Predicting individual-level income from Facebook profiles

Published in: PLoS ONE - March 2019

Written by

Sandra Matz, Jochen Menges, David Stillwell, Andrew Schwarzt

Summary

What we found: In a study with 2,623 U.S. Americans, we found that digital footprints such as Facebook likes or status updates can be used to predict people’s income. Compared to using only socio-demographical information such as gender, adding digital footprints made the prediction of income even more accurate.

Why it matters: Our findings show that digital footprints are a powerful tool to tailor products and services more effectively. At the same time, using digital footprints in research and business raises questions about the digital ethics and moral principles governing these decisions.

What next: Digital footprints can be used to increase the fairness of many organizational processes such as recruiting. But organizations also need to establish clear guidelines on how and when to use digital footprints in an ethical way. This could help organizations in making their offerings not only more user-friendly but also fairer.

Previous
Previous

Changing patient mindsets about non–life-threatening symptoms during oral immunotherapy: a randomized clinical trial

Next
Next

Providers’ demeanor impacts patient perceptions of visit length