Every day, people in many diverse occupations make decisions that have potentially large consequences for human health and safety. Is this plane safe to fly? Given this patient profile, should a physician operate or recommend more conservative therapy? Will this weather pattern develop into a tornado? The answers to these questions have high stakes, and the process of arriving at a decision is often highly probabilistic, involving analysis of ambiguous and often conflicting information.
Researchers in the area of signal detection theory study how people, animals, and machines distinguish meaningful or important information from the background "noise" in their environment. From these studies, behavioral scientists are developing methods to help people make better decisions.
The results are known as decision aids, often computer programs that rely on a systematic, standard method that establishes a decision threshold and maximize accuracy. A decision threshold is the level of evidence deemed necessary to make a decision in the specific situation, e.g. at what probability of a malignancy does one diagnose breast cancer? A fine balance must be struck between a lax threshold which creates many false positives (and their attendant emotional stress and unnecessary surgery), and a strict threshold which misses some positive cases and thereby jeopardizes lives. The decision maker must weigh whether false alarms or undiagnosed conditions are most costly and then must adjust the threshold accordingly. Accuracy can be improved by enhancing the quality of evidence available through basic research and developing better diagnostic tests and instruments. Thus, the methods of behavioral science go hand in hand with the physical and biological sciences.
The practical applications of decision aids are numerous and extensive. Already used experimentally in breast cancer diagnosis, HIV testing, weather forecasting, prostate cancer staging, and testing airplane wings for metal fatigue, the technique promises improved accuracy, and improved public health and safety.