Periodic Reporting for period 2 - PAVE (Creating a Panoramic View of Emotion Communication)
Reporting period: 2024-02-01 to 2025-01-31
People communicate emotions constantly. They express their emotions, intentionally and unintentionally, through their behaviour, language, and body; perceivers readily observe and interpret those expressions in others. The dynamics of this emotion communication process are key to successful social interactions. PAVE aims to capture these dynamics by adopting a panoramic approach to emotional communication: looking at how the same emotional experience is expressed in multiple streams (e.g. face-to-face and on social media), and is perceived by multiple observers (e.g. friends, partners, and automated algorithms). Understanding emotion communication better can help people lead better relationships in their personal and professional lives. This is crucial because not only do most people value good relationships in and of themselves, research has shown that good relationships are associated with better health and wellbeing.
The project aims to (i) advance basic science on emotion communication (ii) develop new tools to help future researchers studying these issues (iii) assist in the development of couple intervention schemes to improve romantic and other relationships (iv) assist companies in developing automated emotion sensing capabilities.
The project advanced basic science on emotion communication by generating findings regarding social inferences and relationships, and emotion dynamics in conversations. These findings can be used in the future to assist in development of couple intervention schemes. It has generated a large open dataset examining emotions in friend relationships, coupled with sensor data, which will serve as a valuable tool for other researchers and companies. It has also resulted in a new software package for location data analysis, which can also assist researchers and companies.
The project aims to (i) advance basic science on emotion communication (ii) develop new tools to help future researchers studying these issues (iii) assist in the development of couple intervention schemes to improve romantic and other relationships (iv) assist companies in developing automated emotion sensing capabilities.
The project advanced basic science on emotion communication by generating findings regarding social inferences and relationships, and emotion dynamics in conversations. These findings can be used in the future to assist in development of couple intervention schemes. It has generated a large open dataset examining emotions in friend relationships, coupled with sensor data, which will serve as a valuable tool for other researchers and companies. It has also resulted in a new software package for location data analysis, which can also assist researchers and companies.
A first large dataset has been collected with data on emotions, emotion communication, relationships and behavior from over 300 participants. A preregistration has been posted. Recommendations for recruiting participants for longitudinal dyadic studies have been shared via blog post. A manuscript is under preparation. The general concepts of the study have been shared with high school students via a student podcast and with elementary school students via classes.
A software package allowing for automated analysis of location data - assigning geographic coordinates to meaningful location types (e.g. shop, residential area, commercial area) in a way that is free, privacy preserving, and reproducible has been developed. The package has been publicized via blog post.
A second large study combining daily emotion data with brain imaging has been designed, measures have been implemented and recruitment has started. A preregistration has been posted.
A study on emotional dynamics in conversation has been conducted. Two preregistrations have been posted, analyses have been completed and a manuscript is under preparations. Results have been presented at two scientific conferences.
A study on social inferences in multiple types of relationships has been conducted and published as a manuscript.
Multiple students have been trained, including four Ph.D. students, one M.A. student and multiple B.A. students.
Note that the project was terminated four months early due to the researcher starting a faculty position.
A software package allowing for automated analysis of location data - assigning geographic coordinates to meaningful location types (e.g. shop, residential area, commercial area) in a way that is free, privacy preserving, and reproducible has been developed. The package has been publicized via blog post.
A second large study combining daily emotion data with brain imaging has been designed, measures have been implemented and recruitment has started. A preregistration has been posted.
A study on emotional dynamics in conversation has been conducted. Two preregistrations have been posted, analyses have been completed and a manuscript is under preparations. Results have been presented at two scientific conferences.
A study on social inferences in multiple types of relationships has been conducted and published as a manuscript.
Multiple students have been trained, including four Ph.D. students, one M.A. student and multiple B.A. students.
Note that the project was terminated four months early due to the researcher starting a faculty position.
The project collected data from over 300 participants in ongoing interpersonal relationships (friendships and romantic relationships) for three weeks each, resulting in over 28,000 questionnaire responses and over 10 years of phone sensor data. To the best of our knowledge this is the largest dataset of its kind (dyad experience sampling with mobile sensing) so far. Importantly, the dataset includes both participants' emotions as they experienced them AND those same emotions as experienced by a friend or partner, allowing for multiple perspectives on the same experience. A de-identified version of the dataset will be released to the scientific public.
The project has resulted in new findings on the interaction between relationships and the ways people make social inferences, which have resulted in a published paper. Its findings show that when examined separately, feeling familiar with someone is associated with specific knowledge about them, feeling that someone is similar to you leads to assumptions that their opinions and beliefs are similar to your own, and liking someone leads to stereotypical thinking about them. These findings can advance scientific understanding of social inferences, especially in cases where familiarity, similarity and liking are not strongly connected (e.g. in the case of parasocial relationships where strong liking may occur in the absence of familiarity).
It has also resulted in new findings regarding the emotional arc of conversations, which have been presented in several conferences. Its findings show that people begin conversations by trying to seek common emotional ground, discussing gradually more similar emotional content, but at some point this dynamic stops and people remain at a certain emotional distance, and even being diverging. These findings can advance scientific understanding of conversation dynamics.
Finally, the project has resulted in the development of a new software package allowing for free offline, privacy preserving and reproducible analysis of location data, assigning meaningful location tags to coordinates. This will allow researchers as well as companies to better analyze location data. will allow researchers as well as companies to better analyze location data.
The project has resulted in new findings on the interaction between relationships and the ways people make social inferences, which have resulted in a published paper. Its findings show that when examined separately, feeling familiar with someone is associated with specific knowledge about them, feeling that someone is similar to you leads to assumptions that their opinions and beliefs are similar to your own, and liking someone leads to stereotypical thinking about them. These findings can advance scientific understanding of social inferences, especially in cases where familiarity, similarity and liking are not strongly connected (e.g. in the case of parasocial relationships where strong liking may occur in the absence of familiarity).
It has also resulted in new findings regarding the emotional arc of conversations, which have been presented in several conferences. Its findings show that people begin conversations by trying to seek common emotional ground, discussing gradually more similar emotional content, but at some point this dynamic stops and people remain at a certain emotional distance, and even being diverging. These findings can advance scientific understanding of conversation dynamics.
Finally, the project has resulted in the development of a new software package allowing for free offline, privacy preserving and reproducible analysis of location data, assigning meaningful location tags to coordinates. This will allow researchers as well as companies to better analyze location data. will allow researchers as well as companies to better analyze location data.