How Reliable Are EEG Metrics in Ad Testing? New Research Sheds Light

By Carla Nagel

Electroencephalography (EEG) has become an increasingly popular tool for assessing the effectiveness of television commercials. But how reliable are the metrics commonly used in neuromarketing? A new study by Van Diepen, Boksem, and Smidts (2024) in the Journal of Advertising provides crucial insights into the reliability of EEG measures and what this means for ad research.

The researchers examined six EEG metrics—alpha, beta, gamma, theta, alpha-asymmetry, and intersubject correlation (ISC)—in a large-scale study involving 116 participants watching 13 TV commercials. Their goal? To determine the consistency of these metrics across different participant samples and viewing conditions.

Key Findings:

ISC is the gold standard. It demonstrated the highest reliability, confirming that synchronized neural responses across viewers can serve as a strong predictor of ad engagement.
Traditional EEG power metrics (alpha, beta) showed moderate reliability, but require careful sample size considerations.
Alpha-asymmetry, a widely used measure of approach motivation, was found to be highly unreliable. This raises questions about its validity in neuromarketing studies.
Larger sample sizes improve reliability. At least 30–40 participants are needed for stable results in most EEG metrics, with ISC requiring the smallest sample.
Repeated exposure enhances measurement accuracy. Watching an ad multiple times significantly improves EEG signal consistency.

Why This Matters

For neuromarketers and advertisers relying on EEG data, choosing the right metric is critical. This study provides much-needed clarity on which EEG signals offer robust insights and which ones might lead to misleading conclusions. It also reinforces the importance of proper study design, ensuring that sample size and experimental setup support meaningful, reproducible findings.

Curious to learn more? Read the full study here: Van Diepen et al. (2024) - Reliability of EEG Metrics for Assessing Video Advertisements