Regression analysis of cursor movement features and personality traits
In this rapidly digitalizing world, the need to understand users better in order to best cater to their demand is also growing. A cost-effective way to conduct user experience research is to examine user’s mouse movement data. This research examines the approach to infer personality from mouse trajectories. Regression and correlation analyses were performed to examine the relationships between cursor movement features, such as pauses, reconsideration, and cursor speed, and Big Five personality traits. Significant correlations and regression coefficients were found in the analyses. These results validate the feasibility of using mouse movement data to predict personality. Findings in this research can be extended to further evaluate the relationship between cursor features and personality traits. The method presented here can also be applied to detect other internal states, such as inattentiveness and emotional state, in related future work.
The data used in this research came from two original studies from the Environmental Neuroscience Lab of the University of Chicago. The first study is a preference-based image rating task. In each trial, participants were asked to look at 12 photos of streets taken at different angles, and pick four images that she or he liked or disliked the most. The second study comes from a similar street psychology project, which also asks participants to select 4 out of 12 street view pictures. The only difference is the selection criteria used. In the second study, participants were asked to select four pictures that include the highest level of a given feature, such as perceived walkability, perceived level of human activity, perceived order, or perceived visual diversity, etc. A screenshot of our instruction page is shown below:
All participants for these two studies were recruited from Amazon Mechanical Turk. For privacy purposes, participants worker ID have been changed/removed. Some of the exact steps taken in data wrangling are not included in the google colab notebook for the same reason.