Propositions (Hp / Hd)
Definition
The two competing hypotheses in evaluative reporting. Hp is the prosecution proposition (for example, the accused is the source of the trace), and Hd is the defence proposition (for example, an unknown person is the source). The likelihood ratio is always calculated with respect to a specific pair of propositions.
- Hp meaning
- Prosecution proposition, e.g. accused is the source
- Hd meaning
- Defence proposition, e.g. unknown person is source
- Function
- Numerator and denominator of the likelihood ratio
- Framework
- Bayesian evidence evaluation
Common questions
How does changing Hd change the likelihood ratio for the same evidence?+
The same DNA match can yield a very different LR depending on whether Hd names a random unrelated person or a close relative of the accused, because the probability of the evidence under Hd depends heavily on how genetically similar the alternative source is assumed to be.
Can a scientist testify to Hp being true rather than reporting the LR?+
No. Good practice restricts the scientist to the likelihood ratio, which weighs the evidence under both propositions; deciding which proposition is actually true is the role of the court, which also weighs evidence outside the forensic report.
Related terms
- Bayes' Theorem
- The mathematical rule that relates prior odds, the likelihood ratio, and posterior odds. In the odds form: O(Hp|E) = O(Hp) x LR....
- Evaluative Reporting
- Assessment of findings against alternative propositions, sometimes expressed using a supported likelihood ratio. An LR compares probabilities of the findings under those...
- Likelihood Ratio (LR)
- A ratio comparing the probability of findings under one stated proposition with their probability under another, conditional on relevant information. Its direction...
- Prosecutor's Fallacy
- The error of treating the RMP (or its reciprocal) as the probability that the defendant is innocent, or as the probability that...
- Random Match Probability (RMP)
- The probability that a randomly chosen unrelated person from the relevant population would match the evidence profile by chance. A very small...