It’s time to test the VI! I have four versions, all based on the initial feedback from the prototype I tested at Playtest Thursday before the break. For reference, here’s the feedback from 10 users that I received about that prototype:

The green rows indicate the users who made choices in agreement with Prospect Theory. The red are those who did not. I showed these results to Clarisa Diaz, my thesis advisor, who told me I was probably trying to fit too many features into one prototype. With her guidance, and using the feedback given in this preliminary round of testing, I developed four new prototypes (one of which is a control) that test for individual features that may influence to varying degrees, which choice people make. They are

VI-V1 – Control Interface

This interface (VI-V1) presents both options as highly similar, visually. They are the same color and size, with no supplementary visual representations.


This interface (VI-V1) presents both options as the same size, but different colors. In the preliminary user tests, users reported that color affected their conception of the “right” decision, so I have changed the less likely (according to Prospect Theory) choice to the more inviting color (according to previous users).


In this interface, the two options are different sizes, but the same color. Users reported that the larger button seemed like a better choice, so the less likely choice is bigger in this prototype, to invite users to press it.


Finally, this prototype includes a representation of the lives at stake, as users reported this image as influencing their decision. The results for each interface were:

In agreement with Prospect Theory, the majority of users chose the more conservative, less risky option, Treatment A. Here is are data visualizations of the choices for VI-V1, VI-V2, VI-V3, and VI-V4:

Results for VI-V1

Results for VI-V2

Results for VI-V3

Results for VI-V4

The most variation occurred in VI-V4, indicating that users may have been persuaded by the moving images of the people potentially affected by their choice. Because the order of the test was scrambled for each user, it is unlikely they choose a different option in VI-V4 because it was the last test–in most cases it was the first of middle test taken.

Elizabeth Henaff mentioned that giving users a countdown when they reach the options screen might provide a better simulation of the gut-reaction choices we make in organic digital environments. For the next set of tests, I will give users a 5-second countdown to make their choice on the third artboard.

Leave a Reply

Your email address will not be published. Required fields are marked *