2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 1

2006PSY Cognitive Psychology Mid-Trimester Exam

Detailed Study Notes for Weeks 1 & 2
Please see document 2 for weeks 3 – 4 (Cannot upload together because too large of a document)

Week 1: Introduction to Cognitive Psychology

Cognitive Psychology is:
 The study of how people perceive, learn, remember, and think about
information
o Visual perception n
o Attention
o Speech and language
o Memory
o Reading
o Decision-making
o Task-switching

Cognitive Psychology
 Concerned with the mental processes that underpin our everyday functioning
 Everyday cognition is rich and complex
o Cognitive processes
 Can occur together or in close succession
 Can proceed rapidly (at speeds that exceed
computers)
 Can proceed automatically (without conscious
thought/control)
 Are covert (not directly observable)
 How to study cognition with sufficient rigor to draw firm conclusions?
o Methods used are almost as important as the findings

History of Cognitive Psychology

 Origins go back to great philosophers of the ages
o Aristotle (384-322 BC) investigated memory and learning
o Aristotle was held an empiricist position, which argued that
knowledge is learned, comes from own experiences, and
environment plays a powerful role in shaping intellectual ability
o Argued that we can study thought and cognition through the
process of introspection
o Also proposed three laws of association

Introspection
 Introspection is the systematic investigation of one's thoughts, through written
or spoken records.
o Aristotle and others believed that we could gain an understanding,
or at least some insight into thinking
o Introspection as a formal technique could be taught
 Wundberg school of "imageless thoughts"
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 2

 Focus on stimuli
 (narrating aloud to an observer) report everything
goes through their head
 Revealed that people consciously think about many
things, including sensations, feelings, images,
conscious mental sets, thoughts.
 Limitation in that it only captures conscious thought,
and not fast automatic processes

Early Schools of Experimental Psychology

 Structuralism: Formal scientific investigation undertaken by Wilhelm Wundt
(1832-1920)
o Established research lab in Lepizig, Germany
o Psychology: scientific study of conscious experience
 Analysed conscious experiences into its basic elements/components, and
showed relationships between them.
 Expanded by Titchener into structuralism

Main Claims
 Raw materials of consciousness are sensory
 With proper training, people can detect and report the workings of their own
minds (introspection)

Evaluation
 First approach to cognitive questions scientifically
 First to use experiments to test cognitive theories
 Introspection technique was problematic
o Not everyone could be trained to use it,
o Only accesses conscious (non-automatic) thought processes
o Much of cognition occurs outside of consciousness (we only
experience the outcome not what goes into it).

History of Cognitive Psychology

 17th - 19th centuries
o John Locke, David Hulme, John Stuart Mill, George Berkeley
(empiricist position)
o Rene Descartes, Emmanuel Kant (nativist position)
 Nativist position emphasises role of biology (nature)
in cognition
 Some hard-wired functions that are present at birth
 Individual differences in abilities are attributed to
biologically endowed capacities

Behaviourism
 Argument made by Watson (1878-1958) that for psychology to be a credible
'hard science', we should only consider observable stimuli and behaviours
o "Not that imaginary stuff we call thoughts"
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 3

 Behaviourism was an approach that would "never use the terms consciousness,
mental states, mind, content, introspectively verifiable, and the like". Watson
(1913. p.165)

And so begins the dark age of cognitive psychology...

Behaviourism
 B.F. Skinner (1904-1990) extended behaviourism, along with other researchers
like Pavlov, into studying classical and operant conditioning.
o Skinner agreed with Watson that we should ignore mental
processes, and focus only on the external.
o Extended Darwin's concept of biological adaption to environment
to cover psychological phenomena
 Work has relevance for learning, especially phobias (extinction) and motivation
(prior systems of reward punishment)

Evaluation
 Focuses on observable responses to stimuli (stimuli-response)
 Criticised introspection as too subjective and unscientific
o 'images' and 'thoughts' were just labels for internal bodily
processes (Skinner)
 Did not study higher order processes (thought, decision-making)

The Cognitive Revolution

 1900-1950, behaviourist tradition dominated psychology
 During and after World War I/II, behaviourist assumptions began to be
questioned, based on developments in psychology and other disciplines
 Renaissance for cognitive psychology, as mental representations were seen
both as:
o Necessary to explain human behaviour
o Able to be formally studied
 Major advance in understanding of language (Chomsky, 1956), short-term
memory (Miller, 1956) and problem solving (Newell and Simon, 1958).
 Snowballed in late 1960's 1970's with advent of computers and information
processing.

Gestalt Psychology
 Study of perceptual processes and reasoning
o Ψ phenomena are best understood when viewed as organised,
structured wholes (whole > sum of parts)
o Opposed to introspection which was about dividing thought
processes into component parts
o Argued couldn't understand perception in pieces, only by looking
at the whole

Genetic Epistemology
 Jean Piaget (1896-1980)
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 4

o Studied cognitive development

o Children at different stages of development use different mental
structures/processes to perceive, remember, and make sense of
the world
o 4 stages of cognitive development
 Sensorimotor (0-24 months)
 Preoperational (2-7 years)
 Concrete operational (7-11 years)
 Formal operation (adolescent to adulthood?)

Individual Differences
 What is the nature of mental abilities?
o Are they uniform or do they differ across individuals
 Are intellectual abilities inherited?
o Genetics of cognition (heritability)
 Or are they instead learned and developed?
o Availability of instruction, practice, schooling
o What factors affect our cognitive abilities
 Study of individual differences
o Many tests and questionnaires to measure intellectual abilities
o Statistical techniques to analyse the data
o Forerunner to psychometrics

Information Processing Approach (Atkinson and Shiffrin, 1968)

 At one time, cognitive psychologists endorsed an information-processing
approach.
 It posited that we're driven by external stimuli (much like behaviourism), and
then that triggers internal cognitive processes
o Attending to the stimuli,
o Perceiving it,

o Thinking about it and making a decision

 These occur sequentially, one after the other (serial processing)
 The way in which the stimulus affects our thoughts is termed bottom-up
processing
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 5

Bottom-up Processing

Top-down Processing
 So in bottom up, its really the stimulus driving successively higher-order
cognitive processes.
o Greatly oversimplified explanation of a far more complex reality
 But our thoughts and feelings often drive how we perceive external stimuli
 What you're thinking of guides what you perceive it as

So it is bottom-up or top-down processing?

2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 6

 It's a combination of the two.

 Lower-level perceptual processes DO drive how we think and perceive the
world. But our thoughts and prior experiences also drive how we perceive and
think about the world as well.
 See examples on page 11 of textbook
 Cognition involved both bottom-up, and top-down processing (interactive
processing).

Parallel Processing
 Another limitation of the traditional information-processing approach was the
assumption that processing is a sequential (i.e. One after the other)
 Different regions of the brain are recruited to attend to stimuli, process it, and
make conscious thoughts/evaluations.
 In most cases, we are engaged in parallel processing.
 I.e. Mental processes occur concurrently
 Scenario:
o Think of the task of driving a vehicle. You're steering the vehicle,
applying pressure to the break pedal, your eyes are watching not
just the car in front of you but also the other vehicles and the
road. You might also be listening to music, or thinking about which
turn you'll have to make to get to work
 Imagine if you had to finish one task before starting
the next?

Contemporary Cognitive Psychology

 Modern cognitive psychology is guided heavily by experimental research and
evidence from a variety of disciplines/fields, including:
o Experimental cognitive psychology
 Recruiting neurologically healthy participants to
complete cognitive tasks under laboratory conditions.
o Cognitive neuroscience
 Involves the additional gathering of physiological
brain measurements (electrical, blood flow) and
sometimes manipulation (transcranial magnetic
stimulation, direct current, electrodes)
o Cognitive neuropsychology
 Recruiting neurologically unhealthy, or brain
damaged patients, in the hope of understanding how
regions of the brain contribute to cognition and
behaviour
o Computational cognitive science
 Computer modelling of brain connections, and
simulation of neural networks to test hypotheses
about brain organisation and structure.

Experimental Cognitive Psychology

 The vast majority of cognitive research falls into this category
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 7

o Carefully controlled research studies present a standardized

stimulus to participants, and take measurements (either
performance on cognitive tasks, reaction time, or both)
o Typically use healthy individuals (often student subject pools) so
that results can be generalised from the sample to the general
population
 But this can be a limitation, because subject pools
substantially differ from general population in
important characteristics, like education,
socioeconomic status, etc.
 Advantages of experimental cognitive psychology
o Increased experimental control over stimulus and participants
o Easier to demonstrate cause-and-effect
o Reduced measurement error, and eliminate confounds
 Disadvantages of experimental cognitive control
o May lack ecological validity to real world cognitive tasks

In a carefully designed experiment on a computer, we can record both ACCURACY of

responses and REACTION TIME.

Accuracy and Reaction Time

 In many cognitive tests, we are interested in the accuracy of responses (how
many items answered correctly, error rate, etc.)
o E.g. How many mathematics problems can you solve in a fixed
time span
o Rationale is that people who are higher on the construct of
interest (mathematics ability) would answer more items correctly
than those who are not
 In other cases, we may be more interested not in accuracy or speed but in how
many errors were made (or of what type).
o Certain cognitive errors can indicate impairment, or give insight
into underlying cognitive processes.
 But how long a person spends on a task is an importance performance indicator
too. Typically this is done on computer, and measured in milliseconds (MS). We
use the term reaction time (RT).
 But the very FIRST cognitive experiment was performed by the philosopher
Franciscus Donders, in 1868 (even before the term cognitive psychology was
used).
o He proposed a simple reaction time task where individuals press a
button when they see a light has been illuminated ~ 30ms
 Donders also designed a more complex task, involving additional cognitive
processing (thinking).
o In a second experiment, he presented subjects with TWO lights,
and they had to indicate which light by pressing the correct button
(forced choice reaction time).
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 8

o Cool part is that the different between these two tasks gives us an
estimate of how much extra time is taken to make the forced
choice (approx. 50 ms).
o Called Donder's Method of Subtraction

Experimental Cognitive Psychology

 Experimental cognitive psychology has provided considerable insights into how
people think, solve problems, and learn.
 Advantage is that you have complete control over the type of stimuli used, the
type of measurements (performance, reaction time, behaviour) taken, and the
environment of testing.
o Can rule out experimental confounds, such as
distraction/inattention, and improve measurement accuracy
o This leads to increased experimental control
o Easier to demonstrate cause-and-effect
 Disadvantage is that we're examining cognitive processes inn a very abstract,
artificial environment
o Cognitive tasks are very different to 'everyday' environment (at
home, at work, in the community)
o This limits generalisability to everyday cognition (ecological
validity)

Cognitive Neuroscience
 Experimental research can be extended by taking physiological measurements
of brain activity. Why?
o People often lack insight into internal cognitive processes,
o People may not tell the truth (self-report)
o People may be unable to communicate verbally (locked-in)
 Additionally we gain a better understanding of the brain by looking at what
parts of the brain are activated for certain tasks (e.g. Fusiform facial area FFA in
face-recognition tasks)
o Also look at how performance on a tasks relates to brain waves
(particularly in attention-disorders).
 Simplest way to investigate the brain is with a CT scan, which shows the
physical structure of the brain. We divided the brain into a series of lobes
(Figure 1.6 in textbook)
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 9

Note: this is a side on, so the front of the brain is on the left, and the back of the brain
(occipital lobe) at the back.

Cognitive Neuroscience
 We can also refer to more specific parts of the brain (rather than an entire
lobe). Anatomist and neurologist Korbinian Brodmann produced a map of the
brain, based on variations in cellular structure.
 While most complex cognitive tasks recruit from a variety of brain regions,
certain specific functions are controlled primarily by neurons located in specific
areas. For example Brodmann's Area 45 (BA45) controls mouth articulation and
is responsible for pronunciation in speech. People with damage to BA45 find it
difficult to speak a word (aphasia), even though they may be able to write or
type it.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 10

Figure 1.7 - The Brodmann areas of the brain on the lateral (top) and medial (bottom)
surfaces.

Looking inside the brain

 When we want a detailed look at structure of the human brain, we can use
magnetic resonance imaging (MRI).
o But this only shows the physical brain structure, not which parts
are active.
o If we want to see activity in the brain, we need to employ other
techniques, such as functional magnetic resonance imaging (fMRI)
o Measures small changes in blood flow, which is sent to replenish
the brain's consumption of glucose and oxygen.
o Changes are slow to register, with a latency of 2-3 seconds.
o Other techniques you'll come across is positron electron
tomography PET scans, and magnetoencephalography (MEG).
 While the information produced from such studies is useful, the major
disadvantage to fMRI and PET studies is cost. These are incredibly expensive
tools (many millions of dollars) that are used for medical diagnosis, and so
access to time on them is short.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 11

 This means that the sample size of such studies is generally very small,
sometimes as few as 5-10. if we want to generalise wider samples, we have to
do so VERY cautiously.
o E.g. Studies showing gender differences in brain structure with
only a handful of people are dubious
o Wide variability in size and interconnections of brain structures
due to differing learning and experience, so hard to show
causation.
o Many brains are atypical, yet perform identical to others in
cognition and behaviour

Other physiological measurements

 Well we could always insert electrodes into subject's brain.... But most people
are a little squeamish about participating!
 So the next best alternative is to use externally placed electrodes on the human
scalp, which provide a rough measurement of electrical activity in the brain
(electroencephalogram EEG).
o Painless, non-invasive measurements can be taken with much
smaller latency that fMRI measurements
o Far cheaper (though cost is still a factor), but often fiddly and
difficult to use outside the laboratory (ecological validity with 'real
world' cognitive tasks).
 Emotiv Insight

How does EEG technology work?

2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 12

 Advantages/disadvantages
o Good temporal resolution (millisecond accuracy), but poor spatial
resolution (each electrode covers a wide area).
 That trade-off makes it worth it through for
examining cognitive performance in a research
setting.
 However expense and fragility of system limited
ecological validity in everyday cognition... For now.
o Newer products like emotive will take the technology our of the
lab though, for consumer electronics (brain machine interfaces)
and also for neurorehabilitation and education.
o Other limitations include that only demonstrates association
between ERP and behaviour. Some tasks actually decrease brain
activation (particularly during concentration), and lots of
background noise (brain is always active to some degree).

Cognitive Neuropsychology
While experimental neuropsychology investigates neurologically healthy individuals, we can
also learn a considerable amount from the study of cognitive deficits in people with brain
damage or neurological disorders.
 Lesion is the term applied to structural damaged caused by injury or disease.
Often this is due to stroke or head injury, but many of our most 'famous' cases
come from industrial accidents and war.
 While intellectually and medically fascinating, remember that these are people
who have suffered serious impairment that diminishes their quality of life.
 However the information gained helps us understand the progression of
neurological diseases, as well as to study ways to help rehabilitate and treat
such conditions.

Cognitive Neuropsychology - dissociation

2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 13

Cognitive neuropsychology seeks to better understand brain function by searching for

dissociations in case studies.

A single dissociation occurs when a patient has a deficit in a specific cognitive function, but
is otherwise normal on other cognitive processes (i.e. Equivalent to health controls). This
means that damage is confined to one specific task.
 E.g. Patient has normal visual functioning, and can recognise and identify
objects. But the patient has difficulty recognising human faces (prosopagnosia).
 E.g. Amnesia patient may perform very well on tasks requiring short-term
memory, but be unable to form new memories (classic case, Patient H.M.)
o This suggests that short-term memory and long-term memory rely
on different brain structures.

Stronger evidence can be found in the form of a double dissociation (see textbook). But
these require matching two patients with opposite pattern, so harder to locate.

Assumptions of Cognitive Neuropsychology

Key assumptions:
 Modularity - cognitive system consists of numerous modules that operate
relatively independently (functional specialisation)
 Consistency - modules are organised consistently across subjects
 Subtractivity - damage to one module reduces performance, but doesn't learn
new processing strategies.
 Depending on the brain region, cognitive process, and individuals, these may be
called into question due to factors such as distributed processing across brain
regions, individual variability neuroplasticity.

Another limitation is gaining access to subjects with a lesion confined to just the location of
interest (and not more widely).

Transcranial magnetic stimulation (TMS)

Developed initially as a therapeutic technique for depression and other mental health
conditions, transcranial magnetic stimulation uses targeted electromagnets placed over the
scalp to deliver a small electromagnetic pulse. This is usually repeated rTMS.

Depending on the strength of the pulse and location delivered, it can either deliver
stimulation or suppress/distort electrical activity. In the later, this stimulated a lesion
without any lasting damage.

Perfect for cognitive neuropsychology, because we can decide precisely which region to
affect.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 14

Direct Current Stimulation

Another alternative (not covered in text), is direct current stimulation (DCS). Rather than
create temporary lesions, this seeks to stimulate specific brain regions and increases brain
activation in these areas.

Why? It's a non-invasive technique that can actually IMPROVE cognitive functioning
(memory, language, attention, problem-solving).

For more information, see:

https://www.sciencedirect.com/topics/neuroscience/transcranial-direct-current-stimulation

Computational Cognitive Science

Computer science and information processing theory has made substantial contributions to
cognitive psychology already (e.g. Modularity, parallel processing, etc.). But cognitive
psychology also contributed to the development of artificial intelligence as a field too!

This came in the form of computational modelling. These are intricate simulations of
cognitive processes and systems. Such computational models aim to do one specific task,
and to do it very well.

As computing technology progressed, the relative inefficiency of these simulations could be

overcome with sheer computational power.

 Yet there are still some tasks that a human infant can do better than a
computer.

Connectionism and Computational Cognitive Science

Rather than attempt to create a computer program to simulate cognitive tasks (dead end for
AI), connectionism is based on studies into the biological of the brain.

Neurons in the brain (both in animal and human models) are:

 More numerous than previously thought
 Arranged in a parallel fashion
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 15

 Massively interconnected and distributed

 Learning occurs when connections between two neurons are strengthened

Neural activity surround a particular cognitive action is typically distributed across a broad
range of neurons, not limited to a single pin-point sized location.

Connectionist networks make use of elementary units (nodes) that are interconnected and
also layered. This has lead to a form of computational modelling called artificial neural
networks, responsible for a rapid growth in machine learning.

Where these neural networks differ from traditional 'code' is that they are mathematical
models of associations and learning... So the internal workings are hidden, and cannot be
easily deciphered. We can only observe their outputs and actions.

These neural networks are trained with a large source of stimuli (such as pictures of objects),
mirroring the learning process.

For example, a neural network can be trained to identify objects (say pictures of cats).
Once trained, it can examine a picture that its never seen, and identify whether it
matches the target.

Incomplete or fault information can be tolerated to some extent --> allows for
imprecision in stimuli

Advantage of connectionist models is that learning can generalise from the initial training
stimuli to new novel-stimuli.

This way they can truly learn how to handle new situations, or when noise is present in
data (such as looking at an object under different lighting conditions, different angles,
etc. The way a human brain does)

Disadvantage of connectionist models is that they are a closed system (black-boxes)

There is no computer code to decipher, or mathematical equations to try and

understand. One trained they generally work, but we don't fully understand how they
make their decisions.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 16

As an example, consider a simple 'identify me as human' task we all encounter, the dreaded
CAPTCHA.

Traditional computer program is unable to read this, but an artificial network model can.

Likely that we'll see a blend of psychology and computer science making real strides in
artificial intelligence.

Combining Approaches
 Cognitive psychology can use evidence from all of these approaches (not an
either-or scenario).
 Textbook uses example of visual perception and visual imagery. Kosslyn et al
argues that visual imagery uses the same processes as visual perception.
 If so, we can test this prediction uses a variety of lines of evidence.
o People should scan visual images similar to basic visual arrays
(behavioural evidence from experimental cognitive psychology)
o Similar brain regions should be activated looking at a complex
image to (cognitive neuroscience imaging)
o Brain damaged patients with a visual-perception disorder should
also have impaired visual imagery, and vice versa (cognitive
neuropsychology)

Reproducibility and Psychology

2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 17

 Rather than relying on a single study (which may have a biased sample, poor
measurement, or just be a statistical fluke) we typically consider the results of a
series of studies before reaching a consensus about a particular topic.
 This is particularly important when looking at neuroimaging (fMRI) or brain
measurement studies (e.g. EEG), where sample sizes may be particularly small.
o Challenge to psychology as a field is the replication crisis, where
many 'intriguing' and 'novel' studies fail to replicate when
independent researchers attempt to do so.
 As cognitive psychologists, we need to be able to weigh up the evidence
ourselves.

Combining evidence from multiple studies

 Initial studies, particular if 'shocking' or 'surprising' should be interpreted
cautiously. Consensus comes from combining a variety of studies on an effect,
to see if generalises across samples.
o Looking at the number of studies for or against (vote-counting) is
not an appropriate method
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 18

o Particularly problematic with small sample sizes from cognitive

neuroscience studies, as they may lack statistical power
 Instead we use the statistical technique of meta-analysis, which is widely used
in psychology and medicine.

Meta-analysis
 Goal of meta-analysis is to aggregate the results of many studies, to provide a
clearer view of the literature.
o Originally developed by psychologists to review entire literature
o Across many studies, and many samples, provide an estimate of
how strong an experimental effect is present
 E.g. Is there a gender difference between males and
females on verbal and language tasks? If so, how
large is the difference?
 But a meta-analysis goes further, and also contributes new information that
might not be apparent from a single study.
o It does this by looking at moderators of the effect
 E.g. Developmental trend towards larger gender gaps
as age of participants increases.

Combining evidence form multiple studies

 Textbook often refers to meta-analyses, as they offer stronger evidence than
individual studies.
o However some (typically non-statisticians) have argued these are
problematic. For example, the 'file drawer problem', which posits
that only statistically significant studies get published, and null
studies finding no difference are lost to the winds.
 Sharpe (1997) reviews these as potential issues, which are readily addressed by
competent meta-analyst.
o For example, file-drawer problem can be easily overcome by
looking at non-published literature (honours, postgrad theses),
o 'apples and oranges' overcome by using strict inclusion/exclusion
criteria.
o 'garbage in garbage out' - We also frequently conduct a moderator
analysis for publication quality overcoming GIGO.

Evaluating evidence in cognitive psychology

 Cognitive psychology employs a variety of lines of evidence, including
experiments, cognitive neuroscience, cognitive neuropsychology and
computational cognitive science.
 However the methods used are almost as important as the conclusions
reached, and you should apply critical analysis to these. New techniques have
emerged in recent years, and still more are coming.
 No single study is sufficient to "prove" or "disprove" cognitive theories. The
"gold standard" is to synthesise results from numerous studies through the
technique of meta-analysis.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 19

Week 2: Visual Perception

Overview
 Visual Perception Overview
 Gestalt laws of perception
 Template versus Feature theories
 Object Recognition/ Face Recognition
 Perception and action
 Inattention/change blindness

What is Visual Perception?

To perceive, to track and to identify external stimuli, objects, people, at a distance, and to
process and interpret them.

Visual Perception
 Importance of visual perception in everyday life
o For most people born without impairment, visual perception is of
tremendous importance in everyday life
 Allows us to orient ourselves, to walk and recognise
our location
 Makes it possible to recognise objects, people's faces
 Supports higher order tasks like reading signs, books
and magazines
 Allows us to watch videos (tv)
 This doesn't mean that life isn't worth living for someone born blind, however -
in fact they still will use many of the brain regions responsible for visual
perception and many of the cognitive processes.
o Echolocation

Visual perception with ambiguous stimuli

 Magic of the visual perception system is that it works under some very trying
circumstances.
o Object recognition from different angles/perspectives, under
different lighting conditions (brightness, colour)
o Neural networks permit a tolerance for ambiguity, distortion and
noise
o Consider the case of different handwriting styles and font faces
o We can recognise people's faces, even if they change their hair
colour, grow a little older, or we see them in an unexpected
context. Yet there's a lot of similarity across faces (two eyes,
mouth, nose, skin colour, hair colour).

Sensation versus Perception

We draw a distinction between sensation and perception.
 Sensation is a biological process, and refers to receptors that react to external
stimuli (photons of light) striking the retina.
 Upon reaching the eye, rods and cones send electrical impulses through the
optic nerve to the visual cortex (located in the occipital lobe)
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 20

 Perception is quite a different concept entirely. Perception involves processing

of visual information in the brain, and the eventual interpretation of visual
stimuli.
o As such, it is a higher order task that relies on a number of
important cognitive processes
o These also draw on a number of different brain regions for object
recognition, memory, analytical thought and emotion.

Understanding perceptual processes

 We can gain a better understanding of the underlying perceptual processes in
the visual system through studying responses to carefully constructed visual
stimuli (aka. Illusions)
o Long tradition of doing this in Gestalt psychology, but also in art as
well

Gestalt Psychology
 Gestalt psychology sought to better understand visual perception, and the rules
that governed how basic sensory stimuli are processed by brain
o Argued Ψ best understood when viewed as organised structured
wholes (alternative to structuralism, believed could be brokenn
down innto components)
o Humans are hardwired to perceive organised patterns and objects,
principle of Prägnanz (laws of grouping)
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 21

 Proposed a series of six perceptual laws, explaining how components of

patterns are seen and perceived together.

1. Law of Similarity
Law of similarities proposes that visually similar objects will be seen as grouped
together

2. Law of Proximity
Law of proximity suggests that the brain uses proximity to group together shapes.

Even though made up of small circles, most people see a cube, followed by three tall
rectangles.
3. Law of Closure
Law of closure suggests that even given incomplete visual stimuli, the mind will tend to
see the complete figure or form and mentally 'fill inn' what is missing.

We see a circle, even though its not fully drawn there. Evolutionary psychologists argue
this is evolved from the need to be able to see predators or threats even if partially
obscured/camouflaged.
4. Law of Continuity
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 22

Law of continuity suggests that the mind will follow the smoothest path when viewing
limes, regardless of how they are actually drawn.

In this case, your brain perceives this as one horizontal path and one curved path...
Even though the paths change colour midway. They are seen as two complete paths.
5. Law of Common Fate
Law of common fate suggests that we see elements moving in a common direction
(referring to the below image: or in this case since its static, pointing in the same
direction) as a whole.

In this case we see a flock of birds flying in the same direction, rather than focusing on
individual elements.
6. Law of Good Form (Good Form/Good Figure)
Law of good form refers to the tendency to group together visually similar objects
(colour, shape, pattern, etc.)

In this case we see these as the six rings of the Olympic symbol, even though if you
look carefully you'll see a small break between the rings (they aren't actually joined).
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 23

Figure-ground Organisation
 Another principle coming from the Gestalt school of psychology is the concept
of figure-ground organisation.
o One part of the image is identified as the figure (the object we
focus on) while the rest of the details are given less attention
(falling into the background)
o At any given moment, you can perceive this as either the vase, or
two faces, but can also switch between them. But we can only
focus on one figure at a time.
o Believed to occur very early in visual processing (preceding object
recognition)

Evaluation Gestalt Visual Processing

 While Gestalt theory provides an explanation of many aspects of visual-
processing, like any theory it should be evaluated carefully.
o For example, gestalt argued that figure-ground organisation occurs
early in visual processing and proceeds object recognition.
o In a series of experiments, Grill-Spector and Kanwisher (2005) had
subjects perform figure-ground segregation tasks and object
recognition. Reaction times were almost identical, suggesting that
they actually occur concurrently.
o Extended by Mack et al. (2008) in an object detection (is it there?)
and object recognition (what is the object?) tasks, with visual
noise to increase difficulty.
o Results showed that under increased cognitive load, harder to
identify objects than perceive them, which would support Gestalt
claims that figure-ground occurs before object identification.
o i.e. Gestalt was right.

Pattern Recognition
 After we have processed and perceived visual information, comes a higher
order process of pattern recognition.
o We have to first recognise the pattern as something familiar
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 24

o Next we have to match the object to the correct concept and

classify it.
o Vital skill in everyday life, and we can learn a lot from those with
pattern recognition deficits.
 Neural networks need to be able to handle less than ideal conditions (partially
obscured, different angles, lighting conditions, distance, distracting stimuli, etc.)
o This allows for flexibility, and tolerance of ambiguity in stimuli.

Template theories versus Feature theories

 Two major perspectives for visual processing, templates and feature theories.
o Templates perspectives come from the Gestalt principle of
precepts (a recorded memory or template of the object)
o Feature theories propose that there are cognitive
processes/regions that detect perceptual elements (like horizontal
and vertical lines) that are then assembled into more complex
patterns.
o Just like with bottom-up and top-down processing, there is no
single 'right' perspective, and both types of pattern recognition are
used.

Template Theories
 Template theory suggests that for every type of object we can recognise, there
is a stored template or exemplar of what the object looks like (plus
accommodations for variations)
o Pattern is recognised on the basis of similarity to the template
o Exactly how is it stored? One way would be an idealised image
(e.g. Upright, face on, etc.) and that we mentally adjust for
variations
o Alternatively we might store many templates (top-down view,
front on, etc.) and many different variations.
o Goldstone (1998) argues that this is the case, and the brain is
capable of maintaining numerous templates.

Feature Detection theories

 Feature detection theories also offer an explanation for pattern/object
recognition.
o Neurological basis for much of feature detection, but also requires
exposure to perceptual stimuli.
o Critical age for environmental stimuli and development of feature
detection (e.g. Kitten experiment)
 Feature detection theory argues that lower level cognitive processes actively
seek out and 'detect' perceptual features, which flow upward to higher level
processes (Jainn & Dunn, 2004)

Horizontal line detected. Two vertical or diagonal lines detected.

2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 25

Support for Feature Theories

 Neisser (1964) compared search speed to locate the letter Z under different
distractor conditions.

Which of these lists is easier to search for it?

o For List 1, only straight lines are present (WEVX).
o For List 2, round objects are distractors (OUQD).
o When distractions were visually similar (overloading straight line feature
detectors), it took longer for List 1 than List 2.
o This supports feature detection theory.
 Feature detection generally focuses on local detail and argue its noticed first,
but in a classical experiment by Navon (1977) it was shown that we also use
global processing.

Figure 2.8 - The kind of stimulus used by Navon n(1977) to demonstrate the
importance of global features.
 Navon provided stimuli (like the picture above) and asked participants either to
identify the small letters (local) or the larger letter formed (global).
o Reaction time was faster for the large letters than the smaller
letters.
o Navon claimed we often see the forest (global structure) before
the trees (local).
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 26

Feature Detectors
 Nobel Prize winners Hubel and Wesel (1962) studied cells in occipital lobe
(visual processing) of the brain.
 Two types of neurons found - simple and complex cells.
 Simple cells respond to particular line orientations, and contrast of light/dark.
 Complex cells don't take orientation into consideration (so can handle different
angles) and then pass information up to higher level pattern-recognition
processes.

Approaches to perception
 Direct perception theories
o Bottom-up processing
o Stimuli flows to upper level processes and guides perception
o Pattern guides object recognition
 Construction perception theories
o Top-down processing
o People's thoughts, their expectations, guide how pattern is
interpreted.
 Most feature theories argue bottom-up processing.

Object Superiority Effect

 A challenge to bottom-up processing comes in the form of the object
superiority effect
o Objects are easier to detect when part of a 3D whole
o Weisstein and Harris (1974) had subjects detect the presence or
absence of a line by briefly flashing stimuli.
o According to feature detection theory, should always be found
o But subjects were better at detecting the line when it was part of a
more complex picture, suggesting that top down processing was
involved.
 Similar example is the word similarity effect
 Easier to detect a letter in a visual search task if part of a word (WORD or
WORK) than in isolation (D or K)

Other evidence for top-down processes

 Duck/rabbit illusion and how your perceptions/thoughts guide what you
interpret it as.
 Other examples include Rubin-Vase , or young-old woman illusion.
 Necker Cube Illusion
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 27

 Below is another example: The image is quite degraded, with noise and missing
parts.
o But does it make it easier for you to perceive if you were told that
it lives near water? Or quacks?
o Top down-processing guide perception

Visual Object Recognition

 One of the most important tasks in visual perception is object recognition -
identification of everyday objects and faces.
 Are we using bottom-up or top-down processes? Templates or feature
detection? A combination of all four perhaps?

Recognition-by-components theory
 Biederman (1987) argued that we analyse visual imagery by breaking it down
into component parts
o Start with basic feature detection of edges and lines (by
brightness, colour, texture)
o Next we work out which edges go together, and break a scene into
basic components (called geons)
o Examples include blocks, cylinders, spheres, arcs, etc.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 28

Figure 2.14 - Some examples of Biederman's geons

 According to Biederman, there are at least 36 basic objects.
o This doesn't sound like very much but,
o We can combine these to form far more complex parts, just like
the 44 phonemes in English make up hundreds of thousands of
words.
o According to Biederman, these more complex objects then form
templates in our long-term memory to compare against.

Figure 2.14 - Some common objects composed of Biederman's geons

 Evaluation:
o Geons are sufficiently varies (36+) that they can be used to
assemble larger more complex shapes (bit like lego blocks).
o Rather than using a static picture, by storing templates as geons
we can handle visual 'noise' and degraded signals.
o Geons can also be viewed from different angles, perspectives, and
so SHOULD be viewpoint-invariant
o But is it?
 The assumption that object recognition was viewpoint-invariant was tested by
Biederman and Gerhardstein (1993), in repetition priming experiment.
o Objects should be recognised more easily when primed (recent in
memory).
o Second time an object seen should have smaller reaction time.
o This was true even when viewpoint was altered by as much as 135
degrees.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 29

Challenges to viewpoint-invariance
 Other theorists (e.g. Friedman, et al. 2005) have challenged this viewpoint, by
pointing to evidence that objects are more quickly recognised from certain
perspectives/angles.
o Study found that in reaction time was faster in humans and in
pigeons, especially those angles we're most familiar with
o Compare those of a jeep-style car below.

o Now try and recognise it from this perspective.

Object recognition more complex than single approach

 Evidence from other theorists (e.g. Foster and Gibson, 2002) suggest that a
variety of processes are used rather than a single system.
 According to Tarr and Bülthoff (1995) viewpoint-invariant systems are used
when making easy perceptual decisions (for example: is it a cat or is it a dog?)
 But that more complex systems (viewpoint-dependent) are used when making
fine discrimination (for example: is it a Labrador or a Chihuahua?)
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 30

What happens when object recognition breaks down?

 We can learn a considerable deal about visual-processing through
understanding of visual deficits and disorders of object recognition.
 Visual agnosia is the term used to describe an object recognition disorder,
typically but not always due to brain damage
o Patients can see the object, would have an understanding from
prior experience what the object is and what it would be used for,
but have difficulty making the identification.
o Impairment varies on a continuum from mild difficulty to complete
inability.
 Two primary forms of visual agnosia:
o Apperceptive agnosia refers to impairment due to deficits in
perceptual processing
o Associative agnosia refers to deficits in retrieval from long-term
memory (where object templates are stored) rather than visual
system per se, or to derive meaning from them
 Refer to the textbook for a good summary. Apperceptive is going from
sensation to perception, whereas associate agnosia is in activating the meaning
of the input even thought there's a base level of recognition.
 How do we tell the difference? Refer to textbook notes.

Facial Recognition
 Most people have adequate face recognition skills, but I think we've all had
those moments where you know you recognise someone but can't quite place
them. Or the name is on the tip of the tongue.
o This is an example of a memory retrieval problem. You've
recognised the face, but can't recall their actual identity.
 Some people are exceptionally good at this task though, and can recognise
changes in age, hair colour etc. Others are relatively poor. What does this mean
for eyewitness recall?

Facial Recognition - Eyewitness Recall

 While we may be experts at recognising close friends, family, most people
aren't very good at recognising and then remembering strangers (particularly
under stressful situations, poor lighting, etc.)
 Davis & Valentine (2009) played moving video and asked participants to identify
and match particular faces. Even when very high quality close-up images used,
participants made a staggering amount of errors.
 Kemp (1997) provided college students with credit cards that used photo ID.
With a matching image, cards were accepted 93% of the time. But when they
used a card with someone who looked slightly similar to them, the incorrect
card was accepted 64% of the time!

Curious Case of Prosopagnosia

 Prosopagnosia is a more extreme form of the difficulty in recognising faces that
we all have from time to time.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 31

o https://www.youtube.com/watch?v=k5bvnXYIQG8
 Contrary to what the textbook suggests, it is not normally caused by brain
damager, but rather is a neurological condition that is hereditary.
o Most prosopagnosia cases show good object recognition, but are
just poor at recognising faces
o Approximately 2% of the population qualify for a diagnosis of
developmental prosopagnosia (Bate & Tree, 2017)
 Regions responsible for facial recognition differ from those involved in general
object recognition (hence single dissociation).
 Functional specialisation inn the fusiform facial area (FFA), also known as
Brodmann 37.

How do we know the location?

 Excellent opportunity to highlight to utility of fMRI in cognitive neuroscience.
 Downing et al. (2007) presented subjects with images of household objects
(tools, fruits, vegetables) in an object recognition task.
 As would be expected, wide diversity of brain regions were recruited. To
contrast this, subjects were also presented with pictures of faces, and the order
of tasks randomly selected so no clear pattern.
 When viewing just pictures of faces, much less neuronal activity was found, but
greatly increased activation of the FFA on both hemispheres of the brain.

An even more peculiar case of super-recognisers

 Just as some people have particular poor face recognition, there are some
gifted individuals that have exceptional face recognition ability.
o Russel et al. (2009) has reported a case study of four individuals,
who could even recognise people as adults from their childhood
photographs.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 32

 Just as with prosopagnosia, its likely that there's a strong genetic contribution.
Twin studies show that identical twins are more likely than fraternal twins to be
better/worse at face recognition.
 But there's also likely to be an environmental (learned) contribution, and
people can improve with training. Interestingly extraverts are generally better
at introverts (Lander and Poyarekar, 2015).
o Why? Likely that they pay more attention to faces, look at them
more often, suggesting a practice effect.

Brief Interlude: Visual Illusions

 One of the most common visual illusions we present to students is called the
Müller-Lyer illusion (see below).

Figure 2.26 (a) The Müller-Lyer illusion

 Most people say the line to the left is longer, but they're actually the same.
 Studies have found this replicated cross-culturally, and is also present in
primates (capuchins). Also greater susceptibility in females (Dewar, 1967)

Perception and Action

 According to Milner and Goodale (1998, 2008), we actually have two visual
systems:
o There is a vision-for-perception system used to identify objects
(fallible to optical illusions).
o Then there is a vision-for-action system used to deal with changing
stimuli such as a moving object that might be a threat.
 These can also be described as the 'what' (vision-for-perception) and the
'where' (vision-for-action) pathways, and are supported by evidence from
cognitive neuroscience.
 The 'what' route goes via the ventral pathway, while the where goes via the
dorsal pathway.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 33

Evaluating Milner and Goodale

 If Milner and Goodale are correct about there being two systems, then if we
were to invoke the vision-for-action system we should be able to reduce or
even eliminate visual illusions like the Müller-Lyer.
o A meta-analysis of 33 different studies by Bruno et al. (2008)
found that when subjects were asked to rapidly point at one of the
figures first, they were dramatically less likely to experience the
illusion (5.5% compared to 22.4% under control condition).
o This was because they were tracking movement (where), rather
than relying on the what pathway.

Inattentional Blindness
 As another way of illustrating the two systems, the topic of inattentional
blindness should be studies. When we're carefully paying attention to one part
of visual field, we can often miss important details.
 A study by Simon and Chabris (1999) examined inattentional blindness.
o Approximately 50% of people miss a very important detail,
illustrating just how susceptible to inattentional blindness.

Change Blindness
 However inattention is not always responsible for us missing information.
Another reason has been termed change blindness.
o Simons and Levin (1998) carried out a study in natural environments, where a
confederate asked a stranger for directions on a college campus.
o https://youtu.be/FWSxSQsspiQ
o They were briefly interrupted as two men carrying a wooden door
briefly come between the confederate and participant.
o A complete stranger (of different height, build and voice) steps
into place while the confederate uses the moment to leave. Over
half of the participants didn't pick up or register the change!
 However we greatly underestimate our susceptibility to inattention and change
blindness.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 34

 However change is also easier to recognise for important objects.

 Refer to the two examples from the textbook: in the first example, relatively
difficult to see as an obscure detail - average 10.4 seconds. In the second
object, given the importance and potential threat, average 2.6 seconds for
identification.

Inattention and Change Blindness

How might we seek to explain inattention and change blindness?
 Visual processing and attention are intrinsically linked.
 We have a relatively small amount of cognitive capacity, so pay attention to the
most salient details.
Rensink argues we detect changes only in the objects we're paying active attention to, and
miss changes in those we are not.
 Inn social situations we're often focused on other tasks, so attention divided.
 Alternate arguments are proposed inn the textbook though, such as decay of
visual memory.

Visual Damage and Blindsight

There is some evidence to suggest a difference between conscious awareness and
unconscious visual processing.
 This is exhibited through the curious phenomenon known as blindsight.
 Some people with damage to the primary visual cortex (V1) could still respond
to external stimuli (e.g. Catching a ball thrown at them) despite being unable to
see.
 Others can experience reactions when shown distressing video or happy video.
 Some parts of the two visual pathways are intact, even though can't fully
perceive or interpret.
2006PSY COGNITIVE PSYCHOLOGY STUDY NOTES 35