Modeling and Comparing Brain Processes in Message and Earned Source Credibility Evaluation.

Area of Science:

• Cognitive Neuroscience
• Social Psychology
• Information Science

Background:

• Assessing information credibility is crucial in the digital age, particularly with the rise of 'fake news'.
• Source credibility is a key factor influencing human judgments of trustworthiness.
• Measuring brain activity offers a direct window into the cognitive processes underlying credibility evaluation.

Purpose of the Study:

• To investigate the neural basis of human source credibility evaluation.
• To develop predictive models of credibility judgments using electroencephalography (EEG) data.
• To compare models of source credibility evaluation with message credibility evaluation.

Main Methods:

• An experiment involving participants evaluating the credibility of fictitious student sources.
• Utilizing electroencephalography (EEG) to record brain activity during these evaluations.
• Developing machine learning models to predict credibility judgments from EEG signals, employing 10-fold cross-validation.

Main Results:

• Identified specific brain areas active during positive and negative source credibility evaluations.
• Achieved an F1 score exceeding 0.7 in predicting human source credibility evaluations using EEG data.
• Successfully modeled and predicted message credibility evaluations with perfect knowledge, enabling comparison with source credibility models.

Conclusions:

• EEG-based brain activity measurements can effectively predict human source credibility evaluations.
• The study provides insights into the neural mechanisms differentiating positive and negative credibility judgments.
• This research contributes to understanding cognitive processes in navigating information credibility.