 STDYX vs STDY - binary outcome    Message/Author  Manon Burgat posted on Wednesday, June 26, 2013 - 5:25 pm
Hello,

I am conducting a path analysis with 1 binary outcome variable (DV), a continuous mediating variable, and 5 continuous IVs. I am using WLSMV as the estimator for this model. My question relates to which standardized coefficients I need to report.

My understanding is that I should report STDYX estimates for pathways between the continuous IVs and the continuous mediating variable. Is that correct?

I'm unsure which estimate to report for the pathway between the continuous mediating variable and the binary outcome variable. I have read in previous posts that I should report the SDTY estimate. Is this correct? If so, I am not able to obtain this from my output, even when I specify STANDARDIZED (STDY).

I have found a formula for STDY in the userguide (bStdY = b/SD(y)) but I am unclear on how to calculate this. Does "SD(y)" stand for the standard deviation of my mediating variable? Can you tell me what 'b' stands for? Where do I obtain these values?

Apologies for the long question.  Linda K. Muthen posted on Thursday, June 27, 2013 - 11:24 am
In a conditional model, use StdYX for continuous covariates and StdY for binary covariates.

b is a regression coefficient. SD(y) is the standard deviation of y.  Manon Burgat posted on Thursday, June 27, 2013 - 5:50 pm
Thanks for this. Okay, so am I correct in thinking that 'b' is the unstandardised estimate provided in the output prior to the standardized model results? And is "SD(y)" the standard deviation of my binary DV?

Kind regards.  Manon Burgat posted on Thursday, June 27, 2013 - 7:07 pm
Hi again,

I also need to clarify that I have a binary dependent (outcome) variable, rather than a binary covariate. For example - "v2 ON v1" - where v2 is binary categorical.

My question is - Can you tell me which standardised estimate I should report for this path - Std StdYX or StdY?

The userguide states the following on pg.642: "StdY should be used for binary covariates because a standard deviation change of a binary variable is not meaningful. The standardized coefficient bStdY is interpreted as the change in y in y standard deviation units when x changes from zero to one."

If Y = DV and X = IV, it does not make sense to me to calculate StdY, as this formula includes the standard deviation of the DV, which is a binary variable.  Linda K. Muthen posted on Friday, June 28, 2013 - 5:52 am
Continuous covariate -- StdYX
Binary covariate -- StdY    Topics | Tree View | Search | Help/Instructions | Program Credits Administration