Assignment

James Gibson’s theory of direct perception emphasizes

one particular aspect of vision. Give a brief overview
about the central claims of the theory and at least two
different applications of these ideas to real-world tasks.
PSYCHOPHYSICS
MEASURING SENSORY CAPACITIES
 MEASURING SENSORY CAPACITIES

► We are going to explore how we sense, experience or

interpret the world.
 We will look at measurement of sensory capacities or
specifically how we measure and make decisions in
psychology.

 The sense organs (ear, skin, eye, tongue and nose)

are our link to reality.

But what are the limits of the sense organs?

 An approach known as psychophysics provides the

insight.
 In psychophysics - physical changes in the stimuli are
measured and related to psychological sensations (e.g.,
brightness, loudness, taste etc).
This is the subject matter of Psychophysics.

Sensation and Perception

 The human brain is insulated from the outside world by

layers of tissue. Knowledge from the outside world is
carried into the brain through the sensory systems of the
eye, nose, tongue, skin and the ear.
Sensation
¨ Sensation refers to the initial detection of energy
from the physical world.

¨ It is the process of the sensory organs transforming

physical energy into neurological impulses the brain
interprets as the five senses of vision, smell, taste, touch
and hearing.

 The sensory systems respond to aspects of the

environment to produce sensations such as light,
smell, sound, taste, touch/pressure.
 We could consider sensation as the psychological
experience of a stimulus.

NB Sensations as we experience them are private or

subjective events.

¨ You may describe your sensations to others, but no one

else can directly experience them.

Example:

If a particular change in the physical environment

evokes a similar response from many people, we can then
assume that we have experienced similar sensations (e.g.,
turning up the volume of a stereo may cause most listeners
to describe the sound as loud).
 NB This shows that there is a consistent relationship
between a physical stimulus (amount of energy coming from
the environment e.g., sound waves from the speakers) and the
listener’s sensations (descriptions of what they hear – sound).

 These relations are referred to as psychophysical

relations because they express a relationship between
physical and psychological variables (stimuli and sensations).

 We can therefore say that the relationship between the

physical changes of the world and the psychological
experience associated with these changes is called
psychophysics.
 SENSORY SYSTEMS

► Our body is equipped with sensory systems (ear, tongue,

eyes, skin and nose) and these systems are sensitive to
some form of physical energy.
  For example:

Our Auditory System responds to rapid variations in air

pressure we call sound.

 Our Visual System responds to specific forms of

electromagnetic energy we call light.

 Taste is the sensation stimulated in taste buds (chemicals)

when food, drink, or other substances are in contact with
them.
Sweetness, saltiness, bitterness and sourness are
considered the four basic taste sensations, and all flavors
combine these in various ways with the sense of smell.
 Conditions in Order to Experience Sensation

. First as human beings, we can only respond to stimuli in

the environment of which we have “receivers” or
technically the receptor organs (ears, nose, eyes etc).
 Some of the stimuli we are unable to respond to.
E.g., electromagnetic energy in the form of radio and
television waves - our bodies cannot receive them unless we
have a radio and a television.

2. Second, the energy that we are receiving must be intense

enough (loud) to produce a noticeable sensation (If the
energy is low in radio frequency, we get poor reception).
For example, source of sound must be intense enough for
us to hear.
The Five Senses
 Sense Modalities and where their Receptors are located

1. Receptors for hearing (Ear) – these are the hair cells in

the inner ear that responds to rapid vibration of air
pressure.
2. Receptors of touch, pain, warmth, cold and other skin
senses (Tactile senses) – These are mostly found at the tip
of the fingers.

A B
3. Receptors for odor (sense of smell). The receptors are
located in olfactory epithelium. It is estimated that humans
have over 40 million olfactory receptors.
4. Receptors for taste (sense of taste – gustation)
these are taste buds and are grouped into structures
called papillae (small projections/protuberations of
tissue on the tongue).
5. Receptors for the eye – retina (these receptors
respond to electromagnetic energy we call light.
PATHWAYS FROM STIMULUS TO RESPOND

 Sensory nerves carry inbound information about the

body's movements and sensations.

 Motor nerves carry impulses outward to the muscles and

glands.
 Mixed nerves perform both functions.
Psychophysical Measurements
How the early psychophysicists measured stimuli and
sensation
Measurement of Sensory Limits
 Early experimental psychologists who engaged in
psychological measurements tried to assess our sensory
limits (the point at which one can experience sensation) by
varying the strength or intensity of a stimulus in order to
determine the minimum level capable of evoking a
sensation.
 Our focus will be on two types of sensory limits:
- Absolute threshold

- Differential threshold.
Terminal threshold:
the level beyond which any increase to a stimulus no
longer changes the perceived intensity.

Recognition threshold:
the level at which a stimulus can not only be detected but
also recognized.
 Absolute Threshold

1. In psychophysics, Absolute Threshold refers to the

hypothetical barrier which incoming stimuli must cross
before they can be perceived.

▬ It is the lowest intensity of some stimulus that produces

a response.
Differential Threshold

2. Another kind of Sensory Limit (Differential

Threshold) is defined by our ability to notice a change in
sensation. E.g.,, a source of light has to be increased or
decreased in some intensity for someone to notice that
change.

 Early experimental psychologists tried to measure

exactly:

What is the minimal amount by which the original

physical stimulus must be increased so that the
difference between the old and new intensities
would be noticed.
  Stated in another way, we are interested in finding out:

the smallest difference in a physical stimulus that

would be noticed.

 This amount is referred to as the Difference threshold

(the level of stimulus change necessary in order for a
person to perceive a difference.

 This is also known as Just Noticeable Difference (JND).

- defined as the amount of stimulation necessary to detect
a difference between two stimuli or the smallest change
from the standard stimulus that an observer can
notice.
  In other words, the difference threshold produces a just
noticeable difference - a psychological entity which
describes the subject’s ability to discriminate.

 With difference threshold, we are simply asking:

●how much must a stimulus change (increase or

decrease) before it becomes just noticeably different?

 These methods Absolute threshold and Difference

threshold were among the earliest methods developed to
attempt precise measurement in psychology.
  They are still of interest, not only on historical grounds, but
also because they are simple and adequate as rough
measures.

 The study of Just Noticeable Difference (JND) led to one

of psychology’s first natural “laws”.

 This is the Weber’s law, stated as:

the amount of change needed to produce a JND is a

constant proportion of the original stimulus intensity.

We will return to the Weber’s law and its significance.

 Absolute Threshold and Difference Threshold (similarity)

 Absolute threshold may be considered as a special case of

a difference threshold.

For example, if we ask a question.

How much stimulus energy must be added to zero
stimulus before the subject can tell the difference between the
old stimulus “I see nothing” and the new “Now I see it”?

 In this situation, the absolute threshold and the differential

threshold will be the same.
 Measurement of Absolute Threshold

 By this we are asking - How much stimulus energy is

needed to trigger a conscious perceptual experience?

Think about the following questions:

 What is the quietest sound that can be heard?

 The weakest light that can be seen?
 The lightest touch that can be felt?

 The minimum detectable amount of light, sound, pressure

or other physical energy is the absolute threshold. For
normal human beings these limits can be amazingly low.

► Estimates (limits) for the absolute threshold:

Galanter, 1962
 Measuring Sensory Intensity
 How do we measure Psychological Intensity?
(quantitative strength of stimulus or sensation)

► For humans, measuring the magnitude of a stimulus is in

principle very easy. For example, we measure physical
stimulus energy in pounds or kilograms, temperature in
degrees centigrade/Fahrenheit and sound in decibels etc.

 This is not so with sensory intensity, and it is extremely

difficult when we try to assess psychological intensity –
the magnitude of a sensation rather than a stimulus.

Measurement

► The founder of psychophysics Gustav Theodor Fechner

(1801-1887), believed that sensations cannot be measured directly.
  He believed sensations and the stimuli that produced
them belong to two different realms.

In philosophical terms this is referred to as the body

(stimuli) and that of the mind (sensation).

 Fechner, believed that while sensations can’t be

compared to physical stimuli, at least, they can be compared
to each other.

Example, We can compare two of our sensations and

judge whether the two are the same or are different.

In trying to resolve the computations, he resorted to the use

of Weber’s Fraction.
 The Weber’s Fraction

 For Fechner, measuring the jnd was just a means to a

larger goal – he wanted to formulate a general law relating
stimulus intensity to sensory magnitude.

 Fechner believed such a law could be built upon the

empirical generalizations which were first proposed by the
German physiologist Ernst H. Weber (1795-1878) in 1834.

 Weber believed that the:

“size of the difference threshold (JND) is a constant

ratio of the standard stimulus”.
 Examples

 Let us assume that we can tell the difference between

100 and 102 candles burning in a dark room.
If Weber is right, then we would be able to distinguish
between:

200 and 204 candles,

400 and 408 candles,
800 and 816 candles and so on.

 Weber’s claim is that the difference threshold (JND)

between two stimulus magnitudes is a certain constant
fraction of the total magnitude.

 He was so impressed with this relationship that he referred

to it as Weber’s law, which is written as:
 The fraction I / I = K is often referred to as the -
Weber’s Fraction.

Example:

If 1 kg must be added to a 10 kg standard weight for a

j.n.d, then 10 kg would have to be added to a 100 kg
standard weight to produce a j.n.d.
 Example: 2

 If the price of a 50-cedi candy goes up by 5 cedis,

shoppers might note the change; similarly, it might take a 5
million cedis price hike in a 50 million cedis Mercedes-Benz to
raise the eyebrows of its potential buyers.

In both cases, the price went up by 10 percent.

Let’s see the Representative Values of Weber’s Fraction:

Table 4.2
  Importance of Weber’s law.

 We may use it to compare the sensitivity of different

sensory modalities. Consider you want to know whether the eye
is more sensitive than the ear.

How can we determine this?

Obviously, we cannot compare the JND for brightness

and JND for loudness.

 The values will certainly be different – e.g., in different

units. For brightness we measure in millilambert and
decibels for sound.
  This difficulty is circumvented by using the Weber’s
fractions for the two sensory modalities. Fractions are
dimensionless and can therefore be used to compare the
sensitivity of different senses.

 When the fraction (I / I ) is small, the discriminating

power of the sense modality is great – that is
proportionately only little must be added to the standard for a
difference to be noticed.

The opposite holds true when the fraction is large. That is,
when the fraction is large, the discriminating power of the
sensory modality is weak.
 Based on this analysis we can now refer to the table and
determine which sensory modality is sensitive or you may
be able to compare the senses.
Weber's law states that the bigger the stimulus, the
bigger the increment needed for a change to be
detectable.

Other Uses

Weber's law is often used in marketing, particularly

with regards to price increases for products and
services. It implies for example, that it is possible to
increase prices by small enough amounts – that fall
under the “absolute threshold” – without your
customers even noticing.
 Weaknesses

  While reading these important contributions from

Weber on psychological measurements, keep in
mind that more recent research indicates that
Weber’s law should be viewed as only rough
characterization of our sensitivity to changes in
stimulation.

 It fails in the case of very weak or very intense

stimuli and is only approximately true for the
middle range of stimuli.
Nevertheless, it is a useful general approximation
of human sensitivity to stimulus differences.
 How do we Detect and Decide the presence of
a stimulus?

 The basic idea behind the threshold (limen) is that it

constitutes a boundary between perceptibility and
imperceptibility or a boundary between “I can see it” and “I
cannot see it”.

 Stimuli below the threshold evoke no response in the

sensory system (subliminal perception); those above the
threshold result in conscious experience.

 This implies that for the Psychophysicists, sensation is an

“all-or-none” affair and that a stimulus is either detectable
or not detectable.
 ► There are three basic problems with the classical
psychophysicists.

1. Sensory experience is not always all-or-none; instead,

we are often unsure whether we are perceiving a weak
stimulus (uncertainty).

2. The second problem involves the methods used to

measure thresholds. For example, how do researchers know
for sure that the subject really perceives a stimulus (e.g., a
tone) whenever the subject claims to hear something?
In other words, since sensation is inherently private, how
could anyone tell if the subject is lying?
 3. Thirdly, in determining the relationship between a subject’s
responses to the characteristics of the physical stimulus,
the decision making process may be influenced by his
expectations or wishes.
 Early psychophysicists disregarded the above factors.

NB Recent approach to psychophysical measurement insists

that they cannot be ignored.

 This current approach is Signal Detection Theory, a very

influential way of thinking about the way people detect and
make decisions.

 Note that in Detection and Decision making, two

important factors are known to affect subject’s responses. They
include: Response bias and Sensitivity of the stimulus.
  Response Bias
Response bias is a person’s willingness or reluctance to
respond to a stimulus, which has to do with motivation,
needs and expectations.

 If the stimulus (signal) is weak the subjects become

uncertain about their judgment. Under such circumstances, their
Response Bias will exert an effect. (It is a preference for one
response over another (either “Yes” or “No”) quite apart from the
nature of the stimuli).

► Sensitivity of the stimulus

 Early psychophysicists determination of thresholds was

largely influenced by the intensity of the stimulus.
  High intensity or sensitivity reduces response bias of NO
and Low intensity also increases response bias (of No)
(intensity is the strength of the signal).

Simply, we can conclude that determination of the

thresholds reflect the two factors.

a. Sensitivity of the stimulus and

b. Response Bias.

 Signal detection theory has developed a more systematic

way of dealing with response bias and that is catch trial
(where no stimulus is presented) becomes part of the
regular procedure to keep the subjects on their toes.
Distance 18/3/23
Signal Detection Theory

 The effects of noise and response bias can never be

entirely eliminated during detection. Researchers have
gone beyond trying to determine thresholds and they have
turned to Signal Detection Theory.

 Signal Detection Theory is a mathematical theory of

detection of physical signals. Its assumption is that the
sensitivity to a signal is not merely a result of its intensity but is
also dependent upon the:

amount of noise present,

the motivation of the subject and

criterion which the subject sets for responding.

 The theory is concerned not only with how people detect
faint lights and sounds but with their response to events of
any kind.

For example,
a. How good are you at finding typing errors in your term
paper?
b. How likely is a football referee to detect a foul?
c. Will the airport security guard detect a hidden explosive
in a hijacker’s x-rayed luggage?
d. What causes “false alarms” that lead the guard to
“see” weapons where there are none?

 Signal detection theory provides a way to understand and

predict correct responses, missed signals and false alarms
in these and many other situations.
► In Signal Detection experiments, there are two kinds of
errors.

 Miss – not reporting a stimulus when one is present.

 False alarm – reporting a stimulus when in fact none is

present.

We also have two different kinds of Correct Responses:

 A hit - reporting a stimulus when it is actually there and

 A correct negative or rejection – not reporting one

when none is present.
Responses in Signal Detection Experiment
 Response Bias can have a serious effect on our mode of
response.

 So what factors underlie or explain Response Bias?

 Why do subjects prefer one response over another (why

do they prefer to say “Yes” or “No”?).

(We are going to use the payoff matrix to produce a

preference for Yes response

The Payoff Matrix

► A Payoff Matrix - is a prescribed schedule of gains and

losses and will help us to understand some aspects of
response bias.
Example 1

► A Payoff matrix that will produce a “YES” bias.

In a detection experiment, if we agree to pay a subject :-

¢100 for every hit and

¢50 for every correct negative,

but penalize him by losing ¢100 for every miss and only

¢10 for every false alarm, we can put this schedule in a

payoff matrix.

 This payoff matrix will obviously lead to a “Yes” bias.

 (hit) (Miss)

(False Alarm) (Corr Neg)

► Suppose there are 50 trials on which the subject has no
sensory information to base his decision on (whether the
stimulus is present or absent). Signal will be presented
50% of the time.

 If he consistently responds “YES” he will on the average be

correct on 25 trials and collect ¢2500 and will be wrong on
the other 25 trials, losing ¢250.

Yes Bias will result in a net gain of ¢2250.

Yes No

Yes No
 In contrast, a consistent “NO” response or decision will
lead to a net loss (+1250 for the correct negatives and
- ¢2500 for miss).

“No” response will result in a net loss of - ¢1250.

 Under these conditions, subjects will do well to adopt a

liberal attitude or criterion and give a “YES” judgment or
response when they are in doubt.

NB. In this case, a response bias becomes a function of the

subject’s motive.
 Lets see other effects in real life

¨ The effect of the payoff matrices may also occur in real

life. Imagine a radiologist using X-ray to look for indications
of a malignant tumor.

¨ What are the penalties for error?

If the physician decides there is no spot when there is

actually one, the miss may result in the death of the
patient.
 If on the other hand, he decides there is a spot when there
is in fact none, the false alarm has its costs – cost of
elaborate and expensive clinical tests, anxiety and even an
operation.

 Generally, physicians will decide on what their eyes tell

them based on the X-ray and also the relative costs of the
two possible errors (Miss and False Alarm) they may commit.
 Separating Sensitivity and Response Bias

► If the intensity (sensitivity) of the stimulus is loud, the

subject will definitely respond YES even if his response bias
is to say NO. But if the stimulus is relatively faint he is likely to
respond NO.

 Can the effect of Stimulus Sensitivity be

disentangled from Response bias?

The detection procedure itself can provide us with a means of

disentangling stimulus sensitivity from response bias:
1. First, you can vary response bias while keeping stimulus
sensitivity constant. This can be done by changing the
payoff matrix (the schedule of gains and losses).

2. Another option is to vary the proportion of trials on which no

stimulus is present – the fewer such trials (of no stimuli),
the greater the YES bias.

3. We can induce the subject to be more conservative in his

response. This can lead to a reduction in the proportion of
trials on which he is guilty of false alarm, but as we induce
subjects to be conservative, this also leads to a reduction in
the proportion of Hits.
4. Similarly, upward shifts in the YES bias will also lead to an
increase in the proportion of both hits and false alarms.

► These effects can be demonstrated on a graph by plotting

the two proportions against each other.

The resulting function is referred to as:

Receiver-Operating-Characteristic curve –
abbreviated ROC curve.
 The proportion of hits is plotted against the proportion of
false alarms. The stimulus (sensitivity) is kept constant but
the response bias is systematically varied by asking the
subjects to adopt a more liberal attitude with their yeses.
 Note that as the proportion of hits increases, so does the
proportion of false alarms.
Signal Detection and the Decision Process

► Why does the subject in an experiment have so

much trouble in making a distinction between the
presence of a stimulus and its absence?
 Signal detection theory proposes an interesting answer.

There is no such thing as zero stimulus.

That is, in every sensory process (e.g., noise generated
either by external stimuli or internal stimuli), there is an
underlying neural activity always going on.
  Signal detection theory proposes that a sensory process
will occur even when no stimulus is administered.

 The factors responsible for such sensory process are

collectively called background noise.

Example:

 Consider a hearing experiment where no sound is actually

presented. Although no sound is presented (external
signal) it doesn’t guarantee the absence of activity in the
auditory system (or that the subject will not hear anything).
  The auditory system may respond to activities from within
e.g., the throbbing of the pulse, activity in the nervous
system (the firing of several cells at random intervals without any
external trigger).

► In a detection experiment the subject’s task is to decide

whether a given sensory process should be attributed to
the background noise alone or to the signal which is
superimposed on the background noise (signal+noise).

The indecision arises because of the fluctuation in the

two sets of sensory processes (the background and
the signal).
Occasionally, the subject’s heartbeat sounds louder than
the experimenter’s stimulus tone.
If the stimulus is intense, little confusion will arise in a
detecting experiment.
 In sum Signal Detection theory offers a different
conception of what a sensation is than did the earlier
psychophysicists.

Its basic assumption is that sensitivity to a signal is not

merely a result of the signal’s intensity but also
dependent upon:

 the amount of noise present within and outside the

subject,
· the motivation and expectations of the subject and the

 criterion which the subject sets for responding.

The decision processes
are captured below:
Noise alone Signal + Noise
 Criterion (Decision Rule)
No Response Yes Response

A B C D

A: CORRECT NO RESPONSES
B: MISSES; INCORRECT NO RESPONSES
C: FALSE ALARMS; INCORRECT YES RESPONSES
D: HITS; CORRECT YES RESPONSES
 Applications of Signal Detection Theory

► The general approach of signal detection theory has

applications to many areas of psychology outside of
psychophysics.

 It is relevant whenever a person has to decide between two

alternatives but cannot be sure of the outcome.

Examples,

1. Signal detection theory has applications to eyewitness

testimony. Justice demands that the detection process is
sensitive as possible to produce the fewest possible
misses and false alarms.
  Eye or Ear Witnesses (victim) usually bring with them a
response bias. Their response criterion may be so low that
they are likely to identify someone in the line-up (as
perpetrator). We call this risky bias.

 Witnesses may be reminded that the criminal may not be

part of the line-up.

But such reminders also tend to lower witnesses

expectations and raise their response criterion.
 2. Signal detection framework has also been applied to the
task facing radiologists reading x-rays. Research has
shown that the accuracy of radiologists’ judgement in these tasks
rarely exceeds 70 %. These misses and false alarms have
implications for the health of the patient.

3. Another example is the selection of college applicants.

One error of admission is a miss. That is, an applicant is
rejected who would have done well. Another is false alarm:
An applicant is accepted who will be unable to graduate
(flunk out).
  If one wants to minimize the misses (rejected applicants
who would have done well had they been accepted), one
has to change the response bias toward “yes” which then
inevitably increases the number of false alarms
(accepted applicants who will flunk out) and also hits.

 As you can see, which trade-off is chosen depends on the

payoff matrix.