Manual Calculation of SEM Factor Scores

Question

I am trying to re-create the factor scores computed by AMOS for a latent construct. However, I get a very different answer when using the "Estimates" for the variables than when AMOS computes the factor scores (using the Impute method). So for example, AMOS reports the weights of the variables as:

V1 ~ Latent_var1 = 1
V2 ~ Latent_var1 = .75
V3 ~ Latent_var1 = .67
V4 ~ Latent_var1 = .45

If I simply multiple the actual scores from any respondent for V1, V2, and V3 by these coefficients, it will not equal the "Factor Score" reported by AMOS (not even close), for that same respondent. If this were standard linear regression, it would work just fine: constant+beta*_value + beta2*_value + beta3*_value, etc.

Does anyone know why?

For those familiar with AMOS, here's snippet from the model output. the Latent construct Personality estimates vp1 - vp6. If I take the actual variable values and multiple them by these coefficients, I do not get the same factor score.