Parameter estimates, CIs and p-values inconsistent across unstandardized and standardized solutions

[ENC]

Hi there,

I am running a cross-lagged model to study the relationships between anxiety and social communication over two time points. I have noticed that when I get parameter estimates and confidence intervals for the unstandardized solution, these are very different to the parameter estimates and confidence intervals for the standardized solution.

I am using the following line of code to get confidence intervals for the unstandardized parameter estimates:

·        summary(fit, ci = TRUE)

And this line of code to get confidence intervals for the standardized parameter estimates:

·        standardizedSolution(fit, ci = TRUE, level = 0.95)

When I do this, one of my model's pathways is significant when I look at the unstandardized parameter estimate (b = .019, 95% CIs [.004, .035], p = .015), but non-significant when I look at the standardized one (β = .185, 95% CIs [-.791, 1.160], p = .710).

Any information on why this may be the case, or advice on how to interpret this discrepancy, would be much appreciated.

Thank you,

Elise