COGNITIVE PSYCHOLOGY: PERCEPTION
(LECTURE ONE)
We have no direct contact with the physical world – only the information conveyed
by our sense organs to our brains. All knowledge of the outside world depends on our
senses. We have the illusion of having direct contact with the world.
Perception can be seen as the psychological processes and underlying physiological
mechanisms by which we gain knowledge of the world via our sense organs.
Our perceptions are an accurate judge of reality - we can judge the speed of cars so we
don’t get run over. But we still see illusions.
Perception is an active process of interpreting sensory information to guide our
interactions with the environment.
We understand things from our sight – things which a camera can’t understand.
Seeing appears easy because we have so much specialized neural circuitry devoted to it –
half of the cortex is devoted to sight.
S TUDYING P ERCEPTION
Discoveries – e.g. all of the colours we can see can be duplicated by mixing 3 basic primary
colours in appropriate amounts.
Psychophysics: The study of the quantitative relationship between sensory
experience (psycho) and environmental stimulation (physics)
The central concept is the measuring of thresholds.
Detection threshold: weakest stimulus value needed to evoke a sensation in the observer.
Discrimination: (difference threshold): smallest difference between 2 stimuli along a
particular dimension that can be detected.
Measuring Detection Thresholds psychophysically
Limits:
- Change stimulus strength until subject says just detectable/undetectable
Adjustment:
- As above but subject adjusts stimulus until just detectable/undetectable
Constant Stimuli:
- Present subject with a fixed set of stimulus strengths in a random order and ask them
each time whether or not they can detect it.
Forced Choice task:
- Solves problem of subject being biased or not telling the truth
- Ask subject to report something about the stimulus, e.g. which side
- Calculate percent correct responses and threshold.
- Threshold = stimulus strength producing 75% correct performance (5 units)
Measuring Discrimination Threshold Psychophysically
Ideally use a forced choice task
One stimulus is presented (first) and then a second one.
, The observer has to report which stimulus has the greater strength.
Calculate percent correct responses, plot a psychometric function and estimate the
threshold.
THRESHOLD PERFORMANCE
- Infer properties of perceptual mechanisms from changes in thresholds.
- For the SAME stimulus strength the subject sometimes responds correctly and
sometimes incorrectly. So transition from chance performance to perfect performance
is gradual, not abrupt.
- Signal Detection Theory (Green & Sewts, 1966)
- Sensory systems are imperfect and inherently noisy – random cell firing
- This internal noise interferes with our perceptual decisions about the world
when the stimulus is weak
- Leads to incorrect decisions some of the time, so thresholds are not absolute
measures of performance – about 75% correct.
N EUROPHYSIOLOGICAL A PPROACH – SINGLE CELL RECORDING
Recording action potentials elicited by a range of visual stimuli.
Electrical impulses from each cell are amplified and sent to a loud speaker to be recorded.
The experimental animal faces a screen on which a range of visual patterns are displayed.
This enables the experimenter to map the receptive field of a cell (the pattern of visual
stimulation on the retina that changes the cells firing rate)
A peri-stimulus time histogram is a plot showing how the firing rate of a cell changes during
the time a stimulus is presented within its receptive field.
Cells potential to detect a threshold can be used to derive neural thresholds.
But single cell recording is essentially a reductionist approach to perception.
It attempts to reduce a complex problem into a set of basic problems we can study in the
lab.
But tells us little about perceptual processes which rely on the combined activity of many
neurones.
Cells as feature detectors?
- Individual cells cannot be considered as feature detectors.
- This is because the activity of ANY one cell is inherently ambiguous with regard to the
actual stimulus within its receptive field that made it respond.
- Although cells typically respond better to some stimuli than others their firing rate
depends on many other factors.
- So the brain must work out what is in the world by the patterns of activity within
populations of neurones.
STUDY OF THE HUMAN BRAIN
Lesion studies / neuropsychological patients – perception can be disrupted by neurological
disease or brain damage
Brain images affect lots of areas of the brain.
(LECTURE ONE)
We have no direct contact with the physical world – only the information conveyed
by our sense organs to our brains. All knowledge of the outside world depends on our
senses. We have the illusion of having direct contact with the world.
Perception can be seen as the psychological processes and underlying physiological
mechanisms by which we gain knowledge of the world via our sense organs.
Our perceptions are an accurate judge of reality - we can judge the speed of cars so we
don’t get run over. But we still see illusions.
Perception is an active process of interpreting sensory information to guide our
interactions with the environment.
We understand things from our sight – things which a camera can’t understand.
Seeing appears easy because we have so much specialized neural circuitry devoted to it –
half of the cortex is devoted to sight.
S TUDYING P ERCEPTION
Discoveries – e.g. all of the colours we can see can be duplicated by mixing 3 basic primary
colours in appropriate amounts.
Psychophysics: The study of the quantitative relationship between sensory
experience (psycho) and environmental stimulation (physics)
The central concept is the measuring of thresholds.
Detection threshold: weakest stimulus value needed to evoke a sensation in the observer.
Discrimination: (difference threshold): smallest difference between 2 stimuli along a
particular dimension that can be detected.
Measuring Detection Thresholds psychophysically
Limits:
- Change stimulus strength until subject says just detectable/undetectable
Adjustment:
- As above but subject adjusts stimulus until just detectable/undetectable
Constant Stimuli:
- Present subject with a fixed set of stimulus strengths in a random order and ask them
each time whether or not they can detect it.
Forced Choice task:
- Solves problem of subject being biased or not telling the truth
- Ask subject to report something about the stimulus, e.g. which side
- Calculate percent correct responses and threshold.
- Threshold = stimulus strength producing 75% correct performance (5 units)
Measuring Discrimination Threshold Psychophysically
Ideally use a forced choice task
One stimulus is presented (first) and then a second one.
, The observer has to report which stimulus has the greater strength.
Calculate percent correct responses, plot a psychometric function and estimate the
threshold.
THRESHOLD PERFORMANCE
- Infer properties of perceptual mechanisms from changes in thresholds.
- For the SAME stimulus strength the subject sometimes responds correctly and
sometimes incorrectly. So transition from chance performance to perfect performance
is gradual, not abrupt.
- Signal Detection Theory (Green & Sewts, 1966)
- Sensory systems are imperfect and inherently noisy – random cell firing
- This internal noise interferes with our perceptual decisions about the world
when the stimulus is weak
- Leads to incorrect decisions some of the time, so thresholds are not absolute
measures of performance – about 75% correct.
N EUROPHYSIOLOGICAL A PPROACH – SINGLE CELL RECORDING
Recording action potentials elicited by a range of visual stimuli.
Electrical impulses from each cell are amplified and sent to a loud speaker to be recorded.
The experimental animal faces a screen on which a range of visual patterns are displayed.
This enables the experimenter to map the receptive field of a cell (the pattern of visual
stimulation on the retina that changes the cells firing rate)
A peri-stimulus time histogram is a plot showing how the firing rate of a cell changes during
the time a stimulus is presented within its receptive field.
Cells potential to detect a threshold can be used to derive neural thresholds.
But single cell recording is essentially a reductionist approach to perception.
It attempts to reduce a complex problem into a set of basic problems we can study in the
lab.
But tells us little about perceptual processes which rely on the combined activity of many
neurones.
Cells as feature detectors?
- Individual cells cannot be considered as feature detectors.
- This is because the activity of ANY one cell is inherently ambiguous with regard to the
actual stimulus within its receptive field that made it respond.
- Although cells typically respond better to some stimuli than others their firing rate
depends on many other factors.
- So the brain must work out what is in the world by the patterns of activity within
populations of neurones.
STUDY OF THE HUMAN BRAIN
Lesion studies / neuropsychological patients – perception can be disrupted by neurological
disease or brain damage
Brain images affect lots of areas of the brain.
