Catarina Vales
cvales.bsky.social
Catarina Vales
@cvales.bsky.social
130 followers 130 following 21 posts
developmental scientist. not-that-kind-of-doctor. coffee lover. cat person. parent. she/her.
Posts Media Videos Starter Packs
thanks for the recommendation!
excited to share this work at #cogsci2025, Fri at 1p!
this is great! here’s another (not browser specific, though slightly more work) way to remove the AI preview from google searches bsky.app/profile/cval...
if (like me) you’d like to remove the AI preview when doing a google search, adding “-ai” to your search should do the trick!
Reposted by Catarina Vales
Also talked about The Onion being inefficient on purpose.

www.status.news/p/the-onion-...
Reposted by Catarina Vales
if (like me) you’d like to remove the AI preview when doing a google search, adding “-ai” to your search should do the trick!
Reposted by Catarina Vales
IU has the oldest Folklore Department in the country--no longer a major. Cognitive Science? gone! PhD programs also sacrificed. Full list here: www.in.gov/che/files/In...
www.in.gov
Reposted by Catarina Vales
Or here's a thought--maybe we could hire more actual human teachers, pay them better, make class sizes smaller, build in more paid prep time, and generally just invest in education instead of investing in AI.
Reposted by Catarina Vales
Even accepting the premise that AI produces useful writing (which no one should), using AI in education is like using a forklift at the gym. The weights do not actually need to be moved from place to place. That is not the work. The work is what happens within you.
Reposted by Catarina Vales
Developmental psychologists and statisticians team up to investigate the accuracy and reliability of research tools used to assess how children learn. Read more about how CMU researchers statistically evaluate spatial arrangement tasks with nontraditional data in this featured article! #research
Developmental Psychologists and Statisticians Come Together To Ensure Research Tools Measure Up - Neuroscience Institute - Carnegie Mellon University
Developmental Psychologists and Statisticians Come Together To Ensure Research Tools Measure Up
www.cmu.edu
Reposted by Catarina Vales
Fellow academics, I know that it can feel like capitulation to think about focusing on your work in the face of so many attacks on us right now.

Remember that knowledge is the enemy of ignorance & ignorance is the tool of tyranny.

Doing our work is now a form of rebellion. Keep rebelling.
Reposted by Catarina Vales
Don’t fall into the trap of normalizing any of this with this sort of language that we’ve seen from so many universities, falsely equating the current situation with other presidential transitions

This is unprecedented, unconscionable, and deliberately disruptive.

Have the backbone to say that.
one thing that helps me figure out which books to read is to look up suggestions for books similar to titles i enjoyed. i’m lucky to have access to a great catalog at our local library, i find reading on an e-reader much more convenient (as much as i love the experience of reading a physical book).
getting back to reading for fun was my resolution for 2024 and i read 50 books this year! i did: an app to track my progress (StoryGraph); a kindle case with a strap for comfortable holding; and downloading more than 1 book at a time (so i could switch if i didn’t get into a book).
Reposted by Catarina Vales
We tend to think of anti-vaxxers as affluent, White, crunchy granola-type moms. But that stereotype ignores how precarity creates vulnerability to anti-vaxx messaging.

Take Tara, whose daughter had a sudden shift in temperament not long after a routine vaccine. 🧵1/
open.substack.com/pub/jesscala...
The Lie That Won't Die
Why We're Still Talking About Autism and Vaccines
open.substack.com
all data and code are openly available on OSF!
… — replicating results from tasks that not not measure intended constructs does little to advance theory & failures to replicate may be due to the tasks lacking reliability (and not to the underlying hypotheses being incorrect).
in the discussion, we connect these results to the larger discussions in the field about replicability and open science.

we argue that the current standard of not examining psychometrics of cognitive tasks is a threat to developmental science…
we additionally did some (even cooler!) permutation analyses to see whether the results could be explained by participants randomly placing items on the board — spoiler alert, that seems unlikely!
we then used this measure to assess the psychometrics of the task, which showed appropriate construct validity (β = 0.40), internal consistency (r2 = 0.20), and test–retest reliability (r2 = 0.41; ICC = 0.56).
to capture the structure of a child’s arrangement, we developed this (very cool!) measure that captures the number of clusters, the homogeneity of the clusters, and how well differentiated they are.