1. OVERVIEW
  2. Problem Definition
  3. Solution
  4. My CONTRIBUTION
  5. Generative Research
  6. Lo-Fi Prototyping
  7. Think-Aloud PROTOCOL
  8. Iterations
  9. DESIGN DECISIONS
  10. PRESENTATION
  11. REFLECTION

Date Assist

User Research | UX Desgin
A plug-in for existing dating apps that implicitly delivers self-affirmation interventions to users to boost their self-concept prior to a first date.
Team: Aaron Bishop, Cori Faklaris, Samantha Reig, Bidisha Roy & Molly Schaefer
September 2019 — December 2019

MY ROLE

  • User Researcher
  • UX Designer

CONTEXT

Final project for Persuasive Design in HCI course, taught by Geoff Kauffman

METHODS

  • Academic Research
  • Semi-Structured Interviews
  • Rapid Prototyping
  • Think Aloud Protocol

Problem Definition

How might we leverage self-affirmation theory to implicitly reduce the anxiety of a first date?
What is self-affirmation? In this process, a person reflects on valued aspects of the self, allowing those positive aspects to counter the effects of negative self-perception. Such a reflection in one domain can reduce threats in unrelated domains by shifting from a narrow focus on the immediate threat to an expanded perspective of one’s self-worth.
Our team set out to embed self-affirmation within a technological solution that users would want to use, which could adequately boost their defenses ahead of a scheduled threat. After exploring the space and considering topics such as public speaking and test-taking, we identified dating apps and the accompanying first dates as the most potential for successful self-affirmation intervention.
Dating app users can face anxiety and apprehension at every step in the process of a date—their profile setup, the matching process, messaging their matches and scheduling up a date, and of course, the date itself. If users face rejection, it can severely damage their self-worth.

SOLUTION

A plug-in for existing dating apps that adds calendar-based scheduling and personalized playlist building.
Our plug-in for online dating apps uses a touch-screen calendar interface to help singles schedule their first date. The plug-in also encourages users to build a personalized, meaningful playlist to play before their date — and, if they so choose, to share one of their favorite songs to break the ice and ease conversation.
FEATURES
Calendar with shared availability
By setting preferred availability before matching, the calendar feature prevents stifling scheduling issues.
Provides date suggestions to users
Prompts users to create personalized playlists
Provides a quick icebreaker, if the user wants one

My Contribution

I led User Research and Prototyping efforts, focusing on the self-affirmation portion of “Date Assist.”
I was fortunate to have a thoroughly multidisciplinary team. Our research team members focused on psychology and experiment design, while myself and the design team took on prototyping and user research, while. Specifically, I focused on the application of self-affirmation theory through experience design. Some of my notable contributions:
• Conceptualizing, prototyping, and testing the music-based self-affirmation solution
• Leading interpretation sessions to synthesize findings from interviews and think-alouds
• Mapping out the user flow of our intervention

Generative Research

By synthesizing 10+ scholarly articles and a session with an expert on self-affirmation, we narrowed our focus to dating.
In addition to analyzing research papers, we met with faculty from the CMU Psychology department who specialized in self-affirmation research.
Through 8 target-user interviews, we found common pain-points in dating (scheduling, conversation topics) and identified music as a hallmark for self-identity.
We asked about dating app usage, matching process, getting ready for a date, how they feel on the date, how they feel after the date. Separately, we asked about core values (a mainstay in the traditional self-affirmation exercises) and music, and how they do or do not fit into their dating routine.
After our interviews, we came together as a team for an interpretation session to synthesize insights across participants.
> INTERVIEW Insight 1
When scheduling doesn’t line up, it can feel like a rejection.
> INTERVIEW Insight 2
For many, music is both an established part of a getting-ready routine and integral to feeling like themselves.
> INTERVIEW Insight 3
Two common pain points were ‘where to go’ and ‘what to talk about’— participants were more willing to accept guidance on the former than the latter.
Even though ‘not knowing what to talk about’ was a near unanimous pain point of the dating experience, the interviewees often said that overcoming this with their date was an important step towards identifying them as a romantic partner.

LO-FI PROTOTYPING

As a team, we ideated many different paths in which the user journey could take shape.
We also began to identify methods towards measuring the impact of our intervention.
Four of us each designed unique lo-fi prototypes, so we could test multiple ideas at once.
I focused on creating the self-affirmation intervention, while other team members focused on scheduling, suggesting locations and events.
The prototypes I created included one value-reflection exercise (framed as a conversation bank) and one music-reflection exercise (framed as a ‘good vibes’ playlist-builder).

THINK-ALOUD PROTOCOL

Testing prototypes with 8 think-aloud studies helped us understand user desires and comfortability.
Participants tested the prototypes on a smartphone, using Figma Mirror.
> TAP Insight 1
To elicit a potentially self-affirmative response, verbiage is very important.
One participant chose music that she personally identified with because it was “chill” or “hype.” However, this did not elicit the self-reflective affirmation experience that we were hoping for.
> TAP Insight 2
Music can be associative due to the memory of the song’s audio, or due to the content of the song’s lyrics.
One participant, after choosing the value “Family,” was asked to think of a specific positive memory with her family (which she was able to do), and then a song that made her think of her family. While she could not think of a song that reminded her of family, she chose “Happy” by Pharrell, because family made her happy. While the music choice in this case did not evoke the self-affirmative response on its own, it opened the opportunity for having our app’s intervention actually form an association
> TAP Insight 3
Users were heavily divided on what is ‘too personal’ to divulge before meeting in person.

ITERATIONS

Through user testing, we were able to identify interventions that did not translate well.
Aside from the “music selection” intervention, we also tested multiple prototypes involving “conversation starters” or “photo selection,” but users had mixed reactions to these, with some very strong negative reactions.
We revised verbiage and framing with each iteration, getting closer to an experience that both felt valuable to the user and could be self-affirmative.
One key element that we added in the second round of tests was giving users a choice to share 1 of their 3 songs (or none). This satisfied users who wanted to share some of themselves before a first date and users who would prefer no interaction before the first date.

Design Decisions

1. Embedded design: frame the self-affirmation exercise as a “playlist builder.”
Rationale: To make the intervention more effective, we wanted to conceal the self-affirmation and ensure that the exercise was as implicit as possible. In addition to framing the self-affirmation exercise, we also intermixed ‘off-topic questions’ with self-affirming questions to further mask the embedded design.
2. Enable users to share their results, but only if they wish.
Rationale: The “song share” aspect justifies the playlist builder’s place in a dating app, and gives a new way to interact with their upcoming date.
3. Employ reciprocity to heighten user comfort
Rationale: We locked the song so that the date only sees it if they share one in return--this reciprocity makes it less nerve-racking. If users choose to be ‘vulnerable’ by sharing a song, at least they know that their date has to share something back before viewing it.
4.  Utilize the scheduler feature to strategically deploy self-affirmation exercises
Rationale: The scheduler provides value for the user by making it easier to see their matches’ availability and schedule dates; it serves a double purpose by also allowing the app to know when the user’s date is scheduled. This allows the app to push notifications at the best times for the self-affirmation exercise (~24 hours before the date) and the reminder of the self-affirmative content (2 hours before the date).

Presentation

We presented our design through testable high-fidelity prototypes and a one-sheeter.

Reflection

We learned self-affirmation is very different from simply boosting someone’s self-esteem.
Our early ideation included enabling friends to boost each other’s self-esteem. Early feedback we got from our faculty advisors was that this would not elicit a proper self-affirmative response. It would have to be implicit, and it would have to involve a specific upcoming threat. This provided a greater, more complex design challenge for us to overcome.
Dating, especially, is a domain where every user wants something different.
It was important to ensure our design could cater to many different types of users. We ensured that the interventions we were offering, while appealing, were optional, and gave the user many ways to customize their experience.
In order to truly test the efficacy of our design, we created a research plan proposal.
To investigate whether our self-affirmation exercises would be effective in online dating, we propose a 4 x 2 between-and-within-subjects study in which 600-800 participants will use our plug-in for a period of several weeks and be exposed to different combinations of self-affirmation activities. All participants will complete an onboarding questionnaire, and will periodically answer intervention questions that require reflecting on (a) a valued domain (the classical self-affirmation exercise), (b) identity-related songs, (c) a valued object, or (d) a domain that is not their self-reported most-important (the control). Various factors such as self-reports and whether the pair had a second date will be used to measure effectiveness.