Definition
A False Positive (Type I Error) occurs when a test indicates a presence or an event when it is actually absent. It is an error in binary classification, common in statistics and signal detection theory.
Why It Matters
The cost of being wrong is rarely symmetrical. In medicine, a false positive might lead to an unnecessary biopsy or treatment. Understanding this error allows us to set our thresholds based on the “cost of the false alarm” ensuring our decision-making systems are optimized for the highest-stakes outcomes.
Core Concepts
- False Positive (Alpha Error): A “false alarm.” (e.g., a fire alarm goes off when there is no fire).
- Specificity: The ability of a test to correctly identify those without the condition (minimizing false positives).
- The Trade-off: Increasing sensitivity usually decreases specificity (leading to more false positives). The optimal balance depends on the cost of the error.