Lightnews — Scholar-powered news

Sholto David @sholtodavid.bsky.social · 5d

Alison found another... I think it might be worse.

Sholto David @sholtodavid.bsky.social · 5d

Adding to the previous diagram by Anna, I can see tiny little football looking things (red arrows) and some of the surrounding area.

1 3

Sholto David @sholtodavid.bsky.social · 5d

Here's the paper I posted about: doi.org/10.3390/jnt5...

doi.org

1 2

Sholto David @sholtodavid.bsky.social · 5d

Great find! The one you found is even better 😂

1

Sholto David @sholtodavid.bsky.social · 5d

Recent MDPI paper including diagram of a bacterial cell with a mitochondrion 🤦‍♀️ The authors propose a mechanism for the toxicity of silver nanoparticles: "Ag+ ions in AgNPs bombard the bacterial mitochondria’s electron transport chain, resulting in cell death".

8 11 52

Sholto David @sholtodavid.bsky.social · 13d

As Fabian says - some people are compulsive fiddlers, manipulators, and (frankly) liars. They would change the data simply for the sake of changing it. Because it is their habit. I have personally met people who lie even when it damages their own credibility. I suspect the same is true of data.

3

Sholto David @sholtodavid.bsky.social · 13d

One thing to consider is that the only good reason to manipulate an image is to make it show something that it didn't show in the first place. Perhaps what was there originally was not background? A defect? A particle? Someone's watermark or label from another paper.

1 1 2

Sholto David @sholtodavid.bsky.social · 16d

Thanks, Dorothy, glad you found it interesting! Funny thing is I dislike cycling in Oxford very strongly, even though they call it a "cycling city". Catch up soon 😊

1 2

Sholto David @sholtodavid.bsky.social · 17d

"Criminal" seems to be any behaviour which Wafik doesn't like 🤷‍♀️

3

Sholto David @sholtodavid.bsky.social · 19d

A CRO has been scamming academic and industry labs in the US with fraudulent data that has been published in Nature and other high impact journals. Leonid has written a great piece on my findings: forbetterscience.com/2025/09/22/i...

In Bad Company

“Authors thank the members of NeuroDigitech for their contribution to data generation, and all the animals that contributed to these studies.”

forbetterscience.com

16 23

Sholto David @sholtodavid.bsky.social · 19d

In 2024 I cycled from the UK to China. I finally finished a vlog, so here it is if you'd like to watch it. It's a glorified holiday album with some complaints and maps 😂 youtu.be/MdgHZPfivVA

Cycling Alone from the UK to China: 10,000 km in 90 days

YouTube video by Sholto David

youtu.be

2 6 28

Sholto David @sholtodavid.bsky.social · 25d

I have been blogging at For Better Science. Revisiting Queen's University in Belfast: forbetterscience.com/2025/09/16/f...

1 6

Sholto David @sholtodavid.bsky.social · 28d

Great spot by Yong‐Chang Zhou who was first on the scene! I wonder how they spotted it? This is a relatively new account on PubPeer that is quite active. pubpeer.com/publications...

PubPeer - WNT signalling control by KDM5C during development affects c...

There are comments on PubPeer for publication: WNT signalling control by KDM5C during development affects cognition (2024)

pubpeer.com

1 1 5

Sholto David @sholtodavid.bsky.social · 28d

"Outsourcing" strikes again!

1 2

Sholto David @sholtodavid.bsky.social · 28d

Four different papers, including one in Nature. The images appear to have been provided by a CDMO (Neurodigitech, LLC) to different teams in the US. Researchers have been adding these beautified and recycled images into their papers like they're filling up a sticker book.

1 4 6

Sholto David @sholtodavid.bsky.social · 28d

Also see here from the same team:
pubpeer.com/publications...
pubpeer.com/publications...

2 2

Sholto David @sholtodavid.bsky.social · 28d

Even in 2024 and 2025 after so much scandal there are still name brand labs publishing photoshopped neuroscience images...

1 2 8

Sholto David @sholtodavid.bsky.social · Sep 11

Yes - I was only being silly. Dove Press have been very quick too.

1 1

Reposted by Sholto David

Derek Lowe @dereklowe.bsky.social · Sep 10

This is bad news.

Shooting people whose views we disagree with can only take us in a terrible direction.

4 2 83

Sholto David @sholtodavid.bsky.social · Sep 10

I don't think this is a problem. Is there any recorded instance of a journal acting faster than several months?

1 4

Reposted by Sholto David

Fabian Wittmers - PubPeer: Archasia Belfragei @academic-integrity.bsky.social · Sep 9

Yes, I really marked every duplication of a particle in this SEM image. It might have taken me longer than it took them to fake it... Khan et al. 2017 (DOI: 10.1016/j.jallcom.2016.11.260)
This Elsevier article is a complete clown show... pubpeer.com/publications... #ImageForensics

3 4 15

Sholto David @sholtodavid.bsky.social · Sep 8

Harvard researchers publishing in Nature, some encouragement for those suffering imposter syndrome. pubpeer.com/publications...

1 13

Reposted by Sholto David

David Sanders @davidsanderssci.bsky.social · Sep 5

A Cuban article on Carlo Croce @OhioState . It doesn't get all the facts correct, but it does include an image of @smutclyde.bsky.social @thatsregrettab1.bsky.social @forbetterscience.bsky.social @elisabethbik.bsky.social www.juventudtecnica.cu/articulos/fr...

Fraudes científicos de este siglo (35): Carlo M. Croce: el timador timado - Juventud Técnica

En 2024 se habían retractado 15 de sus publicaciones sobre oncología, cinco recibieron una ‘expresión de preocupación’ y 23 fueron corregidas.

www.juventudtecnica.cu

1 5 10

Reposted by Sholto David

For Better Science @forbetterscience.bsky.social · Sep 5

They use the artwork of @smutclyde.bsky.social I commissioned from @sholtodavid.bsky.social !
forbetterscience.com/2023/10/09/c...

1 2 2

Reposted by Sholto David

Smut Clyde, X-Ray Haruspex @smutclyde.bsky.social · Aug 31

This is what happens when you entrust a chatbot with the task of writing your paper about the Chimp Optimization Algorithm and the Evolutionary Programming Algorithm.

doi.org/10.1109/acce...

2 4 13