Summer Sprint 2: Probing the details of perceived privacy and agency

During sprint 1, our work allowed us to learn about how people engage with data collection/usage and their reaction towards a privacy meta control. But since our last prototype used a multiplayer VR game(Ping-pong) and a very high-fidelity dashboard, the user’s attention was not focused on the perceived privacy and how they felt about data collection/usage during the testing.

Sprint 1 Test

Thus, during this sprint, we conducted two concurrent tests:

1. The perceived privacy follow-up test with the 4D cinema experience where we aim to learn whether the timing of privacy disclosure affects the user’s experience and the users’ boundaries of unknown data collection.
2. The agency & control test with a personal trainer on elliptical where we aim to gain an understanding of mechanisms of trust-building that let users forego control for an immersive experience

The Perceived Privacy Follow-up Test with 4D Cinema

9 participants, 20 min, 3 A/B testing groups with 3 participants in each group

The Test:

We aim to be more granular for the follow-up test to understand the optimum timing of privacy control/disclosure and whether it will affect the user’s immersive experience and perceived privacy.

Moreover, we aim to learn about the users’ acceptance level/boundaries of unknown data collection and whether they will feel uncomfortable if they are notified at the end? Would they feel violated, and what would they do in response to the lack of trust in the system?

biometric-data-based

During the test, we included three separate groups where:

1. We explain what sensors are present, how data is collected/used, and give users the privacy control in the beginning
2. We explain…the privacy control in the middle
3. We explain…the privacy control in the end

WizOfOz facilitator talks to participants by pretending as an omnipresent voice agent(sits outside, and the voice gets projected to the room)

Sensors used or described(wizard-of-oz)

• A camera on the wall recording
• Watch to collect biometric
• Glasses collecting eye-tracking
• Walls and floors can sense your weight, movement, etc.

Initial findings: ___________________________________________________________________

• When new concepts like the biometric-data-based 4D movie were introduced during an experience, people felt safe engaging with the experience and trusted the system when it included something they were previously familiar with.
• The timing makes a huge difference in participants’ comfort in data collection. Prompting, in the beginning, received positive feedback and neutral feedback when it is at the end. Prompting in the middle noticeable disrupted the experience.

The Agency & Control Test:

6 participants, 25 min, 3 testing groups with 2 participants in each group

“An all-connected world should allow users to forego control with a deep sense of safety to accomplish goals, without sacrificing their personal agency.”

Process

We started with a reverse engineering approach where we defined the end goal first: what we wanted to learn at the end of the study. Each team member individually brainstormed on what concepts we envisioned should be tested.

Brainstorming, discussing, and voting on the concept to test for agency&control

After discussing and voting, we have settled on one concept: How might we persuade users to let go of the control in an immersive experience(autopilot)?

The test went through many iterations because our initial idea was to implement it on a treadmill, but due to the uncertainty of the treadmill, automatically changing levels such as speed may bring some dangers. Moreover, one of the uncertain factors brought about by exercise is that everyone’s limit is different. The reason for calling to stop or uncomfortableness may not only be a result of not having control but because they have reached their own limit.

Brainstorming session to reconsider what could be the best test implemented to test agency & control

Thus we returned to the brainstorming stage to reconsider how the test should be implemented. Because foregoing control is subjective to the domain and the user’s job-to-be-done, we had difficulty developing the ideal scenario to test our goal.

We thought of various solutions, including a less intensive workout or simply keeping away from exercise and returning to the domain of watching/interacting with an immersive experience. But the problem with immersive experience is that
1. We can only test with students we know in the MHCI room.
2. Except for the VR, other experiences, such as immersive movies, are not dynamic enough, and users don’t have much agency over the experience

After hours of discussion and reflecting on the pros and cons, we eventually decided to keep the original idea and intercept people at the gym and ask them to participate, but on an elliptical and amplified the variable for each persuasion technique.

The Test:

Our goal for the test is to understand the mechanisms of trust-building and what kind of persuasion will make the user forego control. We assumed that trust could be established by showing social validation from peers/celebrities/friends or personalizing the experience based on the participant’s biometric data.

During the test, there were 3 groups, each with 2 participants.

1. The group where we automate the experience without persuasion
2. The group where we use social validation as a way to persuade the user to let go of control
3. The group where we use the personalized workout data as a way to convince the user to let go of control

Initial Findings

___________________________________________________________________

1. Data-backed recommendations and personalized routines enhance users’ trust in the system, which had made them willing to forego control when the trainer asked them to push more beyond their limits.
2. Compared to the control group, the personalization group is more willing to accept the instructions and give away control than the social validation group.
3. All participants liked the aspect of automatically controlling the elliptical and adjusting the difficulty levels, but they did not like the routine being determined for them.