top of page

Here are representative quantitative projects. I draw from prominent HCI and social psychology research to identify users' current struggles (not tied to products) and their pain points with existing digital products, develop prototypes with design solutions, and conduct empirical research to validate prototypes and make actionable recommendations.

Project 3. Designing Similarity Cue to Promote Well-being in Social Media Users

To protect social media users’ well-being after engaging in upward social comparison, we designed a “similarity cue” that emphasized the overall similarity between the user and the upward comparison target. A 0% similarity cue emphasized low overall similarity, whereas 50% and 90% similarity cues emphasized high and moderate overall similarity. Would these cues result in positive user outcomes after upward social comparison? Read about the study here

Skills

  • Website prototyping 

  • Survey design

  • Multivariate testing (i.e., a cross-sectional between-subjects experiment with 9 versions of a chatbot tutor)

  • Quantitative data analysis (i.e., ANOVA, correlation, exploratory factor analysis, mediation analysis)

Design Recommendations

  • Show moderate overall similarity in percentage, which can be quickly processed.

  • Social media must explicitly emphasize the overall similarity. Facebook shows users the overall similarity tacitly, but it should not do that (e.g., users see which group of friends they have in common with other users via a feature called “Mutual Friends”).

  • Establish overall similarity by emphasizing their similarities over trivial personal information (e.g., hobbies, movie preferences). 

This image shows correlation between major study outcome variables.
The effect of similarity cues on key user outcomes after upward social comparison.
This image shows the mediating role of perceived overall similarity.

Project 1. Designing Chatbot Intervention to Reduce Imposter Syndrome

This study prototyped 3 versions of a chatbot-based growth mindset intervention to reduce imposter syndrome in undergraduate students. We examined whether students who teach a chatbot mentee about a growth mindset would report a greater reduction in imposter syndrome than students who are taught the mindset by a chatbot mentor, the scenario that is prevalent in many chatbot-based therapies. You can read the study's short report here. 

Skills

  • Chatbot prototyping and conversational script design

  • Survey design

  • Multivariate testing (i.e., a cross-sectional between-subjects experiment with 3 versions of a chatbot-based growth mindset intervention)

  • Quantitative data analysis (i.e., trend analysis, correlation, ANOVA, binomial regression, exploratory factor analysis, multiple regression)

Design Recommendations

  • A growth mindset intervention should introduce a chatbot as a user's mentee, not a mentor.

  • Let a chatbot asks a user to share about the time they had positive beliefs in effort to succeed in class.​

  • A chatbot mentee demonstrates it has learned a lesson from a user.

    • A chatbot repeats what a user has said about a lesson.

    • Include a notebook feature where a chatbot notes what it has learned from a user, who can check the note at any time. 

This image shows a number of words exchanged between participants and their respective chatbot partner.
This image shows correlation between study variables.
This plot shows the relationship between imposter syndome and each predictor.

Project 2. Exploring Students' Feedback Acceptance from Chatbot Tutor

This study examined whether students are more likely to accept negative performance feedback coming from a chatbot tutor than from a human tutor. I hypothesized that a chatbot tutor’s objectivity and lack of a mind would minimize the negative feedback’s threat to self-esteem and public self-image. You can read the study's short report here

Skills

Design Recommendations

This image shows participants estimated mean on outcome measures, broken down by conditions.
This image shows correlation between major study outcome variables.

Try out a chatbot prototype for this study.

  • Chatbot prototyping and conversational script design

  • Survey design

  • Multivariate testing (i.e., a cross-sectional between-subjects experiment with 9 versions of a chatbot tutor)

  • Quantitative data analysis (i.e., correlation, ANOVA, exploratory factor analysis, path analysis)

  • Let a chatbot tutor communicate negative feedback asynchronously (vs. real-time feedback).

  • Include positive emojis and a "sandwich" technique (i.e., negative feedback is placed between two positive feedback) to deliver negative feedback.

  • A user does not want to be reminded of how bad they are performing by receiving negative feedback, and a learning platform can empower users to opt out from social competition features (e.g., Duolingo's Leaderboards).

Project 4. How Immersive Journalism Affects User Perceptions and Cognitions

We investigated how reading stories with VR headsets and 360°-video will influence users' reading experiences, including a sense of presence and their perception of the company's expertise and story-sharing intention. Would these storytelling modalities result in better user reading outcomes than stories written in text? You can read the study's short report here

The main effect of storytelling modality on key user reading outcomes
The indirect effect of storytelling modality on trustworthiness and expertise (VR-to-Text comparison)
The indirect effect of storytelling modality on empathy and story-sharing intention.
bottom of page