A method that seems designed to allow researchers to draw causal conclusions from single-source, concurrent observational data is a problem. The management literature is awash with this nonsense. Sure, the users are to blame but PROCESS seems almost designed to facilitate this nonsense.

So sorry for your loss Simine. 😢 We don't deserve dogs and yet they put up with us.

There are so many types of skullduggery - why limit yourself to research on just one?

I think that a prior editor often signed onto COPE but subsequent editors were not informed.

Many journals are members of COPE and their flowcharts and guidelines would supposedly apply but in my experience these are often ignored - sometimes because editors have no idea that their journal is a COPE signatory.

I just wish that grad programs would follow the evidence and not base admissions decisions on personal statements at all - because we have good meta-analytic data showing that they are not predictive of success and because we can't even know who wrote them.

I struggle to think of a statistical "method" that has been more damaging to psychology (and related disciplines) than mediation analysis based on observational data. IMHO all such manuscripts should be desk rejected.

And, of course, such a model would have even worse fit.

Perhaps I am not following correctly but the authors seem to show that a model with two higher-order factors does not fit data very well and then rely on a model with a single higher-order factor for most other analyses without ever showing (or even testing) that this model exhibits adequate fit.

The only systematic review that I know of has also demonstrated that personal statements offer no incremental validity in the prediction of success indicators in grad school.
proquest.com/docview/225605798?pq-origsite=gscholarcbl=1059mod=article_inline&fromopenview=true

Horrifying. Of course, many psychologists simple "forget" to even include a control group. In one notable instance that led to about 100 million viewers being misled about the efficacy of power posing.

Hard to tell without means, standard deviations, and correlations among variables. Pretty surprising that none of that is reported anywhere.