This is a new method so please try it out and let us know if you run into any difficulties! Text at arxiv.org/pdf/2511.01784 , SPM toolbox and example at github.com/pzeidman/vRS... . Congrats to Alex Lepauvre for this excellent work! [4/4]

vRSA for M/EEG has advantages: 1) Bayesian model comparison for testing hypotheses, e.g. presence vs absence of effects at specific peri-stimulus times. 2) Handles full factorial designs with interactions. 3) Computationally efficient, no cross-validation. 4) Group analysis with covariates. [3/4]

RSA has proven popular in neuroimaging, for testing whether the (dis-)similarity of stimuli or conditions are encoded in the brain. But it has drawbacks. In particular, it's typically not backed by a formal statistical model that separates signal from noise, or handles factorial designs. [2/4]

Looks brilliant, congratulations!!!

"The journal offers a magnificent fireworks display of reflections on the past, present and future of SPM. Yet, the debate extends far beyond SPM. It touches on crucial issues such as how to interpret the growing body of neuroimaging data and explain it in a biologically plausible way" - Ed. [2/2]

* Keen observers of neuroimaging history will note that we've marked 30 years rather arbitrarily, from around the time the code was re-written in 1994, which introduced SPM as we know it today. But as Prof Bullmore points out in his commentary, SPM is a few years older than that!

Congratulations Mona, very well deserved!

Many congrats Nadine and Steve! This looks very elegant indeed!

Thanks so much Rimona! Really nice you were there!

Brilliant! haha not at all, it's a badge of honour! 🤓🔬

Sounds like we have a winner!