Lie Detection: Testing Truth or Falsehood?

  • Thread starter Thread starter Mrencko
  • Start date Start date
  • Tags Tags
    Detection Testing
Click For Summary
SUMMARY

The forum discussion revolves around the application of lie detection in a legal context, specifically analyzing the probabilities associated with a lie detector's readings. The lie detector has a 10% false positive rate for truthful statements and a 95% true positive rate for lies. Users debated the interpretation of the problem, focusing on calculating conditional probabilities, particularly the probability of lying given a positive reading from the detector. The discussion highlights the necessity of understanding Bayesian probability to fully grasp the implications of the lie detector's accuracy.

PREREQUISITES
  • Understanding of conditional probability
  • Familiarity with Bayesian statistics
  • Knowledge of probability theory terminology
  • Experience with interpreting statistical data
NEXT STEPS
  • Study Bayesian inference and its applications in real-world scenarios
  • Learn about the mathematical foundations of conditional probability
  • Explore the implications of false positives and false negatives in statistical testing
  • Research the reliability and validity of lie detection technology
USEFUL FOR

Statisticians, legal professionals, psychologists, and anyone interested in the accuracy and interpretation of lie detection methods.

Mrencko
Messages
109
Reaction score
0

Homework Statement



A crime has been committed and the police have only captured a suspect, although he obviously claims to be innocent. To give a verdict the judge authorizes the use of the lie detector. The manufacturer of the device warns the authority that the lie detector in 10% of the cases in which it has been used showed a positive reading (that is to say lie), since in fact the suspect spoke with the truth. So too, the detector in 95% of the cases, has given a positive reading when the suspects really lied.
A) tell a lie the detector shows a positive reading
B) do not tell a lie and the detector shows positive reading
C) tell a lie and the detector shows negative reading
D) do not lie and the detector shows a negative reading

Homework Equations

The Attempt at a Solution


well this might be too simple or to complex i don't know, but i need to know if i am right in my responses because my career is in play.
a)95%
b)10%
c)5%
d)90%
i use the complement of probabilities to estimate the ungiven probabilities if i am wrong please help me
pr(F/T)=0.10 pr(Fc/T)=0.90
pr(F/L)=0.05 pr(Fc/L)=0.95
 
Physics news on Phys.org
Is that the exact problem statement in English? It has a strange phrasing.
 
Yes it is, in fact very weird problem
 
(A) to (D) are not even complete sentences, and there is no actual problem statement.

I guess your interpretation is reasonable, but it is really guesswork. With the given numbers you cannot calculate more than what you did.
 
Because it is a translation it is hard to be sure, but in English the questions read as though they refer to joint probabilities (lie AND positive). That would require some a priori probability of a lie, which we do not have, but if we were to have such the whole question setting would make more sense. It would lead into asking the probability of guilt given the result of the lie detector. Very Bayesian.
 
i can swear that the problem is like that, i think the problems are designed by the profesor, can't find it in the internet, so its very weird
just gives two data as you can see above and 4 questions i can't put the questions in this way
a) probability of lie given the detector shows positive reading i think this mean pr(L/Fc)
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
3K
Replies
1
Views
2K
  • · Replies 25 ·
Replies
25
Views
5K
  • · Replies 15 ·
Replies
15
Views
7K
Replies
9
Views
3K
  • · Replies 61 ·
3
Replies
61
Views
12K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 2 ·
Replies
2
Views
4K
Replies
2
Views
3K
  • · Replies 152 ·
6
Replies
152
Views
11K