I guess that’s the kind of governing body that thinks this is a good idea.

Read about it here:
pubpeer.com/publications...
pubpeer.com/publications...
and here: www.linkedin.com/posts/peter-...
#EconSky #ResearchIntegrity #Economics #Finance @pubpeer.com

These are just two of many articles in #Economics journals that I have found similar issues with -- some are in journals with Impact Factors >10.

I plan to comment on additional papers, but there more than I will have time to address. #EconSky #OpenScience @pubpeer.com @datacolada.bsky.social

Instead of replying on PubPeer, the lead author of the other paper emailed me directly, which didn’t address why the figures are virtually indistinguishable.
I encourage them to post a short reproducibility package and link it on PubPeer, so readers can verify.
pubpeer.com/publications...

Coşkun et al. (2023) looks legit: based on a PhD dissertation; data/code on GitHub; public post on LinkedIn.

Yet this specific figure doesn’t match their data on Github. Also, signs suggest the upper panel is returns (±), and lower panel is volume (>0).🤔
pubpeer.com/publications...

Full paper here: arxiv.org/abs/2410.23587
Moments by Integrating the Moment Generating Function co-authored with Chen Tong (Xiamen).

I believe this result is useful in many fields. Anyone know the proper outlet for this type of result? #arXiv #OpenScience
arxiv.org/abs/2410.23587

The CMGF method provides a simple integral formula that yields fractional, absolute, central, and even complex moments — without relying on the density function or any derivatives.
For example,
E|X|^0.7 or tail moments
E[(X - ξ)^3 1{X > ξ}].
It is fast: NIG example:

he new method avoids derivatives entirely, is fast, general, and easy to implement.
We call it CMGF, because it uses a complex argument for the MGF (also nests the characteristic function).
And it applies to complex moments too. #Physics

Unfortunately, for fractional moments — or in dynamic models where MGFs are defined recursively — computing derivatives can be difficult, slow, or even infeasible. Existing methods often involve fractional derivatives or integrals with messy expressions.

It's well known that the k-th integer moment is given by the k-th derivative of the MGF evaluated at zero. This is simple, clean, and taught in many probability courses.
But it has limitations. It only works if you want an integer moment, and can take the derivatives you need.