Forensic likelihood ratio for two competing hypotheses

Given the log-likelihood of the evidence under a natural/target hypothesis (loglik_h1) and an alternative (loglik_h2), returns the likelihood ratio \(LR = P(E \mid H_1) / P(E \mid H_2)\) with its base-10 logarithm and the ENFSI (2015) verbal equivalent. Computed in log space for numerical stability. This is the standard way to state forensic evidence: the LR reports how much the evidence favours \(H_1\) over \(H_2\); it is not the posterior odds and does not "prove" either hypothesis.

Usage

morie_taphonomy_likelihood_ratio(loglik_h1, loglik_h2)

Arguments

loglik_h1, loglik_h2

Log-likelihoods of the evidence under \(H_1\) and \(H_2\) (e.g. from morie_taphonomy_evidence_loglik).

Value

A named list (double where numeric): lr, log10_lr, log_lr, verbal, and interpretation.

References

ENFSI (2015). Guideline for Evaluative Reporting in Forensic Science. European Network of Forensic Science Institutes. Aitken CGG, Taroni F (2004). Statistics and the Evaluation of Evidence for Forensic Scientists (2nd ed.). Wiley.

Examples

morie_taphonomy_likelihood_ratio(loglik_h1 = -3.1, loglik_h2 = -12.7)$verbal
#> [1] "very strong support for H1"