Timeline for Partial Least Squares structural equation modeling
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 4, 2014 at 13:31 | comment | added | phx | Just calculate the the unstandardised solutions of your dependent variable using the weights of the responding measurement models and the manifest scores of the indicators as you have described in your question. From this unstandardised score you can derive the mean and the sd for your further calculations. I am not completely sure but that should do... | |
| Jul 4, 2014 at 13:24 | comment | added | schlusie | Yeah. I know. But I dont know the sd and mean of the unstandardized prediction. I know my StandLV and my unstandLV. But prediction gives me only the StandLV | |
| Jul 4, 2014 at 13:14 | comment | added | phx | Standardisation only means to substract the mean and divide by the standard deviation ((x-mean)/sd). Just transform your standardised LV score by using the inverse (i.e.: x = Stand.LVScore * SD + mean). | |
| Jul 4, 2014 at 13:08 | comment | added | schlusie |
Thanks. Im using smartPLS to build my model and than the sempls in R to estimate it. I know how to predict the standardized LV scores with my path coeffcients and the other variables. But I have no idea how to predict the unstandardized LV score.
|
|
| Jul 4, 2014 at 12:50 | history | answered | phx | CC BY-SA 3.0 |