Credibility theory
\ˌkrɛdəˈbɪlɪti\ \ˈθɪri\
A statistical theory used to combine multiple estimates from different sources when predicting future events.
The different estimates are assigned a credibility weighting based on the size and variance of their underlying data sets. They are then combined into one estimate of the future event with greater weight placed on the estimates with higher credibility.
With respect to longevity, this approach is often used to create a baseline longevity assumption for a specific population (for example a pension plan) by combining results from a rating factor model and an experience analysis of the population (for example the plan’s retiree population). Smaller pension plans will have less experience data to rely on, so tend to have a lower credibility assigned to their experience analysis compared to larger pension plans.