Dr Mircea Zloteanu 🍁 🍃
@mzloteanu.bsky.social
Lecturer Psych & Crim @kingscollegelondon.bsky.social | Deception Detection; Emotions; JDM | Open Science; R; Bayes | @ukrepro ReproTea & StatsTea | #statstab | 🇷🇴 🇬🇧🌍
Stats blog: https://mzloteanu.substack.com/
Stats blog: https://mzloteanu.substack.com/
November 27, 2025 at 8:25 AM
All well and good until someone (me) brings up the "decision p-value" vs the "divergence p-value" distinction. doi.org/10.1111/sjos...
Divergence versus decision P‐values: A distinction worth making in theory and keeping in practice: Or, how divergence P‐values measure evidence even when decision P‐values do not
There are two distinct definitions of “P-value” for evaluating a proposed hypothesis or model for the process generating an observed dataset. The original definition starts with a measure of the dive...
doi.org
November 26, 2025 at 8:46 AM
All well and good until someone (me) brings up the "decision p-value" vs the "divergence p-value" distinction. doi.org/10.1111/sjos...
"what do you mean there is error in the model!? Take it out before the shareholders arrive!"
November 22, 2025 at 10:06 PM
"what do you mean there is error in the model!? Take it out before the shareholders arrive!"
* looks at bank balance *
Inner voice: you should buy that Steam Deck. Treat yourself. You don't need food this month.
Inner voice: you should buy that Steam Deck. Treat yourself. You don't need food this month.
November 22, 2025 at 1:40 AM
* looks at bank balance *
Inner voice: you should buy that Steam Deck. Treat yourself. You don't need food this month.
Inner voice: you should buy that Steam Deck. Treat yourself. You don't need food this month.
Thanks. I know that one, but I want to have code that is more explicit on what it does. Still, I might use this for now.
November 20, 2025 at 9:56 PM
Thanks. I know that one, but I want to have code that is more explicit on what it does. Still, I might use this for now.
The principle is "pick the thing that makes my crappy study look amazing"
November 20, 2025 at 7:37 PM
The principle is "pick the thing that makes my crappy study look amazing"
Very true. Most researchers would not like the sample size they need to make the inferences they really want. I wrote a little piece about this issue. If you want to do range tests and set small SESOI or MCID, you need a huge N mzloteanu.substack.com/p/your-study...
Your Study Is Too Small (If You Care About Practically Significant Effects)
Power ≠ Precision, the misguided comfort of 80% power
mzloteanu.substack.com
November 19, 2025 at 11:26 AM
Very true. Most researchers would not like the sample size they need to make the inferences they really want. I wrote a little piece about this issue. If you want to do range tests and set small SESOI or MCID, you need a huge N mzloteanu.substack.com/p/your-study...
I'm realising now I'll need a co-author for this paper on negative binomial reg. This would have been a bit blunder...
November 18, 2025 at 8:02 PM
I'm realising now I'll need a co-author for this paper on negative binomial reg. This would have been a bit blunder...
This is a useful thread for me to stumble upon tonight... I was just writing up this stuff 😅
November 18, 2025 at 7:41 PM
This is a useful thread for me to stumble upon tonight... I was just writing up this stuff 😅