Ink dating part II: Interpretation of results in a legal perspective

In the first part of this paper three ageing parameters were selected as promising from the population of 25 ink entries aged during 4 to 304 days: the quantity of phenoxyethanol (PE), the difference between the PE quantities contained in a naturally aged sample and an artificially aged sample (RNORM) and the solvent loss ratio (R%). In the current part, each model was tested using the three selected ageing parameters. Results showed that threshold definition remains a simple model easily applicable in practice, but that the risk of false positive cannot be completely avoided without reducing significantly the feasibility of the ink dating approaches. The trend tests from the literature showed unreliable results and an alternative had to be developed yielding encouraging results. The likelihood ratio calculation introduced a degree of certainty to the ink dating conclusion in comparison to the threshold approach. The proposed model remains quite simple to apply in practice, but should be further developed in order to yield reliable results in practice.