PSY
429 Experimental Psychology
Class Exercise #4
Psychophysics
Answers
1. Use the method of limits to determine the absolute threshold for a visual stimulus. The data from one subject are provided below.
Light Intensity 1 2 3 4 5 6 10 YES 9 YES YES YES 8 YES YES YES YES 7 YES NO YES YES YES YES 6 NO NO YES NO YES NO 5 NO NO NO NO NO 4 NO NO NO 3 NO NO NO 2 NO NO 1 NO
6.5 7.5 5.5 6.5 5.5 6.5
absolute threshold = 38/6 = 6.33The point at which the observer can detect the light is 6.33 units.
2. Use the method of limits to measure an observer's ability to detect sweetness. The data from one subject are provided below.
Arbitrary Units of
Sweetness 1 2 3 4 5 6 1 Not Sweet NO NO NO 2 NO NO NO NO NO NO 3 NO YES NO YES YES YES 4 YES YES YES YES YES 5 Very Sweet YES YES YES
3.5 2.5 3.5 2.5 2.5 2.5
absolute threshold = 17/6 = 2.83The point at which the observer can detect the presence of sugar is 2.83 units.
3. Suppose you manufacture candy bars and you want to produce a new version of a popular favorite that is noticeably larger, but you don't want it to be too big because of the expense. Use the method of limits to determine what size is perceived to be larger than the original bar. Suppose the original bar of candy is 6 ounces. For the data below, L stands for "larger" and S stands for "smaller" and E stands for "equal." [Assume that the six trials below are representative of the data from 1000 trials.]
Ounces 1 2 3 4 5 6 9.0 L L 8.5 L L L 8.0 L L L 7.5 E E L 7.0 E L E L L 6.5 E E E E E 6.0 S E S L E E 5.5 E E S E 5.0 E E S 4.5 S E S 4.0 S S S 3.5 S S S 3.0 S S S 2.5 S S S 2.0 S S 1.5 S 1.0 S
upper
threshold 7.75 6.75 7.25 5.75 6.75 6.75
lower
threshold 6.25 4.75 5.75 4.25 5.75 5.25
mean upper threshold = 41/6 = 6.83mean lower threshold = 32/6 = 5.33
interval of uncertainity = 6.83 - 5.33 = 1.5
difference threshold = 1.5/2 = .75
Thus, approximately .75 oz of a difference between the new version and the old version is necessary before a difference will be noticed.
4. Draw the signal + noise (signal present) distribution and the noise (signal absent) distribution under the circumstances in which the observer has occasional difficulty detecting the presence of the signal.
5. Draw the signal + noise (signal present) distribution and the noise (signal absent) distribution under the circumstances in which the observer has the ability to detect the presence of the signal 100% of the time.
6. Draw the signal + noise (signal present) distribution and place the decision criterion to reflect a HIT rate of
a. 0.95b. 0.75
c. 0.50
d. 0.25
7. Draw the noise (signal absent) distribution and place the decision criterion to reflect a FALSE ALARM rate of
a. 0.25b. 0.50
c. 0.75
d. 0.95
8. Define a "liberal" decision criterion.
more "yes" than "no" answers resulting in more HITS and FALSE ALARMS
9. Define a "conservative" decision criterion.
more "no" than "yes" answers resulting in fewer HITS and FALSE ALARMS
10. Draw an ROC (receiver operating characteristic) function and label the two axes, identify d' and identify the part of the graph reflecting a liberal decision criterion and the part reflecting a conservative decision criterion.
11. What factors or conditions can affect the decision criterion, whether it will be liberal or conservative?
1. payoff: assessing costs versus benefits-higher costs of saying "yes, the signal is present" will tend to result in fewer HITS and the adoption of a conservative decision criterion2. probability of signal
-knowledge that the signal will occur more often than not will tend to result in more HITS but also in more FALSE ALARMS3. discriminability of signal
-if the signal is easily distinguished from the environmental noise, that will tend to result in more HITS
12.What does d' reflect?
The ability of the subject to deterct the presence of the signal.
13. What does B (beta) reflect?
The decision criterion
14. Typical results are presented in the table below of a subject's response pattern when told that the stimulus (signal) is presented 90% of the time. Calculate d'.
Present 0.95 0.05 Absent 0.63 0.37
HITS = .95FALSE ALARMS = .63
d' = zn - zs+n
d' = -.33 - (-1.65)
d' = 1.32
15. Typical results are presented in the table below of a subject's response pattern when told that the stimulus (signal) is presented 10% of the time. Calculate d'.
Present 0.35 0.65 Absent 0.04 0.96
HITS = .35FALSE ALARMS = .04
d' = zn - zs+n
d' = 1.75 - (.39)
d' = 1.36
To go back to Dr. Ryan's Home Page, click on kitty.