Calculation about unstandardized LV score, Dr. Bido's method

[allenjia]

I am confused with this method. For example, we have 200 lines of original data; and run 500 sample bootstrapping. Thus we have 500 SOW/Std, and thus 500 lines of relative weight for each sample. When we get the relative weight, we can get lv score by using "w1*original value of indicator 1+……". However, we just have 200 lines of original value. Here is the problem, 500 lines weight but only 200 lines original value. There will be 300 lines only have relative weight but no original value. I do not know whether I interpret wrongly. Please help me. Thanks


viewtopic.php?t=812&highlight=unstandardized++score

[Diogenes]

Hi,

1) To compute the unstandardized LV scores, we do not use “bootstrap algorithm”, just the results of the “PLS algorithm”.

2) The std.dev. are computed from the indicators (MV – original), in your case 200 rows.

Best regards,

Bido

[allenjia]

Dear Dr. Bido,

Thank you sooooooooooooo much! I think I got it.

Lin

1) Run the example---PLS algorithm
2) Copy the standardized outer weights--Weights of items on relevant factors
3) Compute the standardized deviation to each indicator and copy/past to the side of the weight----use Manifest variable score-original in Model specification to compute std
4) Compute (outer weights / standard deviation for each indicator)
5) Compute the unstandardized weights as “relative weights” to have the unstandardized LV scores in the same scale of the indicators.