Recent content by spunky

Given the way theoretical reliability and validity are defined, it's mathematically impossible for a validity coefficient to be greater than the reliability. A formal proof of this statement can be found in Chapter 3, Section 3.9 of Lord & Novick's Statistical Theories of Mental Test Scores...

Mostly yes. The computation of the (sample) TLI could be negative and it essentially implies that the independence model (i.e., the model where all correlations are assumed to be 0) fits the data better than the model you propose. So, yes, it does imply that your model is really bad because it...

Yeah, this is definitely a bad model. What's considered an "acceptable" model means RMSEA < 0.06 and CFI >0.90 as per the Hu & Bentler (1999) criteria. There's also the issue of your sample size which Barrett (2006) commented on:

What are the values of the RMSEA and the CFI? And what estimation method did you use?

Whatsapp! The world's most popular messaging app! I don't get why it isn't so popular in North America!

If we know Corr(1/a, 1/(a+b)), Corr(1/a, (1/(1-(b/a))) and Corr(1/(a+b), (1/(1-(b/a))) we can work something out. Nevertheless, preliminary simulations show me that these various correlations are very, very, VERY close to 0 even if cor(a,b)=0.9999. If the OP can safely assume them to be 0, we...

My posts are infinite and eternal. They can't be deleted. They only transform.

And how exactly do you account for sampling error if you only offer descriptives?

After reading your post it made realize that the problem, as I posted it, is ill-conceived and allows for a variety of trivial solutions. For instance, take two standard normal random variables and add a positive constant (say 2 for argument's sake). They are both positively correlated and one...

Ok, I'll bite now because this thread is starting to get interesting. For a bivariate distribution (let's keep it simple), under which assumptions might one be able to claim that a positive correlation implies a larger mean in one marginal than the other? The bivariate normal is obviously out...

This is a good article with examples and solutions to use in R and SPSS: https://scholarworks.umass.edu/pare/vol24/iss1/1/

No. Without more information about the data, nothing can be inferred about the correlation and the averages.

Can you point on where in the STATA documentation they describe these models? I recognize a few but (with the exception of PCA) those acronyms are meaningless to anyone who doesn't use STATA. We need to know how STATA defines the models to tell the difference.

Mediation models like yours are very, VERY specific to subfields within the social sciences. It's usually hard to get help for those models unless you work with someone who specializes on them because almost nobody outside of those subfields uses them or is familiar with them. Have you tried to...

The most immediate reason of why this is a suboptimal approach is because your PCA scores are estimated from the data. These are datapoints that you didn't collect and observe so by treating them as just another collection of data points you underestimate the true uncertainty associated with...