The human touch: Evidence from an eye-tracking experiment on financial advice

This study empirically investigates whether the source of financial advice—human or algorithmic— affects individuals’ evaluation of consumer credit loans, either directly, or indirectly via fixation time on cost-related information. We do so leveraging eye-tracking data gathered during a lab experiment, where participants evaluated the financial convenience of various loans after reviewing an informational document and receiving financial recommendations from either a human or an algorithmic advisor. Results show that the source of advice does not directly influence the accuracy of the financial evaluation. However, human advice exerts an indirect effect by modulating attention: when advice is congruent with the objective financial (in)convenience of the loan, human advisors facilitate more efficient information processing, thereby enhancing evaluation accuracy. Conversely, when advice is incongruent, human advisors elicit longer fixation times, signaling greater cognitive effort without improving performance.