Likelihood ratio and logical thinking

Question 1

Homo oeconomicus v. homo heuristicus

Answer 1

Homo oeconomicus:

1. Acts rationally
2. Optimal weighing of all facts

Homo heuristicus:

1. Effective and efficient results by ignoring irrelevant information
2. Good in uncertainty

Question 2

Bounded rationality

Answer 2

Embrace the benefits of satisfactory results and avoid excessive demands of optimal ones

Question 3

Ecological rationality

Answer 3

Match between a strategy, the environment of application and the abilities and skills of the researcher

Question 4

Fast and frugal trees

Answer 4

Each node has only 2 valid nodes of which at least one is an exit  Example of heuristics

Question 5

Dual process theory

Answer 5

System 1  Intuitive, automatic, fast, frugal and effortless

System 2  Analytical, slower, deliberate, costly and effortful

Question 6

Person v. system approach

Answer 6

Person approach  Blame the person, errors are due to moral issues

System approach  Blame the system, accepts human errors as intrinsic and tries to establish systematic barriers and safeguards to human errors

Question 7

Swiss cheese model of errors

Answer 7

1. Different slices show different barriers/safeguards
2. Errors require:
   - Active failure (unsafe acts committed by the person who is in direct contact with the patient or system) and
   - Latent conditions (inevitable resident pathogens of the system)
3. Once these two are aligned, there will be an error

Question 8

Prosecutor’s fallacy

Answer 8

Fallacy of the transposed conditional, incorrect reversal of a conditional proposition  Because something is likely does not mean that it is

Question 9

Likelihood ratio

Answer 9

1. Also called diagnostic value or positive diagnostic value
2. Likelihood of a hypothesis under two mutually exclusive scenarios (A = by suspect and B = by random person)
3. = Sensitivity/1-speciticity
4. = Probability of the evidence supporting the prosecutor’s hypothesis / Probability of the evidence supporting the defence’s hypothesis

Question 10

Expression of LR

Answer 10

1. = Finding the evidence is … times more likely if the suspect left it than if a random member of the population did
2. = The evidence provides strong support for the theory that the suspect is the donor of the evidence

Question 11

Specificity and sensitivity

Answer 11

Specificity = Probability of correct positives the test produces

Sensitivity = Probability of correct negatives the test produces

Question 12

Odds

Answer 12

Pre-test odds = Pre-test probability/1 – pre-test probability

Post-test odds = Prior probability x likelihood ratio = likelihood ratio / (likelihood ratio + odds of alternative hypothesis)

If odds are in percentage, first multiply the percentage by LR, and then divide that number by that same number + alternative scenario percentage

Question 13

Logical approach to the interpretation of forensic evidence

Answer 13

Uses the concept of likelihood ratio by considering the probability of the evidence under two competing hypothesis (the hypothesis of the defence and that of the police)

Question 14

P-value

Answer 14

Probability of obtaining the study results (that are equal to or more extreme than the current ones) if the hypothesis (usually null) is true  Does not express the probability of the hypothesis given the evidence (= prosecutor’s fallacy)