Jaime Castrellon
LightNews LightNews
banner
jjfcastrellon.bsky.social
Jaime Castrellon
@jjfcastrellon.bsky.social
720 followers 320 following 8 posts
Assistant Prof @UCLA | studying decision making & motivation | #LatinxInSTEM | #FirstGen | he/him

https://www.psych.ucla.edu/faculty-page/jcastrellon/
Posts Replies Media Videos
jjfcastrellon.bsky.social
Overall, students gained a greater appreciation for SAN research and even though this is not a stats or programming course, on average, they felt more comfortable running tests and familiar with code.
plot showing data comparing students familiarity with social and affective neuroscience during week 1 vs week 10 of a course. Plot shows significant increase in mean familiarity level. plot showing data comparing students familiarity with Python during week 1 vs week 10 of a course. Plot shows significant increase in mean familiarity level. plot showing data comparing students comfort with statistics during week 1 vs week 10 of a course. Plot shows significant increase in mean comfort level. Example student testimonials from course: Question: "What do you feel you got out of this course?" 1.) "I feel like it demystified a lot of psychological research methods - I too can produce incredibly complicated and convoluted research findings!" 2.) "I feel like the topic of social and affective neuroscience is more approachable to me and it solidified my interest in social psychology and healthcare as I found myself wondering how to apply these findings in a clinical setting. The guided nature of the course helped me understand concepts without the stress usually associated with learning that I have experienced thus far in college." 3.) "I learned a lot about how neuroscience works in a more psychological than medical context. Labs that focused on neuroscience used to intimidate me but they don't anymore." 4.) "I liked the experience in data analysis and using different softwares and python to draw conclusions. I also never thought of how to define/measure emotion, which has a large influence on our behavior. I'm interested in studying memory, which is enhanced for highly emotional events so it was fun to connect the content in this course to my interests."
December 20, 2024 at 6:16 PM
jjfcastrellon.bsky.social
3rd project: after learning about fMRI, students used preprocessed data on OpenNeuro from @dvsmith.bsky.social, Dominic Fareri et al., for ROI analysis. They chose 1 of 3 social decision tasks, defined ROIs via Neurosynth coordinates they searched, & tested hypotheses. Code ran in Colab via Jupyter.
Screenshot of Jupyter notebook for ROI analysis. Screenshot of one student team's class presentation. Slide shows the motivation for their selected region of interest. Screenshot of one student team's class presentation. Slide shows a specific hypothesis formed by the student team. Screenshot of one student team's class presentation. Slide shows the ROI analysis results with a plot and reported statistics.
December 20, 2024 at 6:16 PM
jjfcastrellon.bsky.social
2nd project: After learning about experience sampling of emotion, students formed their own hypotheses about daily emotions/behavior. Using expiwell, they designed & deployed 5-day surveys for classmates, analyzed data in Jamovi, and presented findings in scientific posters! www.expiwell.com
one student team's poster presentation from experience sampling data on the relationship between media consumption and emotion one student team's poster presentation from experience sampling data on the relationship between in-person versus online social interactions and emotion
December 20, 2024 at 6:16 PM
jjfcastrellon.bsky.social
1st project: After learning about facial EMG, students searched for face images showing emotions (joy,fear,etc). In Google Colab, they used py-feat (@eshjolly.bsky.social, @lukejchang.bsky.social, et al.) to predict emotions, comparing model vs human judgments. Code was novice-friendly w/few tweaks!
example outputs of emotion expression prediction from face images that students selected screenshot of class crowd-sourced data on facial expression predictions slide presented in class to help students visualize differences across their teams in how well the model predicted the emotions
December 20, 2024 at 6:16 PM