1. What is Cognitive Psychology?
Hit the space bar for next slide
Slide Shows in Your Browser
- Works across browsers and is operated like PowerPoint
- Advance to next slide with mouse click or space bar
- Move forward/backward between slides with Cursor Left,
Cursor Right, Pg Up and Pg Dn
keys
- Home key for first slide, End
key for last slide
- The "C" key for an automatically generated
table of contents (or click on "contents" on the toolbar)
- Function F11 to go full screen and back
- The "F" key toggles the display of the footer
- The "A" key toggles display of current vs all slides
- Use the "A" key when you want to view or print all slides
- Try it now to see how to include notes for handouts (this is
explained in the notes following this slide)
- Font sizes automatically adapt to browser window size
- use S and B keys for
manual control (or < and >, or the - and
+ keys on the number pad
- Switching off JavaScript reveals all slides
Course structure
- What is Cognitive Psychology?
- Attention and Focus 1: focused attention
- Attention and Focus 2: split attention
- Object recognition 1: recognising patterns & words
- Object recognition 2: recognising auditory stimuli
- Cognitive Representations: Paivo’s dual-route hypothesis
Aims of the course
Demonstrate that cognitive psychology is an approach, not a specific set of experiments
Show you how different types of evidence are used to evaluate theories, in cognitive and other branches of psychology
Illustrate with evidence from four major domains (attention, vision, language, knowledge representation)
History
- 1879
- Establishment of first psychology laboratory (George Wundt, Leipzig, Germany) – Structuralism
- 1890
- Armchair psychology (e.g., James, 1890)
Problems with the early - introspective - methods
Both structuralism and functionalism referred to mentalistic contents of mind that could not be directly observed.
Participants (or researchers) are not always aware of the procedure they follow to perform a task (can you tell me how you do to remember something?).
Behaviourist’s reaction
...to stick to the only thing we can study objectively: the behaviour that follows an input.
...and to avoid to introduce mental variables (or unseen variables) to explain behaviour.
Behaviourist Psychology
1920s
Behaviourist Psychology (derived from learning theory)
“[The behaviourist] dropped from his scientific vocabulary all subjective terms such as sensation, perception, image, desire, purpose, and even thinking and emotions as they were subjectively defined.” (Watson, 1930)
Black box metaphor (Skinner?).
Times of change
Von Neuman (1950)
Computer: Representations and processes (memory, processing system) between input (key presses) and response (screen display).
Computers:
- Process information in complex ways
- Store lots of information in memory
- Retrieve information from memory and use it
- Have an input and an output
- Have hardware and software
- Have limited capacity
Information Processing Theory
Shannon (1952)
- Mental processes occur in steps or stages.
- Each stage has an input and an output.
- Each stage transforms the output of the previous stage in some way.
- Each stage has a minimum duration.
- Each stage has a limited capacity.
- Serial or parallel processing is possible.
Birth of Cognitive Psychology
1956
Chomsky (preliminary paper on theory of language)
Miller (“The magic number seven, plus or minus two”)
Newell and Simon (human problem solving)
Also Broadbent (1955), human factors
Cognitive Psychology
Behaviourism (1910s-1950s)
Cognitivism (from 1950s)
General framework
Model of human information processing (Wickens, 1992)
Key concepts
The Cognitive Psychology approach is concerned with:
- the ways in which information is represented in the mind
- and with the processes which act on that information
- It does not need to be concerned with the conscious level (the one accessible to introspection)
It typically uses an information processing model (expressed as a sequence of processing stages) to predict the time or accuracy of a decision.
Information Processing Analysis Example
Source: http://psych.colorado.edu/~tcurran/psyc2145/lectures/Lec_01_16.pdf
Methods
1. Empirical methods
“Cognitive processes and structures are inferred from participants’ behaviour (typically speed and/or accuracy of performance) obtained under well controlled conditions”
Methodology
- Cognitive psychologists have developed techniques for inferring properties of processing by analyzing relative response times, error rates, or types of judgments.
- These methodologies are used to test Hypotheses derived from theories against data.
- For example, if there is a discrete module that compares input to lists stored in short-term memory, then it should be possible to describe the sequence of steps or stages through which processing is accomplished
Sternberg Experiment
Saul Sternberg (1966) proposed a method of studying how people search short-term memory (STM) to determine whether certain information is present.
It is one of the 'classic' examples of the information processing paradigm in cognitive psychology.
See http://web.uct.ac.za/depts/psychology/psy300/sternb.html for a description.
Reference: Sternberg, S. (1969) Memory-scanning: Mental processes revealed by reaction-time experiments. American Scientist vol.57, 421-457.
Information Processing Analysis Example
(In this condition, digits rather than letters)
Sternberg (1966, 1969) developed a procedure that permits a test of two questions about the nature of the search of STM.
- whether the contents of STM are searched all at once (parallel search) or one item after another (serial search).
- whether the search stops when the item searched for is found (self-terminating search) or whether all items in STM must be compared to the item searched for (exhaustive search).
Sternberg's (1969)
What do you expect?
- Parallel or serial search?
- Exhaustive or self-terminating search?
Sternberg's (1969) findings
What do these data show?
- Parallel or serial search?
(look at positive set, and how RT changes with increasing set size)
- Exhaustive or self-terminating?
(contrast positive and negative sets)
Sternberg's (1969) findings
- 38 msec per digit
=> Serial search
- Similar RT (reaction times) for positive and negative answers
=> Exhaustive search
Sternberg's (1969) model
Limitations
- Indirect information about the internal representations and processes.
- Artificial set-up, that rarely corresponds to real-life situations.
1. Cognitive Neuropsychology
“studies the performance of brain-damaged patients to infer the mechanisms involved -- in normal cognitive functioning”
Cognitive Neuropsychology
Based on a Modularity assumption: distinct systems which can suffer damage separately from each other.
Concept of Double Dissociation
Example: Semantic vs syntax: Are the processes responsible for generating meaning independent of those responsible for the structure of sentences
Broca’s aphasia
Difficulty speaking (telegraphic speech) with poor syntax (“agrammatical aphasia”) but semantically appropriate words.
Wernicke’s aphasia
Speech is grammatical, but meaningless
Limitations
- It is difficult to carry out group study (when patients are group according to a syndrome, there are variations in performance, in the group).
- The strong locality (or modularity of the brain) assumption is not well supported by data. It is not clear that damage to one module “has” only local effect. It is not clear that areas of the brain are fully specialized.
3. Computational models
(mathematical equations, computer programs, connectionist networks)
Box and Arrows Models
Single reaction time model (Newell)
Limits:
- Flowcharts models are never specific enough.
- "What happens in the boxes?”
- “What do the arrows do?”
- “How can the brain be organised like that?”
Implementing a theory as a program is a good method for checking that it contains no hidden assumption or vague terms.
A computer model forces to clearly specify the format of the representations and the nature of the processes.
Production systems as computer programs
- The mind is a seen as a symbol manipulation device.
- Note that the speed of the program is not directly compared to the participants’ performance.
- Rather, difference in conditions (Task A takes longer than Task B, for both the model and the participants)
Connectionist networks
1. all input neurons are connected to all output neurons
2. If units A and B and C are simultaneously active, the strength of the connection between them will increase
(and B will be triggered faster, next time A is presented)
Connectionist breakthrough: NETtalk
Sejnowski & Roseberg (1986)
Problem of the level of description
Three levels of description (David Marr)
- Computational level : what is computed and why, abstract description of the set of procesing modules required to solve the computation (flowchart, schematic representation).
- Algorithmic level : the procedure and representations used. Theorists attempt to discover the way in which processing is actually accomplished within each box (cf. Software program)
- Implementational level : the physical instantiation (the brain, the computer)
=> What level matters most between Algorithmic and Implementational?
4. Cognitive Neuroscience
“[Attempts] to establish where in the brain certain cognitive processes occur, and when these processes occur… with a tendency to combine functional and physiological concepts”
Neuroscience
Localisation of brain functions in vivo.
Conclusion
The idea is to look for converging evidence from different approaches
Summary
- Cognitive psychologists study how information is represented in the mind, and how those representations are processed
- This approach lets us make and test theories about the workings of the mind
- These processes don't have to be conscious
- Neuropsychological evidence may provide insights into the ways these processes interact
Summary (cntd)
- There are various ways of modelling information processes; either as a computer program (theory checking) or as a more brain-like neural net
- Different types of models may simply describe the same processes at different levels of description
