I'm excited to share a new preprint: In a novel experimental paradigm, we found evidence that stress induces a computational bias during alcohol-related decision-making in favor of alcohol, but this bias was only sometimes strong enough to overcome competing considerations (e.g., taste preferences).