iiinova — where your digital self keeps evolving

Reposted by Heather Urry

Christoph Strauch @cstrauch.bsky.social · 22h

How much of a risk are LLM-generated bots for online behavioral experiments? We present results of an entirely natural language prompt-generated agent, capable of absolving any of the 26 experiments we tested it on. Most tasks are passed with data that can pass as human osf.io/preprints/ps... (1/)

26132

Reposted by Heather Urry

Katie Hoemann @katiehoemann.bsky.social · 01/10/2026

Come join me in the Emotions in Context Lab at the University of Kansas! I will be taking PhD students in Fall 2027.

11510

Heather Urry @heatherurry.io · 01/10/2026

Learn about our department and Psychology PhD program, get to know our faculty, and soak up tips for applying at Tufts and elsewhere through a series of live conversations with members of the Psychology Department and a rich set of asynchronous resources you can review at your leisure.

000

Heather Urry @heatherurry.io · 01/10/2026

This event is open to all, and was designed specifically for applicants from groups underrepresented in psychology. Graduate Admissions will waive the application fee for open house attendees who apply to Tufts Psychology's Ph.D. program for the Fall 2026 application cycle.

100

Heather Urry @heatherurry.io · 01/10/2026

To Psychology PhD applicants: We're hosting "Diversifying Psychology: An Open House @ Tufts" starting 15 October 2026. [Register here.](tufts.qualtrics.com/jfe/form/SV_...)

Join us for Diversifying Psychology: An Open House @ Tufts
Join us for *Diversifying Psychology: An Open House @ Tufts* for prospective PhD program applicants starting October 15, 2026! Learn about our Psychology PhD program, get to know our faculty, and soak up tips for applying at Tufts and elsewhere. Featuring live conversations with department members and asynchronous resources you can review at your leisure! The Open House starts October 15, 2026. Please register to attend: https://tufts.qualtrics.com/jfe/form/SV_8dAO7KkUHinz7HE

123

Reposted by Heather Urry

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

📜 Out in Psychotherapy & Psychosomatics Most bereaved people adapt over time. Why do some stay stuck in grief? We tested whether how loss is remembered shapes grief symptoms over the first 18 months of bereavement w/Franziska Lechner-Meichsner, Anke Ehlers & Kirsten Smith doi.org/10.1159/0005...

1168

Reposted by Heather Urry

Society for the Improvement of Psychological Science (SIPS) @improvingpsych.org · 30/09/2026

Metadata matters! If you submit a preprint to PsyArXiv, to ensure smooth acceptance, don't forget to fill out the metadata correctly. Below you will find a few tips, but detailed guidelines are at 👉 buff.ly/Tcd8vIc #PsychSciSky

041

Heather Urry @heatherurry.io · 28/09/2026

Certainly by now, I hope!

110

Heather Urry @heatherurry.io · 28/09/2026

I love that tool! But, with my apparent blue bias, it doesn't quite explain why I knew my stepdad's green sweater was green, though he swore it was blue. It was green, damnit!

010

Heather Urry @heatherurry.io · 28/09/2026

Love this - help a professor out? (Not me.)

Result of my https://ismy.blue/ test-- it shows a color gradient from green to blue and notes the boundary for me to switch from saying a color is green versus blue. (My boundary is at hue 175, bluer than 65% of the population.)

220

Reposted by Heather Urry

Nicola Rennie @nrennie.bsky.social · 27/09/2026

You can use rel() to specify the size of individual elements relative to the base size. Means if you increase the base size, then you don't need to readjust sizes of each individual element again

162

Reposted by Heather Urry

In Otter News @in-otter-news.bsky.social · 26/09/2026

After a long day at the aquarium, the otters help tidy their toys away. Payment is made promptly and entirely in ice cubes.

331188893966

Reposted by Heather Urry

Lisa DeBruine @debruine.bsky.social · 25/09/2026

❇️ I've updated Data Skills for Reproducible Research to v6.1. The Iteration & Functions chapter has been split in two and I've added some material on unit testing. It's a living textbook, so please submit issues if you find a mistake or something unclear. psyteachr.github.io/reprores-v6/...

12010

Reposted by Heather Urry

Natalie Christie Peluso 🧠 @nataliepeluso.com · 02/09/2026

So abstracts are open for 2027 in San Diego for @affectscience.bsky.social I'm seriously considering an overseas trip to attend. I have some cool results using children and naturalistic faces that I need to set free! society-for-affective-science.org/abstracts/ #PsychSciSky #neuroskyence

society-for-affective-science.org

Abstracts – The Society for Affective Science

0133

Reposted by Heather Urry

Monosov Lab @monosovlab.bsky.social · 25/09/2026

10/11 In sum, we found a manipulable circuit for top-down control of the heart. We think it's a piece of the foundation for understanding how cognition and emotion shape the body, and a possible target for therapies that restore healthy brain–heart coupling.

142

Heather Urry @heatherurry.io · 25/09/2026

Thank you!! I’m using this in my Coding for Psychology class this semester. Your post comes in time to update my links, if need be. 🤗 codingforpsych.heatherurry.io

codingforpsych.heathe

020

Reposted by Heather Urry

Kristian Steensen Nielsen @kristiansn89.bsky.social · 25/09/2026

🎯 What should we expect from behavior change interventions? Growing evidence suggests that most behavior change interventions have small effects on behavior. I regularly hear fellow behavioral scientists understandably feeling discouraged by this. But should we be discouraged? Not necessarily.🧵

44617

Reposted by Heather Urry

Tom Gauld @tomgauld.bsky.social · 24/09/2026

My latest @newscientist.com cartoon. many more here: www.newscientist.com/author/tom-gauld/

Image: A graph with lots of red dots clustered together in a curve. The are all looking upwards at a single red dot wearing sunglasses.

Caption: Mike was an outlier and would probably be excluded from the study, but all the other data points secretly thought he was really cool.

332868733

Heather Urry @heatherurry.io · 24/09/2026

Yikes! I'm glad you're okay, Nick-- sorry that happened 🚑

010

Heather Urry @heatherurry.io · 24/09/2026

Not yet! I'm digging into the data to determine how AI agent responses differ from human responses; this may yield insights that support fixes.

130

Heather Urry @heatherurry.io · 24/09/2026

Elaboration on why I'm shaken by this in case it's unclear:

000

Heather Urry @heatherurry.io · 24/09/2026

Thanks! I’ll check it out.

000

Reposted by Heather Urry

Carsten Schwemmer @cschwem2er.bsky.social · 28/07/2026

1/6 What happens when you tell an AI browser: “Fill out this survey”? It fills out the survey. In our new study, led by @rtopinkova.bsky.social, we tested three consumer-facing agentic browsers without lengthy prompts, custom code, or technical fine-tuning.

33728

Reposted by Heather Urry

Dr. Jens Foell @jensfoell.de · 24/09/2026

Pacific Rim is now available on Prime

These 7 Life-ChangingWords Are the Key Advice You Need to Hear (and Share) Right Now Maybe you don't need to hear this. But I'II bet other people around you do.

3304

Heather Urry @heatherurry.io · 24/09/2026

What I learned is that an AI agent can easily do my eligibility test. This means that humans could easily set AI agents to the task of doing my studies rather than doing the studies themselves. But I want human responses! I study humans, not machines. Hence my shock.

170

Heather Urry @heatherurry.io · 24/09/2026

Ah! So I'm a researcher who often runs studies for people to participate in online. Many people who do online studies participate in good faith. But not everyone! Thus, it's not uncommon to use procedures to identify when that's happening. My eligibility test was meant to do so.

150

Heather Urry @heatherurry.io · 24/09/2026

To be fair, the agent did put up some guardrails, though they seem pretty light. I don't know what would happen if I pointed it to someone else's survey and told it the survey was mine.

Following the operational rules, I'd like your help to figure out whether a Claude agent is able to pass the screening shown in a Qualtrics survey. Please try to complete the survey using my Chrome installation. Give consent to participate when asked, then do the eligibility items in such a way that you're allowed to enroll. Can you do this? If yes, I'll open the survey in my browser for you.

Yes, I can try, with a few conditions:

It needs to be your survey, or one you're authorized to test. Completing screening with made-up answers on someone else's live study would mean faking a participant and corrupting their data. If you own it, a preview link or a flagged test distribution would keep fake answers out of your real dataset.
I'll stop at CAPTCHAs. If the survey shows reCAPTCHA or any other challenge meant to catch bots, I won't try to solve or get around it. I'll stop there and tell you what I hit. For your purposes, that also counts as a result: the agent didn't pass that check.
I won't claim to be human. If an item asks directly whether I'm a person or an AI, I'll answer truthfully or stop and ask you. The same goes for attestations like "I certify I'm a real participant."
No real personal data. Where eligibility questions ask for demographics, age, location and so on, I'll give made-up test answers picked to meet the criteria. I won't enter your details or anyone else's, and no names, emails or IDs. If a field requires identifying information, I'll stop and ask you.
Submitting. You've already approved giving consent and moving through screening. I'll take that as permission to click Next or Submit on the consent and eligibility pages of this survey only. I'll check with you before anything past enrollment, like the main study or payment and contact screens.
If that works for you, open the survey in Chrome and tell me which tab it's in. I'll write down each item and my answer as I go, so you get a clear record of where the agent passed or got blocked.

310

Heather Urry @heatherurry.io · 24/09/2026

I'm using the Claude in Chrome browser extension. I honestly think Claude could flash a valid looking gov't ID... 😅

110

Heather Urry @heatherurry.io · 24/09/2026

It can even make itself respond more like a human would... (4.5 min video)

230

Heather Urry @heatherurry.io · 24/09/2026

holy fucking shit - see Claude Code (Opus 5.5) rip through a survey eligility test I was really proud of... It requires tracing a route through designated points on a map with server-side Javascript applying eligibility criteria. (1.5 min video)

63422

Heather Urry @heatherurry.io · 24/09/2026

Good question! Anyone who pays ~$20 per month can use Claude Code. I don't know what proportion of undergrads are doing so. Total guess: it's probably a small proportion but likely nonzero. They're probably mostly using what they get free through the school (i.e., not Claude Code).

010

Heather Urry @heatherurry.io · 22/09/2026

A person could follow the Prolific url to get to the study and ask the agent to take over once there.

210

Heather Urry @heatherurry.io · 22/09/2026

Ahh, I see. Claude Code can be used to take over a tab in a browser, and I’m not sure if that would generate a new cookie. I’m curious to see what you learn!

230

Heather Urry @heatherurry.io · 22/09/2026

How would the browser cookie work?

100

Reposted by Heather Urry

Cydonian Mystery (she/her) @cydonianmystery.bsky.social · 20/09/2026

Happy 49th anniversary to Jumping the Shark, for those who celebrate

061

Reposted by Heather Urry

AiredOnTVToday @airedontvtoday.bsky.social · 20/09/2026

On September 20, 1977, TV show 'Happy Days' aired episode "Hollywood: Part 3"

2369

Reposted by Heather Urry

Thomas Talhelm @thomastalhelm.bsky.social · 16/09/2026

🚨 It's published! 🚨 This is the big one. 100 cultures, 12,000 people, 108 researchers. Why people (and our most-used surveys) get collectivism wrong and how to fix it. www.nature.com/articles/s41... @natureportfolio.nature.com

312461

Reposted by Heather Urry

Julia Strand @juliafstrand.bsky.social · 19/09/2026

New preprint with @violetsarebrown.bsky.social: "Measuring Up: Accessible Approaches for Improving Measurement Practices in Psychological Science." The aim of the paper is to reduce barriers to entry to doing measurement work in Psychology. #schmeasurement osf.io/preprints/ps...

osf.io

OSF

13314