Thank you so much Mariam!!

Thank you Aaron!

Thank you so much Steve!

Thank you Christian! Hope to see you at the next PPFP event and finally chat in person 😁

I am incredibly thankful for my PhD mentor Dr. David Whitney and my postdoc mentor Dr. Katherine Rankin for supporting me as a researcher and championing me. A huge thank you to @UC_PPFP for this incredible support!

Excited to share that I've been awarded the UC President's Postdoctoral Fellowship @UC_PPFP! I'll be continuing my research on emotion perception and context integration in frontotemporal dementia at UCSF's Memory and Aging Center (@UCSFmac) with the @Rankin_Lab!

Integration of affective cues in context-rich and dynamic scenes varies across individuals - Nature Communications Here, the authors examine how people combine facial expressions and context to understand others’ emotions. They find that most people use a Bayesian integration strategy, where cues are weighted base...

I want to thank my advisor David Whitney for his endless guidance and support on this project, my collaborator @ymurai_eng for helping make this project possible, and to @BerkeleyPsych for being an amazing place to do research!

Read more at: doi.org/10.1038/s414...

Key takeaway: How you read the room might be totally different from your friend!

The biggest surprise? Everyone has a unique strategy! While many observers' judgments were best predicted by the Bayesian integration model, others' integration strategies resembled the simpler "averaging" method more, ignoring cue ambiguity.

We found that the Bayesian model outpeformed the simple Hueistic model in predicting individual human judgments of emotion. This suggests that most of us constantly calculate how reliable a cue is — if a person's facial expression is ambiguous, we place more trust in the context!

One thing was clear from the data: single-cue models failed. Models that relied only on the face or only on the context could not predict human emotional judgments. To truly understand the social world, we must integrate multiple sources of information.

We tested two main models: Do humans combine cues using complex calculus" (the Bayesian model), weighing less ambiguous cues more heavily? Or does the brain take a simpler approach and just "average" everything together without considering cue ambiguity (the Heuristic model)?

How do you know how someone is feeling? We don't just use facial expressions (a), we combine expressions with "context", like body language and scenes, to truly understand emotions (b). Our study investigated the brain's computational mechanisms for integrating these cues in real-time.

Big thanks to my advisor David Whitney for his necessary guidance and support on this project, and to
@BerkeleyPsych for being an amazing place to do research!

These findings demonstrate robust individual differences in affect overestimation, indicating that some observers may recall affective events more vividly than others. Such differences may reflect a predisposition toward underlying psychopathology. (7/7)

Moreover, the magnitude of overestimation was consistent within individuals, remaining stable across different days and generalizing across different stimuli. (6/7)

In our study, we examined whether overestimation in observers’ recollections of affective events is unique to each individual. We found that affect overestimation occurred when observers made summary judgments about the emotions of people in recently viewed videos. (5/7)

a group of cartoon characters are standing next to each other ALT: a group of cartoon characters are standing next to each other

Prior research has shown that ensemble judgments of emotions are prone to overestimation biases, suggesting that people tend to amplify emotional content in memory. (4/7)

This summarization, known as ensemble perception, refers to the visual system’s ability to extract summary statistics from large sets of visual input. (3/7)

a man taking a picture with a camera with #schitts creek on the bottom ALT: a man taking a picture with a camera with #schitts creek on the bottom

Humans rely on recalling past events to make effective future decisions, but this process often requires summarizing large amounts of information. (2/7)

Excited to share that our paper "Continuous affect tracking reveals that overestimation during the recollection of affect is idiosyncratic and stable" is out now in @ARVOJOV !

doi.org/10.1167/jov....

A short thread on the findings below 👇🏽📷(1/7)

Four days left to donate! All funds raised will provide research opportunities for underrepresented students to get one on mentoring on a research project. Learn more about the program by visiting the fundraising page!

Two weeks left to donate to the campaign! Please share with your network if you can!

With your support, we will be able to further support students accepted into the program and continue developing and improving the program for future cohorts! 🧵(4/4)

We invite you to check out our page and get more familiarized with the core aspects of the program! If you’d like to directly support the program and its mission you can donate directly and/or share the link with your network on social media to increase its exposure! 🧵(3/4)

The REP program was created to provide underserved and underrepresented students with immersive research experiences, mentoring, and professional development, as well as communal activities with aims to advance diversity and multicultural excellence in academia. 🧵(2/4)

Hi BlueSky!

The Research Experience Pathways (REP) in Psychology program is participating in Berkeley's DEIBJ fundraising campaign for the month of November!

crowdfund.berkeley.edu/project/39859

🧵(1/4)