Investigating Interpersonal Trust


I proposed an honors thesis project studying social interactions and factors that influence the way we connect with others. I am particularly interested in trust in novel interactions. This was a student project that was carried out with the Social Emotions Lab at Northeastern University.


Dr. David DeSteno (Advisor)
Dr. Ajay Satpute (Secondary Advisor)
Shanyu Kates (Graduate Student Advisor)
Fred Duong (Graduate student Advisor)
Matthew Coleman (Graduate Student Advisor)


Researcher, Main Author


September 2021 – May 2022 (9 months)

Research Question

After participating in research assistant positions, I developed an interest in social interactions and the factors that influence how we connect with others. When classes were moved online during the pandemic, I noticed that my classmates mentioned how they found it difficult to make new friends or keep in touch with others. The increase in remote work (in school and in professional settings) led me to wonder what factors influence trust formation. In this project, I focused on whether positive emotional states might influence our propensity to trust strangers.


Feeling grateful, a specific positive emotion, will be associated with increased trust formation towards a stranger. On the other hand, neutral or other general positive states (like happiness) will not increase trust formation.


I was not able to identify any significant correlation between an emotional state (grateful, happy) and differences in trust (as measured by the economic game that I used). In my discussion, I examined my existing research plan and proposed changes to my experimental design that could improve my results.

Final Paper

I wrote my thesis paper and defended it asynchronously with my two advisors. The document was published in Northeastern University's Digital Repository Services, and can be accessed here.
Show the title of my research paper: "The Effect of Gratitude on Interpersonal Trust"

Research Process

This was my first full research project, and it started when I brought up the question of trust formation and online, asynchronous introductions to Shanyu. I had been volunteering in the Social Emotions Lab (SEL) for a semester and during the summer, and I wanted the chance to explore a research question independently.

Preparing for the Study

First, I spent around 2 weeks finding articles related to emotion inductions, gratitude, trust formation, and video conferencing tools on Google Scholar and Scholar OneSearch (Northeastern University's library resource). To keep track of the papers, I organized the most relevant articles into a Google Sheets document and wrote a short summary of each paper to note how it supported my research process.

Research Goals

After my literature review, I was able to articulate my question: “Does our emotional state influence the formation of trust when we meet others?” From the papers that I read, I identified two methods I wanted to use in my study:

1. 5-minute autobiographical recall (an emotion induction task developed by Dr. DeSteno)
2. Berg’s trust game (a method to quantify trustingness towards a partner)

Drafting the IRB Proposal

This portion of the project took the most time and thought. I drafted the initial research plan, with a background section summarizing the existing literature, and a methods section to explain why I chose these two methods. Since I was studying during the first semester, I created a survey with the emotion induction and the trust game, and planned for remote onboarding via Zoom.

Based on my understanding of the literature, I predicted that the gratitude induction would lead to higher trusting behavior towards strangers, because of the prosocial nature of the emotion. I also outlined the analyses that I planned to use with the survey data. 

Conducting the Experiment

Synchronous Onboarding Session

Each session had one researcher (usually me) and one "confederate", which is a person who appears to be a participant but was actually a SEL member. The steps of the study were explained to participants in a synchronous Zoom call, where they were introduced to the confederate but not allowed to speak with them. At the end of the Zoom call, participants were given the survey link. Overall, I conducted 150+ experiments with the help of the lab’s research assistants.

Survey Design

I built the survey in Qualtrics, with the emotion induction, emotion induction check (Likert scale), and trust game embedded within. First, participants would write for 5 minutes in response to a randomly assigned prompt that asked them to recall a grateful, happy, or neutral event. Next, participants were given an emotion induction check so that I could see whether the induction was effective. Finally, participants would play Berg's trust game, where they were asked to give a certain amount of points to the confederate "player". The amount of points they chose to give was used to estimate their trustingness towards strangers, where more points given indicate higher trust.

If you'd like to check out the survey in more detail, please check out the PDF linked here.

Screenshot of the Qualtrics trust survey questions, part 1.
Screenshot of the Qualtrics trust survey questions, part 2.
Screenshot of the Qualtrics trust survey questions, part 3.
Screenshots of the Qualtrics survey questions and prompts.

Literature Review


I analyzed the data with SPSS to investigate the relevant correlations (namely, if a grateful state was correlated with changes in trustingness). An initial check around halfway through the project did show a slight positive correlation between gratitude and trustingness; however, the final dataset for analysis did not show this relationship between the variables.

Although I was disappointed that there were no significant correlations, I learned how to plan for and conduct a research study. In the future, if I plan to explore the future directions that I discussed in my paper, I would know how to set up a project structure to answer my questions. 

Screenshot of the SPSS software analysis of the trust survey data.
Screenshot of the descriptive statistics from the trust research project. Shows that there are no significant differences between the three treatment groups.


Future Directions

Explore how emotional expression and perception affect our judgments of others’ trustworthiness. In this experiment, I focused on how our internal emotional states (experienced emotion) affects the way that we form judgments of others. With more time, I'd like to investigate the other side: how might others’ emotional expressions (e.g., facial expressions, body language) affect the way that we judge them? Also, prior research indicates that there are cultural differences between emotion expression and perception of others’ emotions. Given these differences, would there be differences in preferred emotions for building trust?​

Baseline measurements of emotion, and measuring individual levels of change. 
In my thesis project, I implemented an emotion induction for each participant, but didn’t measure their initial baseline emotional state. After the experiment, I was curious about whether initial emotional states may affect the efficacy of the induction, and if those may be correlated with the resulting trust ratings. In future studies, I would like to have baseline measurements so that I could conduct within-subject analyses to look at changes in individuals, along with the between-subject groups (from the different emotion conditions).​

Trust formation when using digital technologies like VR/AR spaces. Although this study focused on Zoom video conferencing, I'm also interested in how we interact with others in virtual and augmented spaces. Facial emotional expressions are a key part of social functioning, but emotion perception research in VR is currently lacking. One paper indicates perception of some emotions is higher than others. I’m would like to examine where we look when perceiving emotions (eye-tracking), and how virtual spaces impact trust and relationship building.

Back-to-top button. Image is an arrow pointing towards the top of the page, with a circle border.

Other Projects