Difference between revisions of "Graduate Research Day"
|Line 36:||Line 36:|
==== Wei (Vivian) Fang ====
==== Wei (Vivian) Fang ====
<p style="margin-left:40px"> <br>
==== Hannah Anderson ====
==== Hannah Anderson ====
Revision as of 15:47, 2 February 2021
Graduate Research Day - February 11, 2021
Graduate Research Day 2021 will be held on the Airmeet.com platform.
The link to the conference is:
9:30-10:05 AM Keynote 1: Dr. Mayu Nishimura
10:05-10:50 AM Symposium 1
Effect of Experience on Collective Decision-Making and Social Organization
Tovah Kashetsky, Grant Doering, and Reuven Dukas
Expertise built from experience allows individuals to perform significantly better than novices on a complex task. Social groups can also demonstrate expertise. Within social groups, collective decision-making is crucial for maintaining cohesion, but it is unknown whether a group’s collective decision-making skills can improve with experience. To investigate this, we tested whether repeated experience with choosing between multiple nests during emigration in house-hunting ants (Temnothorax ambiguus) would improve the speed and efficiency with which colonies reach consensus. We hypothesize that experience with decision-making would improve colony performance on future decisions. We first ran preliminary experiments to quantify nest features that colonies prefer in order to construct artificial nests of varying attractiveness. We will provide 20 colonies experience with a choice between a good- and poor-quality nest during emigration, and 20 colonies with no choice during emigration (a single nest). Lastly, we will test all colonies to decide between a good- and poor-quality nest during a final emigration. So far, we found that colonies with experience decision-making do indeed appear to be faster and more efficient at decisions than colonies without experience decision-making. We will also run a social network analysis on 3 colonies from both groups to examine temporal changes in social organization. This will provide us with a mechanistic explanation for how improvements in collective decision-making arise from the actions of individuals. Studying decision-making in ants will allow us to achieve an improved understanding of the development and mechanisms behind expertise.
Depth-specific IOR effect when attention shifts from far to near space relative to viewer
Hanna Haponenko, Hong Jin Sun
Inhibition of return (IOR) is a phenomenon where responses to a peripheral target are delayed if the target appears more than 300ms in the same location as a previous cue. IOR has been extensively shown to operate in 2D scenes. It is not fully understood whether IOR is determined by relative location between cue and target in retinal coordinates or world coordinates. Such a question can be studied by examining IOR in 3D scenes. We compared IOR when cues and targets appeared at same or different depth planes and when depth information was provided by monocular cues. When the cue and target appeared at different depths, a vertical offset was created on-screen, a potential confound to depth perception. We removed the contribution of this confound by contrasting the 3D condition with a 2D control condition that matched cue and target positions but removed all context simulating 3D space. Results showed that IOR magnitude decreased for the different-depth condition compared to the same-depth condition in 3D displays. IOR magnitude also decreased as a function of vertical offset in corresponding 2D displays. Most importantly, such magnitude reduction in 3D displays was higher than that in the 2D displays, but only when the difference in depth was caused by the target appearing at a nearer position compared to the cue. We thus have identified a depth-specific IOR effect in a setting strictly comprised of monocular depth cues, which occurs only when attention shifts from far to near space relative to the viewer.
Memory for global musical structures: Dissecting musical features for their contribution to memory for nonadjacent tonal centers
Joanna Spyra & Dr. Matthew Woolhouse
Memory for musical keys is exceptionally poor. Studies have found that participants maintain a memory for key for only 11-20 seconds after key-change occurs. But music is a complex stimulus with many features; how do these features, such as rhythmical activity or timbre, contribute to the maintenance of memory for past musical sequences? In the Digital Music Lab, we employ a paradigm called “nonadjacent key relationships” to tease apart these musical features and examine their unique effects on memory for key. This paradigm divides stimuli into three sections: (1) a key-defining nonadjacent section, (2) an intervening section in a different key, and (3) a probe cadence either in the original key or in a third key (forming an ABA or CBA relationship between the three sections). Participants are asked to rate the probe for its goodness-of-completion, the idea being that if a memory for the original key remains—despite intervening information—participants will rate the ABA condition higher than the CBA condition. Using this as a baseline, we can manipulate various musical features and compare the strength of completion ratings. If a feature boosts memory, goodness-of-completion should receive a similar boost when compared to CBA conditions. Indeed, this is a pattern we found in many musical features. Results confirm that though memory for key itself may be weak, it is supported by common features we use in music composition every day.
10:50–11:00 AM Break 1
11:00-11:45 AM Symposium 2
Body sway reflects nonverbal communication in a string quartet learning to play unfamiliar music together Emily Wood, Dobri Dotov, Andrew Chang, Dan Bosnyak, Lucas Klein & Laurel Trainor Ensemble musicians must anticipate their partners’ actions to coordinate playing a piece together. To achieve this, musicians attend to sensorimotor signals embedded in their partners’ body sway movements. Indeed, a musician’s body sway movements reveal their upcoming intent regarding phrasing, tempo, and dynamics, which helps their partners anticipate how and when to play next. We have previously measured the body sway of expert musicians in small ensembles with motion capture, and used Granger Causality (GC) to calculate bidirectional influence, or information flow, between the body sway of each musician in the ensemble. We showed that information flow was greater from assigned leaders to assigned followers than vice versa, and that group information flow was greater when musicians played with emotional expression than without. Here, we show how information flow changes in an ensemble that learns to play unfamiliar music together. A professional string quartet came into the LIVELab and played two unfamiliar pieces of music together eight times in succession while body sway motion data was recorded. Linear mixed effect modelling showed that information flow within the group decreased significantly across trials for both pieces, suggesting that musicians relied on body sway to help them play together when the pieces were most novel (trial 1), but this reliance decreased as they gained familiarity with playing the pieces together. We are currently completing cross-correlation analyses to examine how the similarity of group body sway movements changes across trials. Overall, our studies show that body sway reflects nonverbal communication in musical ensembles.
Wei (Vivian) Fang
Dominance pulls faces closer
Wei Fang, Cristina I. Galusca, Zhe Wang, Yu-Hao Sun, Olivier Pascalis, Naiqi G. Xiao
Perceived social traits, such as dominance and trustworthiness, affect other people’s behaviors. While the impact of social traits has been consistently found in high-level cognitive processing, it is unclear whether social traits also modulates perceptual processing of faces. To this end, we investigated how facial dominance affects the perceived distance of faces.
We used an implicit but highly robust perceptual illusion to measure the perceived distance: when two identical faces are presented vertically (one above the other), relative to the top face, the bottom one appears closer. Observers exhibit a strong bias to indicate the bottom face is bigger.
We examined how facial dominance influences the perceived distance with a set of computer-generated Dominant and Submissive faces. If facial dominance makes faces perceived closer, participants will likely report the bottom one is bigger. To probe the generality of this effect, we tested this effect in Canada, China, and France (N = 30/country) with faces from three races (African, Asian, and Caucasian).
Across the three countries, participants showed a significant bias in choosing the bottom face as the bigger (Mean bottom responses = 72.03%, p < .001), replicating the illusion. Moreover, Dominant faces led to a stronger illusion than Submissive faces (p = .009), suggesting that facial dominance led faces to be perceived closer. No effect of face race or country were found.
As facial dominance is often associated with negative signals, our finding suggests an evolutional mechanism in the visual system, which amplifies dangerous signals in the environment.
- Variation and correlations of behavioral lateralization
1:00-2:00 PM Symposium 3
2:00 - 2:45 PM Poster Session
- Rachael Finnerty
- Cindy Tran
- Seyedbehrad Dehnadi
- Konrad Swierczek
- Emma Marsden
- Carly McIntyre-Wood
- Peter Najdzionek
- Maya Flannery
- Jiali Song
3:00 - 3:30 PM Lightning Talks
- Vidhi Patel
- Lucas Klein
- Jamie Cochrane
- Janice Yan
- James Mirabelli
- Oral Presentation: $75
- Poster Presentation: $75
- Lightning Talks: $50
Winners to be announced...