refererenceklassforudsigelse

Refererenceklassforudsigelse, which translates from Danish to "reference class prediction," is a term used primarily in the field of forensic psychology and criminology. It refers to the practice of using statistical methods and historical data to predict the likelihood of an individual belonging to a specific group or "reference class" with certain characteristics. This is often applied in risk assessment, particularly concerning recidivism or reoffending.

The core idea behind reference class forecasting, a broader concept, is to improve prediction accuracy by considering the performance of similar past cases. Instead of relying solely on subjective judgment or case-specific factors, reference class prediction anchors predictions in empirical data from a relevant comparison group. For instance, when predicting the likelihood of a particular offender reoffending, a reference class might consist of all individuals with similar offense types, demographic profiles, and prior criminal histories who have been released from custody.

The process typically involves identifying a suitable reference class and then calculating the base rate of the outcome of interest (e.g., reoffending) within that class. This base rate serves as a starting point for the prediction. Adjustments can then be made based on individual-specific information, but the reference class prediction provides a crucial statistical anchor, aiming to mitigate common biases such as overconfidence or the neglect of base rates. While it is a statistical tool, its application requires careful consideration of the definition and relevance of the chosen reference class.