Statistical Threshold

George Carmody

The KADAP spent time considering and discussing the population makeup of the World Trade Center victims, including the population substructure that might influence the posterior probability of an identification. In the end, our consensus was to be conservative in all calculations to guard against false identifications.

In the WTC identification effort, the statistical threshold (i.e., posterior probability) that was established before a DNA-based identification could be made was high: 99.9%. To attain this high level of confidence, the DNA laboratory targeted a likelihood ratio that, combined with the prior probability, met the statistical threshold. This likelihood ratio (L) was based on the following
formula:

DNA Laboratory's Likelihood Ratio: 1 - (1-1/L) raised to the Nth power is less than or equal to p

In this formula, N = the estimated number of RMs (5,000 in the WTC); and p = the defined acceptable chance of error (1/1,000,000). Solving for the likelihood ratio:

Solving for the likelihood ratio: L is greater than or equal to 10,000,000,000