Signal Detection Theory (SDT)

SDT Values

Description

P(FA)
P("NS"|N)

0.106
[Empirical Value]

Probability of False Alarm outcome.
Probability of saying "Signal" when there was no signal.

P(CD)
P("NS"|NS)

0.690
[Empirical Value]

Probability of Correct Detection Trial.
Probability of saying "Signal" when the signal did indeed occur.

z(FA)

1.250
[z-table lookup]

Number of z-score units from the modal response of the "Noise Only" distribution to the response criterion.

z(CD)

-0.500
[z-table lookup]

Number of z-score units from the modal response of the "Noise + Signal" distribution to the response criterion.

f(Z_FA)

0.183
[z-table lookup]

Height of the "Noise Only" distribution at the location of the response criterion. Given that a particular strength of evidence occurred, what is the relative likelihood it was produced by "Noise Only" condition.

f(Z_CD)

0.32
[z-table lookup]

Height of the "Noise + Signal" distribution at the response criterion. Given that a particular strength of evidence occurred, what is the relative likelihood it was produced by the "Noise + Signal" condition.

b

1.928
[Calculate]

Response Bias.
Minimum ratio of likelihoods to respond by saying "Signal"

d'

1.750
[Calculate]

Sensitivity. d' = z(FA) - z(CD) = 1.250 - (-0.500) = 1.750
How distinct the two distributions are for this subject.

P(Miss)
P("N"|NS)

0.310
[Calculate]

1- P("NS"|N) =1.000 - 0.690 = 0.310

P(Correct Rejection)
P("N"|N)

0.894
[Calculate]

1- P("N"|N) = 1.000 - 0.106 = 0.894