|Original language||English (US)|
|Title of host publication||The Encyclopedia of Clinical Psychology|
|Editors||Robin L Cautin, Scott O Lilienfeld|
|State||Published - 2015|
The ideal of measurement for a multiple-indicator scale is that it should be unidimensional. However, there are many latent variables or constructs of interest within the field of clinical psychology widely believed to be inherently multifaceted. Although scales that adequately assess such constructs are unlikely to be strictly unidimensional, they should still primarily measure one latent variable. Though many believe that coefficient alpha is a good index of the extent to which a scale is primarily a measure of one underlying latent variable, it is not. This entry describes an alternative to alpha, which is derived from the hierarchical factor analysis model—coefficient omegahierarchical. This entry includes discussion of the extant research comparing different methods of estimating coefficient omegahiearchical to each other and to coefficient alpha, as well as some important future directions for research on coefficient omegahieararchical.