The effect of varying the number of contributors in the prosecution and alternate propositions.

ABSTRACT

Using a simplified model, we examine the effect of varying the number of contributors in the prosecution and alternate propositions for a number of simulated examples. We compare the Slooten and Caliebe [1] solution, with several existing practices. Our own experience is that most laboratories, and ourselves, assign the number of contributors, N = n, by allele count and a manual examination of peak heights. The LRn for one or a very few values is calculated and typically one of these is presented, usually the most conservative. This gives an acceptable approximation. Reassessing the number of contributors if LR = 0 and adding one to N under both Hp and Ha to "fit" the POI may lead to a substantial overstatement of the LR. A more reasonable option is to allow optimisation of the assignment under Hp and Ha separately. We show that an additional contributor explained the single locus profile better when PHR≥0.51. This is pleasingly in line with current interpretation approaches. Collectively these trials, and the solid theoretical development, suggest that the Slooten and Caliebe approach preforms well.