Using AI for self-reflection feels like therapy, but uses completely different mechanisms, leaving the 'patient' in the driver's seat - as for the quality of the responses, it definitely requires your own critical thinking. That’s what makes it so interesting.

#CogPsyc #PsychSciSky

Excuse me, BREAKING NEWS:
are we being controlled by cat-borne brain parasites 🧪 #BehEcon, #CogPsyc data science dataskyence epidemiology, VetMed #catlife #you’vegottobekittenme

BeyondTheLab #cogpsyc #AcademicSky #Psychology #Perception 🧠🧪doi.org/10.1037/xhp0001458

You can use AI to get starting information when exploring the outside world.
But if you're using AI for self-reflection, it's better to do your own homework and research the topic before discussing it with AI. Usually, people do the exact opposite
#CogPsyc #PsychSciSky

#cogpsyc #AcademicSky🧪 🧠 A good figure makes the structure of an effect visible at a glance. Great figures, like great scientific writing, are precise and self-explanatory. A high information/ink ratio in both domains makes your argument easier to see, and harder to dismiss.

#cogpsyc #AcademicSky🧪 🧠 If you justify your sample size as "comparable to prior work", we will ask you to remove that claim. Precedent is not a justification — it moves the question one paper back. Prior work can still inform your expected effect size, with due caution about published estimates.

I’ve found a useful way to use AI before a difficult conversation. The important part is the middle. #cogpsyc

#cogpsyc #AcademicSky🧪 🧠 Small-N designs can be informative, with a lot of data per S. Estimate the effect within S, then ask if it replicates across S. A pooled analysis across all trials can produce impressive, but misleading, precision.

See Smith & Little (2018), doi.org/10.3758/s134...