, fatigue, or both? A bifactor modeling approach to the conceptualization and measurement of vitality
Energy, fatigue, or both? A bifactor modeling approach to the conceptualization and measurement of vitality
复制标题
DOI：
10.1007/s11136-014-0839-9
复制
发表时间：
2015-01-01
影响因子：
3.5
通讯作者：
Ware, John E., Jr.
中科院分区：
文献类型：
--
作者：
Deng, Nina;Guyer, Rick;Ware, John E., Jr.
活力是反映与健康相关的生活质量的身体和情感成分的一个重要领域。由于它的复杂性，它的定义和衡量既有广义的也有狭义的。我们探索了一个非常全面的物项库的维度假设来衡量活力及其相关概念。第二次分析使用了1343名代表美国普通人群的成年人对基于互联网的调查的回应，包括从多个量表（例如，SF-36 Vitality, promise - fatigue）编制的42个项目，涵盖了广泛的与活力相关的内容领域（能量、疲劳及其对身体、心理、社会活动和生活质量的干扰）。探索性和验证性因素模型分别使用对半样本进行评估。与传统的单维模型、多维模型和分层模型相比，采用双因子模型来评估项目的基本单维性。方法通过相关误差对常用量表或短语的影响进行建模。探索性因子分析确定了一个主导因素。验证性因子分析确定了一个最佳拟合（CFI = 0.964, RMSEA = 0.084）双因素模型，其中一个一般（活力）因素和两个组（能量和疲劳）因素，解释了总方差的69.3%和4%。考虑到方法影响的相关误差对确定项目的实质性维度很重要。双因子模型被证明对评估复杂结构的维度是有用的。结果支持将活力作为一个单维能量疲劳结构进行概念化和测量。我们鼓励今后的研究比较基于更广泛和更狭窄的活力概念的措施的实际影响。
Vitality is an important domain reflecting both the physical and emotional components of health-related quality of life. Because of its complexity, it has been defined and measured both broadly and narrowly. We explored the dimensionality of a very comprehensive item bank hypothesized to measure vitality and its related concepts.Secondary analyses were conducted using the responses of 1,343 adults representative of the US general population to Internet-based surveys including 42 items compiled from multiple scales (e.g., SF-36 Vitality, PROMIS-Fatigue), covering a broad range of vitality-related content areas (energy, fatigue, and their interference with physical, mental, social activities, and quality of life). Exploratory and confirmatory factor models were evaluated independently using split-half samples. Bifactor model was used to assess the essential unidimensionality of the items, in comparison with traditional unidimensional, multidimensional, and hierarchical models. Method effects of a common scale or phrase were modeled via correlating errors.The exploratory factor analysis identified one dominant factor. The confirmatory factor analysis identified a best-fitting (CFI = 0.964, RMSEA = 0.084) bifactor model with one general (vitality) and two group (energy and fatigue) factors, explaining 69, 3, and 4 % of total variance. Correlating errors accounting for the method effects were important in identifying the substantive dimensionality of the items.The bifactor model proved to be useful for evaluating the dimensionality of a complex construct. Results supported conceptualizing and measuring vitality as a unidimensional energy-fatigue construct. We encourage future studies comparing practical implications of measures based on the broader and narrower conceptualizations of vitality.