# Perceptual Responses to Weak Stimuli by 53401g

VIEWS: 7 PAGES: 45

• pg 1
```									Perceptual Responses to Weak
Stimuli
Detection Thresholds
Signal Detection Theory

Responses to Weak Stimuli   1
Objectives: Be able to
• Discuss the threshold, its shape and
assumptions.
• Discuss assumptions, what’s
measured, and errors for thresholds.
• Discuss assumptions, what’s
measured, and errors for signal
detection measurement.
• Discuss how criterion affects
sensitivity and response bias
• Describe practical examples of the use
of thresholds and signal detection.      2
Responses to Weak Stimuli
Detection Threshold
• If the stimulus is too weak, it is not
detected
• Stimulus grows stronger, still not
detected
• When stimulus reaches threshold
strength, it is detected
• At all levels above threshold strength it
is detected

Responses to Weak Stimuli     3
Threshold – Single Measurement

Responses to Weak Stimuli   4
Responses to Weak Stimuli   5
Problems with Detection Thresholds

• Measured threshold can vary from trial
to trial
• Variation in threshold fits a normal
curve
• Measured threshold functions look like
an S
• Threshold function is a cumulated
normal curve
• Not a squared off function
Responses to Weak Stimuli   6
Multiple Measurements of a Threshold

Responses to Weak Stimuli   7
Discussion Question 1

• If the threshold concept is correct, why
does the “percent of seeing” curve in
threshold experiments produce an S-
shaped curve rather than a squared off,
“threshold shape” function? Explain.

Responses to Weak Stimuli   8
Threshold Assumptions
• A. Sensitivity Varies - thus threshold
varies
• B. No Response Bias - else threshold
may be lower or higher than measured

Responses to Weak Stimuli   9
What’s Measured with Thresholds?

• A stimulus intensity: The stimulus
intensity needed for the stimulus to be
detected half the time.

Responses to Weak Stimuli   10
Three Kinds of Threshold
Measures

• Method of Limits
• Staircase Method
• Method of Constant Stimuli

Responses to Weak Stimuli   11
Method of Limits
• Present stimuli in an ascending order
until observer detects it.
• Present stimuli in a descending order
until observer fails to detect it.

Responses to Weak Stimuli   12
Problems with Method of Limits
• The observer knows whether the next
stimulus will be larger or smaller than
the previous one.
• Response Bias Errors: Anticipation
and Habituation.

Responses to Weak Stimuli   13
Anticipation and Habituation
• Anticipation - response changes before
percept does – jumps the gun
• Habituation - percept changes,
response does not – fail to respond to
change
• These errors eliminated in Method of
constant stimuli
• Stimulus order randomized

Responses to Weak Stimuli   14
Staircase Method
• Begins like method of limits: ascending order
until detected
• From there, tracks the threshold - Quick
Threshold
• Stimulus detected, next stimulus weaker
• Stimulus not detected, next stimulus
Stronger
• Are anticipation and habituation possible?

Responses to Weak Stimuli    15
Responses to Weak Stimuli   16
Method of Constant Stimuli

• Stimulus order randomized - no
anticipation or habituation possible
• Other kinds of response bias possible
• Takes many trials - slow, but accurate
measure

Responses to Weak Stimuli   17
Discussion Question 2
• Compare and contrast the detection
threshold measures: Method of Limits
(MOL), Method of Constant Stimuli
(MCS), and Staircase Method (Stair).
What are the assumptions, what is
measured, and what kinds of errors of
measurement are made using each
method?
• What does compare and contrast
mean?

Responses to Weak Stimuli   18
Discussion Question 3
• Give an example of a practical
application in which threshold
measures would be useful. Be sure
to describe what is measured, give
an example of a practical
application, and show how the
application fits what is measured.

Responses to Weak Stimuli   19
Q3
• What is measured?
– The stimulus intensity necessary for the
stimulus to be detected half the time.
Example.
Fit.

Responses to Weak Stimuli     20
These Two Thresholds Differ:
Is Laurie more sensitive or more willing to say yes?

Responses to Weak Stimuli    21
• Staircase thresholds can be measured
quickly, but there can be problems.
• Because differences in thresholds
could mean either differences in
sensitivity or differences in response or
decision making processes, we can’t
be sure what differences in thresholds
mean.
• When motivation cannot be controlled,
we need another method.

Responses to Weak Stimuli   22
Signal Detection Theory (SDT)
• Assumes that variability in thresholds
is due to variable sensory noise, not
variable sensitivity.
• Assumes that response bias cannot be
eliminated, as required by threshold
measures.
• SDT attempts to measure response
bias and sensitivity

Responses to Weak Stimuli   23
Sources of Sensation

• Sensory Noise - normally distributed.
• Sensory Noise Plus Stimulus - Noise
plus a fixed amount of sensation added
by the stimulus.
• Unlike Thresholds, where the stimulus
is always presented, the observer does
not know whether a stimulus was
presented.

Responses to Weak Stimuli   24
Responses to Weak Stimuli   25
Responses to Weak Stimuli   26
Responses to Weak Stimuli   27
Sensitivity
• Sensitivity is the distance between the
mean of the noise only distribution and
the mean of the noise plus stimulus
distribution in standard deviation units.
• Sensitivity is assumed to be fixed,
unchanging.
• Sensitivity = d’ or d prime.

Responses to Weak Stimuli   28
Response Bias
• Response bias (for thresholds also) is
the willingness to report the presence
or absence of the stimulus, unrelated to
sensation.
• Can be influenced by motivational
aspects of the situation as well as the
probabilities of a stimulus being
present.

Responses to Weak Stimuli   29
Response Bias
• Response bias sets the criterion to
decide whether there was a stimulus
• If the sensation level exceeds her
criterion, the observer says, “I saw it.”
• If the sensation level does not exceed
her criterion, the observer says, “I did
not see it.”

Responses to Weak Stimuli    30
Responses to Weak Stimuli   31
Responses to Weak Stimuli   32
Responses to Weak Stimuli   33
Responses to Weak Stimuli   34
Assumptions of Signal Detection
• Sensitivity is fixed - thus it can be
measured
• There is internal sensory noise that
mimics the effect of the stimulus
• This noise causes variation in the level
of sensation
• There is response bias and it can be
measured

Responses to Weak Stimuli    35
What is Measured by SDT

•   Sensitivity
•   Response Bias
•   These are observer qualities
•   Thresholds measure stimulus intensity
levels

Responses to Weak Stimuli   36
Responses to Weak Stimuli   37
Responses to Weak Stimuli   38
Responses to Weak Stimuli   39
Discussion Question 4
• Compare and contrast the constant
stimuli method (MCS) of detection
threshold measurement with signal
detection (SDT) measurement. What
are the assumptions, what is
measured, and what kinds of errors
of measurement are made using
each method?

Responses to Weak Stimuli   40
Discussion Question 5

• Discuss the effect of changing one’s
criterion from lax to strict on both
sensitivity and response bias. Explain
how those results occur. Drawing a
diagram may be helpful.

Responses to Weak Stimuli   41
Responses to Weak Stimuli   42
Discussion Question 6
• Give an example of a practical
application in which signal detection
measures would be useful. Be sure to
describe what is measured, give an
example of a practical application, and
show how the application fits what is
measured.

Responses to Weak Stimuli   43
Q6
• What is measured?
– Sensitivity, d’
– Response bias, β
• Example.
• Fit.

Responses to Weak Stimuli   44
Objectives: Be able to
• Discuss the threshold, its shape and
assumptions.
• Discuss assumptions, what’s
measured, and errors for thresholds.
• Discuss assumptions, what’s
measured, and errors for signal
detection measurement.
• Discuss how criterion affects
sensitivity and response bias
• Describe practical examples of the use
of thresholds and signal detection.       45
Responses to Weak Stimuli

```
To top