This is a project for the Action and Cognition: Computational Cognition Course.
Our focus is to explore the mouse-tracking method which provides an accessible experimental technique to study conflict in decision making.
We used data published by Caruso et al, at https://data.mendeley.com/datasets/6gphjrkjtd/1 from a Promise Decision task, developed to explore promise keeping and promise breaking behaviour in humans.
Key questions:
- Do trustees keep their promise more often when they make a binding promise i.e. ‘always’ as against an ambiguous promise i.e. ‘sometimes’ ?
- In which situation does a trustee undergo higher cognitive conflict- when keeping a promise made or when breaking a promise?
- Is a trustee’s decision to keep or break the promise reflected in the mouse trajectory of their response?
(We have defined how we determine ‘keep’ or ‘break’ a promise in the notebook)
Overview of the project workflow:
a. Collated the data, cleaned the data(removed obvious errors) and transformed relevant fields to ensure we can analyse features we wanted. A definition of key data-fields is in the notebook.
b. Performed exploratory analysis to understand the main behavioral patterns
c. Calculated specific metrics to quantify the mouse trajectories and explored patterns in mouse trajectories
d. Ran statistical tests to validate key hypothesis
e. Explored methodological issues concerning mouse-tracking
-Ali and Saurabh