me running descriptives: oh, cool i have an interesting finding, let me tell someone. me again running sensitivity checks before emailing them...nvm
Maria Diaz Mize
@msmariadiaz.bsky.social
🇩🇴 Non-traditional Doctoral student in criminology at Florida State University. Interests: community corrections, courts, jails, crime and the media, pretrial justice, AI and surveillance. Mixed methods kinda gal.
A psychology journal has retracted a paper on how using generative AI can affect confidence in work tasks after its sole author refused to share data with the editors.
CEO of AI company loses paper on cognitive impact of generative AI
A psychology journal has retracted a paper on how using generative AI can affect confidence in work tasks after its sole author refused to share data with the editors. The April paper, publis…
retractionwatch.com
Love to use the web these days, where every single page on earth wants your email address
producing a CSV file of random likert scores is not "violating confidentiality" let's be so for real right now
Excited to share new study in Criminology (open access) on GLP-1s & violence: onlinelibrary.wiley.com/doi/full/10.... Among current GLP-1 users, the link b/t impulsivity and violence was flat across impulsivity levels. Similar story w/alcohol use. More to come from me & @daniel-semenza.bsky.social
So the one thing we learned is that there WAS ethics board approval, but still so many questions... retractionwatch.com/2026/06/16/t...
Journal investigating paper on cognitive impact of generative AI
A paper about the effects of generative AI use on confidence in work tasks is under investigation after critics raised questions about the study design, data analysis and ethics approval for the re…
retractionwatch.com
NGL I love a Waymo. Felt much safer in them as a woman.
Vibe coded for the first time yesterday with Claude Code. It was cool, but a lot of back and forth as I set up Claude to ask me for permission at every change (which I probably didn't need, this was a simple Python script). Anyway, feel free to call me a woman in tech now.
One of the ways I'm prepping for my methods comps in the fall is going through methods sections on papers. This one was a good exercise in thinking about mediation. It also reminded me what a Sobel test is (which I learned fall semester and promptly forgot).
Ian Hussey@ianhussey.mmmdata.io · 4mo agoMediation analysis is surprisingly easy to verify using reported betas, p values, and Ns. Eg the indirect a*b path should be, well, a multiplied by b. Sometimes its not, like in this article claiming SSRI efficacy is mediated by expectations. Whoops. pubpeer.com/publications...
Notes from the transcription trenches: Whisper transcription tool is utter garbage, I spend so much time fixing the errors. Still better than transcribing entirely by hand but I'm actually shocked that it's this bad
Meanwhile on threads everyones favorite AI researcher is talking about not violating confidentiality. Not sure how releasing full regression output or a CSV of random likert scores violates confidentiality but okay go off I guess
very cool to see online sentiment in all kinds of online corners
Open Measures@openmeasures.io · 4mo agoWe launched a new tool in our platform called “Research Assistant.” It’s an LLM agent that summarizes large datasets. AI can’t replace skilled researchers, and this isn’t a substitute. Instead, we geared this tool so it can reliably do things that save their time. openmeasures.io/ai-research-...
People really are just posting the AI output huh. We are done as a society, wrap it up
i'm 95% done with my PhD. I still have to pass comps and write/defend my dissertation but I'm just about there.
When you know even a little about methods, it's hard to get excited about clinical trial press releases, esp for GLP1s. Show me the receipts!!! We need critical reviews of these trials that are accessible to the layperson and that aren't driven by activism.
Allowing students to only use LLMs for finding sources, which it is known to be incredibly bad at, is an interesting policy choice
Nathan Kalman-Lamb@nkalamb.bsky.social · 4mo agoUC Berkeley Law creates new AI policy that bans AI for "conceptualizing, outlining, drafting, revising, translating, or editing any work submitted for credit" or exams because "thinking remains the sine qua non of good lawyering (and of a quality legal education)." That's what I'm talking about.
This review finds the evidence on the effectiveness of offender-based interventions (eg using cognitive behavioural therapy) to prevent repeat domestic abuse is weak. Lots of places are spending money on this, so we really need better evidence. doi.org/10.1007/s108...
The Effectiveness of Intervention with Intimate Partner Violence Perpetrators: An Umbrella Review - Journal of Family Violence
Purpose To advance knowledge on effective interventions for intimate partner violence (IPV) perpetrators, it is essential to synthesize evidence from meta-analyses evaluating the effectiveness of…
doi.org
If you want to join a criminology program, know you're not going to spend your time profiling killers, you're going to spend it reading dense sociology and learning statistics.
Brown suggests that maybe it's best if analysis is left to statisticians and not subject matter experts. Interesting idea, but then it totally lacks context/theory.
Retraction Watch@retractionwatch.com · 4mo agoHow the media hypes “research that is absurd on its face”
"Using an AI tool can generate qualitative data, but it is not the same as conducting qualitative inquiry"
AI interviewers can’t connect with people the way human researchers can – they can produce only data, not meaning
Anthropic developed an AI interviewer it claims can help social scientists conduct research at scale. But AI models lack the human qualities that make qualitative research unique.
theconversation.com
I don't want to be dropped into an "interactive AI experience", I want you to give me a list of links like I asked
Do not disrespect qualitative methods in my presence, please. You can get so much more from interviews. #PhDSky
Three specialists on meta analysis on your PubPeer thread and you're going on about haters on your socials? Just issue the retraction and let's wrap this up bc it's getting embarrassing #AcademicSky #ScienceSky
Where were you the first time you learned what a comprehensive exam was? #PhDSky #AcademicSky