lecture 1 - introduction and measuring perception Flashcards
History of Psychophysics
assumption about mind and body until the mid 1800s
assumption that mind and body are separate from each
other:
- Body = physical: can be measured & studied
- Mind = non-physical/invisible: cannot be measured & studied
History of psychophysics
1850s
1860s
1850 (22nd of October)
Gustav Fechner: body and mind are not independent!
→ mind can be studied by measuring the relationship between changes
in physical stimulation (body) and a person’s experience (mind)
- 1860: Fechner published: “Elements of Psychophysics”: proposes a
number of methods to study the relationship between the mental and
the physical → includes proposal of 3 techniques to measure thresholds
what is psychophysics
the scientific study of the relationship between the physical stimulus and the perceptions evoked by it.
what is a threshold
2 types of thresholds
-measures the limits of the sensory system
absolute thresholds -measure the minimum of a stimulus that can be
perceived by one of our senses (e.g., smallest amount of light energy we can
see, smallest concentration of chemicals we can smell or taste etc…)
Difference thresholds- measure the smallest difference between two stimuli a person can detect
Psychophysics deconstructed
MTM
method - physical stimulus , how do we select and present the stimulus ?
-methods of limits
-methods of adjustment
- method of constant stimuli (classical psychophysics)
Task
what does the participant do?
-magnitude estimation
-detection
-discrimination
-matching
-recognition
measure (behaviour/ experience)
-how do we characterise behaviour?
(dependent variable)
-phenomenological report (what do you see?)
-reaction time
-physical tasks and judgements (lift it, how heavy is it etc)
Psychophysical methods to measure thresholds
-classically used to determine absolute thresholds
1) methods of limits
2) methods of adjustment
3) methods of constant stimuli
what is the methods of limits, how does it measure thresholds
-what is a cross over point
-repeat?
Experimenter presents stimuli in
ascending (intensity increases) or
descending (intensity decreases) order
- Determines “cross-over” points – where participants perception changes (from noticed to unnoticed or vice versa) (receiving stimulus to not receiving)
- Several runs starting equally often
above and below threshold - the absolute Threshold = average of cross-over
points (after all the runs)
what is Method of adjustment
-give an example
-pros/cons
The participant (not the
experimenter!) adjusts the stimulus
intensity
example of this is getting participant to reduce contrast until stimulus is no longer visible
- Relatively fast method as
participants can adjust the stimulus
themselves in a few trials - However, is “frowned” upon (i.e.,
quick and dirty approach)
what is the method of constant stimuli
-repeats ?
-how do you get to the threshold
-the experimenter presents stimuli of different intensity in random order , and the participants indicate their perception
-Stimuli must be chosen so that the
weakest is never detected and the
strongest always detected (typically 5-9 different intensities
- Many repeated presentations –
determine the percentage of correct
detections - Threshold = intensity at which stimulus is detected 50% of all time (look at graph) (the percentage of stimuli detected being 50%)
so etc intensity 180 is detected 50% of the time
method of constant stimuli
-pros/ cons
pros
-more accurate than methods of limits and adjustment
-involves more observations and stimuli are presented in random order (reduces order and expectation effects)
cons
-very time consuming
methods of constant stimulus ,
how can some of the problems related to expectations of subjects and order effects can be solved ?
-forced choice procedures
-adaptive staircase method
explain forced choice procedures (2-AFC and 2-IFC)
-what would the ideal observer see compared to the real human
-forced choice procedure patterns are presented side by side
(or one after the other), participants
task which side (or time interval)
contains the grating (until grating is
to faint to see).
If participants can’t see the grating at all they should obtain about 50%
correct (guessing chance).
Threshold determined at 75% correct.
look at graphs of ideal observer (no noise or uncertainty etc) compared to the real human - probability that a target is detected increases gradually
-decreases the chance of expectation of order and expectation effects just like methods of constant stimuli)
adaptive staircase procedure
up down
-what is it
-what issues does it deal with
-what issues are there with this
Adaptive procedure knows, every time it knows you you hear the sound/received the stimulus it’ll go down and vice versa when you cant hear the sound
Previous example: 11 contrasts – to get reliable threshold
measures many repetitions required (e.g., N=40 per
contrast) = 440 trials
- Present intensities that actually include the threshold
(otherwise pointless)! - Staircase deals with those issues: starts with clearly
perceivable stimulus: if correctly identified decreases the
intensity – until an error occurs → increases intensity
again (fast and very few trials required) - Averaging over the transition or reversal points to
determine threshold - Issues:
Where to start? Which step size? Which procedure?
When to stop? Participant expectation…
psychophysical tasks to determine thresholds
detection
discrimination
matching
detection
discrimination
matching
1.Detection: “Is the stimulus there?”
2-AFC: On which side is the stimulus
presented?
- Discrimination: “Which stimulus is
stronger?”
2-AFC: Which side contains the higher contrast stimulus? - Matching: “Adjust one out of two
stimuli so that they look/sound/feel
the same”
