Timeline for Conversion from factor loadings to alpha scale reliability
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 26, 2023 at 19:43 | comment | added | Ben Ogorek | I'm also looking at Cho's paper Cronbach’s Coefficient Alpha: Well Known but Poorly Understood, and on p 33 he's got the unidimension version of omega with that includes the latent factor variance. If it's set to 1 during estimation, then the formula reduces to yours (with the parentheses added). I'm thinking at if you set the first loadign to 1 and let the factor variance float, you'd have to use (sum of loadings) ^2 * latent factor variance in the numerator and denominator | |
| May 26, 2023 at 19:24 | comment | added | Ben Ogorek | This is a helpful answer, but noting that you've got an equals where I think you meant to put a plus in 4.41/(4.41=1.53)=.74. Also I think you need an extra set of parentheses in the denomonator in your first formula | |
| Apr 29, 2021 at 15:37 | history | answered | jsakaluk | CC BY-SA 4.0 |