The Scaling of Latent and Observed Variables in State-Trait Models

The Scaling of Latent and Observed Variables in State-Trait Models
复制标题

状态特质模型中潜在变量和观察变量的缩放

DOI:
--
复制
发表时间:
2003
期刊:
影响因子:
--
通讯作者:
Daniel J. Bauer
Daniel J. Bauer
中科院分区:
--
文献类型:
--
作者:
Daniel J. Bauer

文献摘要

参考文献

被引文献

相似文献

Eid和Langeheine(本期)是值得赞扬的,因为他们提供了一个容易理解的解释,说明当预期人口亚组表现出不同的稳定性和变化模式时,潜在的类分析如何应用于评估状态-性状模型。这种发展是受欢迎的,因为状态-特质模型代表了在个体内部变化中概念化个体差异的一种重要方式(参见Meredith & Tisak,1990; Nesselroade,2001)。而不是批评状态-特质模型本身,我直接我的评论主要是在缩放问题,即,缩放的潜变量和缩放的观察变量,以及缩放问题如何可能影响程序的有效性。我对缩放的担忧是双重的。首先,我质疑特征和状态是否最好用分类潜变量(类型学)而不是连续潜变量(维度)来表示。其次,我担心的是,估计的比例的情况下,每个潜在的亚组(或链)可能会受到影响,利用粗略分类的观察变量作为指标。也就是说,我想知道模型的估计值在多大程度上反映了观察到的测量值的缩放,而不是个体差异的实际结构。我首先解决潜变量的缩放,然后继续考虑观察变量的缩放。测量:跨学科研究和展望,1(3),207-231版权所有© 2003年,劳伦斯Erlbaum联营公司。
Eid and Langeheine (this issue) are to be commended for providing an easily understandabledepictionofhowlatentclassanalysismaybeapplied toevaluatestate–trait models when population subgroups are expected to show different patterns of stability and change. Such developments are welcome, as state–trait models represent one important way of conceptualizing individual differences in intraindividual change (see also Meredith & Tisak, 1990; Nesselroade, 2001). Rather than critique the state–trait model itself, I direct my commentary primarily at scaling issues, namely, the scaling of the latent variables and the scaling of the observed variables, and how scaling issues may impact on the validity of the procedure. My concerns about scaling are two-fold. First, I question whether traits and states are best represented as categorical latent variables (typologies) rather than continuous latent variables (dimensions). Second, I am concerned that estimates of the proportion of cases within each latent subgroup (or chain) may be affected by utilizing crudely categorized observed variables as indicators. That is, I wonder to what extent the estimates of the model reflect the scaling of the observed measures rather than the actual structure of individual differences. I first address the scaling of the latent variables, and then proceed to consider the scaling of the observed variables. MEASUREMENT: INTERDISCIPLINARY RESEARCH AND PERSPECTIVES, 1(3), 207–231 Copyright © 2003, Lawrence Erlbaum Associates, Inc.
DOI: 10.1037//0022-3514.67.4.674
发表时间: 1994
影响因子: 7.6
作者:
Shoda,Y;Mischel,W;Wright,JC
通讯作者: Wright,JC
童年晚期和青春期早期的基因型-环境相关性:反社会行为问题和强制养育。
DOI: 10.1037//0012-1649.34.5.970
发表时间: 1998
期刊: Developmental psychology.
影响因子: --
作者:
O'Connor,TG;Deater-Deckard,K;Fulker,D;Rutter,M;Plomin,R
通讯作者: Plomin,R
DOI: 10.1001/archpsyc.1993.01820210024003
发表时间: 1993
影响因子: --
作者:
Kendler,KS;Heath,AC;Neale,MC;Kessler,RC;Eaves,LJ
通讯作者: Eaves,LJ