PROJECT
Keypad Comparison Study
Using Paired Comparative Multidimensional Scaling to identify which keypad variables appeal most to users.
My role: User Researcher
My contribution: I used Paired Comparative Multidimensional Scaling to identify which variables appeal most to end-users and that by focusing on those features you can optimize the user experience, which in turn will drive more product sales.
​
Introduction:
How do you determine which variables make certain keypads more desirable than others? One method is to use Paired Comparative Multidimensional Scaling to identify and call out which of these variables appeal most to users. Surprisingly, there are quite a few variables involved in creating a keypad that can affect the user experience.
Methodology
How does Paired Comparative Multidimensional Scaling work?
First, you identify clusters of variables that are closely related, and then identify which clusters represent desirable variables and why.
​
So to start, we created some custom keypads with a variety of differences between the keys. Participants were then asked in a randomized order to compare and contrast different keys, and then select from an 11 point Likert scale which key they preferred and by how much. Here are some example paired comparisons with the participant's responses bolded.
A 5 4 3 2 1 0 1 2 3 4 5 B
A 5 4 3 2 1 0 1 2 3 4 5 C
B 5 4 3 2 1 0 1 2 3 4 5 C
Results
The data output was a cluster analysis, which I reviewed across three different dimensions to identify all of the different clusters. If you only looked at the cluster analysis across two dimensions for example, you might miss identifying a cluster that only appears along the third dimension.
After identifying the clusters, I reviewed the clusters with the participant, identifying which represented good experiences, and which represented bad ones. Then I asked the participant to talk about each cluster and what they specifically liked or disliked, teasing out which variables were important to a keypad's user experience.
So, what were some of my findings?
Participants preferred a lower pitch click sound. This is because participants likened a lower pitch click to an impression of better quality, versus a more 'tinny', loud, high-pitched sound. This was most likely due to their impression that heavier or thicker materials make a deeper clicking sound than thinner or lighter materials. One participant likened a deeper, lower pitched sound to the sound that is made when you shut the door of a well-engineered car. Participants also preferred keys with a more pronounced snap/break over. Participants felt that keys with a more pronounced snap/break over felt faster, easier to use, and less fatiguing. The keys that did not have a more pronounced snap/break over were often described as squishy or dead feeling.
After going through a number of evaluations using this methodology, we landed on an optimized keypad user experience, that resonated well with end-users.
​
