|
|
P-values standardised vs unstandardised? |
|
Message/Author |
|
|
Hello, I wonder how come the p-values for standardised and unstandardised coefficients are sometimes so large? This only seems to happen in SEM, in regression models standardised and unstandardised effects seem to get the same p-values. Or is it something about how Mplus calculated these coefficients? Best, Fredrik Falkenström |
|
|
Do you mean that there are large p-value differences between raw and standardized - or that the p-values are large? |
|
|
Sorry, I was unclear. The difference in p-values between raw and standardised coefficients is large. In this case, the p-value for the unstandardised coefficient is .28 and for the same path but standardised coefficient p = .09. I see this now and then, although most of the time standardised and unstandardised coefficients have fairly similar p-values. Incidentally, in DSEM, p-values for standardised and unstandardised coefficients seem to be identical? |
|
|
See the FAQ on our website: Standardized coefficient can have different significance than unstandardized |
|
|
Great, thanks! |
|
Back to top |
|
|