I. Lecture 2 – Measuring Perception with Psychophysics (April 4, 2017)
1. Detection, Discrimination, and Signal Detection
a. Detection: Did you hear anything?
b. Discrimination: Are two sounds different from each other?
c. Signal Detection: How do you decide if you’ve heard your phone ring?
2. Detection (recall from Lecture 1)
a. How much sensory stimulation is needed in order for our senses to detect it?
b. What is the softest sound, the lightest touch, the faintest light, smell, or taste? (Recall absolute
a. How small of a difference between two stimuli can be detected?
b. Is something heavier, brighter, or more painful than another?
4. Signal Detection Theory
a. When do we decide something has happened?
b. Noise is anything else in the environment that is not the signal you’re looking for
B. Detection (recall from Lecture 1)
1. What is the faintest light you can see?
2. Find absolute threshold for perception by increasing or
decreasing intensity and ask subjects whether or not they
perceive a stimulus
a. The probability of detection is gradual
b. “Absolute” threshold is point where detection occurs
50% of the time
c. Graphs: absolute threshold – detection 50% on curve,
trace down to x-axis
1. Has anything changed?
a. What is the smallest detectable difference between two stimuli? Or the minimum change in a stimulus
that can be correctly judged as being different from another?
b. Just Noticeable Difference (JND) – the ability to detect a difference between two stimuli
2. Examples for weight:
a. Which would be detected as a change in weight?
i. Watermelon would exceed the JDN
ii. Possibly the grapefruit, but not the chocolate or feathers
b. Which would be detected as a change in weight now?
i. Grapefruit is heavy in respect to the chocolate box
ii. *JND is relative to context to original stimulus intensity
3. A change in weight must be 2.5% of beginning weight to be noticed as different
a. A proportion of the original weight
b. 20lb bag JDN = 0.5lbs; 2lb chocolate box JDN = 0.05lbs (0.8oz)
4. Other examples:
a. Image holding two envelopes. One has one quarter in it. How many in the other before you can tell
that the two envelopes have different weights?
i. Adding another quarter (doubles weight)
b. BUT, now imagine that those quarters are in your hiking boots. Now, about how many?
i. Maybe need like a roll of quarters?
c. Imagine buying a phone that costs $75. Would you pay an extra $20 for a case? What if the phone cost
$800? d. Imagine speaking to someone in the library vs a rock concert – how much louder do you have to be in
the concert? (Think of the loudness of your voice as a fraction of the background noise.)
i. It’s relative; a psychological property of what we perceive
5. Weber Fraction – the smallest change in a stimulus that can be detected is a constant proportion of the
a. ΔI (JND) = K (proportion) * I (stimulus)
b. K depends on the stimulus (different proportion values for different stimuli)
i. Weight 1:40
ii. Length 1:100
c. Fechner thought that ΔI = “unit of the mind” that has an orderly relationship to a unit of physical
6. Weber’s Law – the greater the magnitude of the stimulus, the bigger the differ you need to detect a change
a. JND change detection needs to increase the absolute physical stimulus intensity
i. Physical measurement of stimulus increases much more than noticeable change unit
ii. JND change is constant; perceived as equal change, but is not the same increase in physical
b. Not all stimuli have this same curve
c. Consequence: If the radio is on softly, you only have to turn it up a little to make it twice as loud. If it
on so loud that your neighbor can hear it – how much would you have to turn it up to be twice as loud?
7. Caveat to Weber’s Law
a. The graph illustrates that the “perceived” magnitude of a noticed change is not always experienced as
b. Each sense is adapted to how the organism functions in the environment.
8. But wait…how does the context affect our thresholds?
a. Most real world signals are embedded in noise (i.e. are not always very easy to detect) and depend on
the individual’s bias
i. Bias – our state of mind of how important something is to an individual
b. Need to know how well we can detect a stimulus amid noise, given a certain tendency to say
something is there or not.
D. Signal Detection Theory (SDT)