Mobile Info-Art

Design and evaluation of a lock screen wallpaper that shows customized information in the form of artwork.

What if we could hide data in art so that only the person looking at it understands the information the artwork is conveying? These displays are referred to as info-art. Furthermore, is the recall rate higher for a display like this when compared to equivalent text-based displays?

Project Goals

Perform literature review of existing peripheral info-art displays

Collect data on what information people would like to see and in what form

Design the info-art display and build a legend of what data corresponds to which element

Design and conduct an experiment to evaluate recall rates of info-art

Jump to results

Related Work

In Prof. John Stasko’s InfoCanvas project, users had a higher recall rate for data embedded in art than standard text for a web display. We want to bring this to a mobile display and test its effectiveness as a peripheral source of information. Further background and related work can be found in my research paper.

Data Collection

We sent out a survey to collect what kind of data users like having access to on a daily and regular basis. We also collected data on what visual element changes they associate with what kind of data. For example, a majority of respondents said that they associate temperature changes with colour while they associated step count with position.

Visual Design

The first step was to design some artistic lock screen wallpapers. The designs were made while keeping in mind that they need to have multiple discernible elements that can convey various data points. I designed these initial ideas on Adobe Photoshop. Some of them used Adobe's new generative AI feature for ideation.

I then decided on one design and defined the various elements and their changes that corresponded with various types of data.

Experiment Design

To assess the efficacy of the designed info-art display, it's crucial to compare it with an equivalent text-centric display. I developed a mobile display showcasing identical data points, but in the form of text widgets. Subsequently, I designed three separate versions of this display with distinct data and constructed the corresponding info-art displays.

Experiment Setup

To evaluate the capabilities of the two displays as peripheral sources of information, the experiment needed to be in a dual-task format.

The primary task given to the participant was to browse through a Wikipedia article on a primary device i.e. a laptop. The chosen article was about Australiformis, a parasitic worm from Australia.

The secondary device was a phone screen with either the info-art or the text information displayed on it and kept next to the laptop. The participants were asked to interact with the Wiki page and glance over at the phone screen for a duration of 4 minutes.

Since the two tasks were unrelated, it’s expected that participants would use a discrete split of their full attention to attend to both. I did not place any indication of priority or urgency on any aspect of the experiment. Since both tasks required a moderate level of cognitive processing, it can be considered a fair setup to test ambient information consumption.

Participants

Control Group

CONTROL

GROUP

EXPERIMENTAL Group

EXPERIMENTAL

GROUP

Participants were recruited by sending out a Microsoft Form calling for voluntary participation. The form asked for their demographic information, gave a brief overview of the nature of the experiment, and asked them to fill out a consent form. I then contacted the qualifying respondents and scheduled meeting times with them. Five respondents were put in the control group and another five in the experimental group.

Test Results

After the duration of the experiment, participants were each asked to take a test on the information they saw and remembered from the mobile screen. Each correct answer is given 1 point. The maximum possible score on this test is 10 points. The results are as follows.

Thus, we see that recall rates are almost consistently higher in the experimental group. This shows that the info-art display is a more effective mode of transmitting information peripherally than a text-based equivalent. The info-art display performs almost 19% better in terms of recall rate!

Limitations

Limiting Graphics

The art elements are not currently customizable
Mappings may be unintuitive
Flashy art style may be distracting to some

Short period of observation

The 4 minute exercise may not be as accurate as results after prolonged usage

The 4 minute exercise may not be as accurate as results after prolonged usage

Primary task validity

Lack of personal stake in the primary task may have caused an inaccurate distribution of attention to the peripheral device

Slightly unfair comparison

The control text-based display is quite bad, current systems are more standardized (including icons)

Testimonials

"I thought it was very cute and I loved it. I would definitely use such an application as my wallpaper."

"I would definitely use it! It would be great as a lock screen of a digital device."

"Will need someone to look at it a couple of times to get all the information mappings."

Reflection

Creating this display and assessing its effectiveness offered me the chance to concentrate on usability-testing and user inquiry. I understand the significance of supporting innovation and design with evidence of efficacy. If given further time, I’d conduct additional surveys with a broader population of participants, perhaps with more custom data visuals, to examine mobile info-art as a real-world system.

This project represents a fusion of my effort in two courses at Georgia Tech. In CS7470-Mobile and Ubiquitous Computing guided by Dr. Thad Starner, I designed the info-art system post an extensive literature review of peripheral and ambient information systems. With my project group in the class, we designed an app that dynamically updated the screen with data sourced from the web. For CS8803-Learning Research Design under Dr. Jessica Roberts, I was tasked with conceptualizing a research exploration based on a hypothesis statement. The presumption was that our info-art display as a peripheral display will record a higher info recall rate than an equivalent text-based display. I carefully designed the experiment based on relevant literature.