Class Notes (1,100,000)
CA (650,000)
Brock U (10,000)
ASTR (200)
Lecture

ASTR 1P02 Lecture Notes - Unconscious Mind, Long-Term Memory, Tachistoscope


Department
Astronomy
Course Code
ASTR 1P02
Professor
Jenny Janke

This preview shows pages 1-3. to view the full 9 pages of the document.
I. Pattern recognition the way in which we conclude that some constellation of
stimulus features represents a single identifiable object. Actual objects. Most research
is done in the domain of vision. Most of it is done with psych 100 students, blank
computer screens with random objects popping up, headphones with random sounds.
All pattern recognition occurs in all of sensory modalities. Sound of baseball hitting
off of a bat, ice cubes dropping into a glass. Touch, smell, sound, taste. Process that
we do all the time, quickly, easily, accurately. The more quickly, easily, effortlessly,
accurately a human does something makes it really hard to study. A good research
strategy is to find people who can make mistakes in accuracy and effort = good
research. Look for situations in which this is difficult. One thing we could do is put a
letter on a computer screen for a brief period of time then take it off.
1. Cognition broadly conceived memory, decision making, thinking
2. Overwhelming majority of cognitive activity is unconscious. 95% of all
activity is unconscious.
3. Conscious mind relative to unconscious to really slow, inefficient and
primarily serial in nature can only do one thing at a time. Unconscious mind
is really fast and effect. parallel in nature, can do many things at once.
4. Cognitive psychology- 1 second is a very long period of time. Most measured
in milliseconds. 1/1000 of a second.
5. Tachistoscope (t-scope) controls exposure interval for looking at a stimulous.
A flap can block out the image after a few milliseconds. More commonly
done with a computer screen nowadays. Can be as small as 1 millisecond.
People commonly struggle with this task. A good strategy is to look for
situations in which people don’t do certain things easily, effortlessly,
accurately. Use little kids instead possibly. Want people to make mistakes.
II. When I show you a stimulus, you form a mental description of the stimulus, then you
compare that to all mental descriptions of all different objects that are stored in long
term memory. When you find a close match, then you have determined with the
stimulus is. What is a mental description? Form? How do we compare descriptions?
III. Template theories each class of objects has a image or template that corresponds to
it. unAnalyzed wholistic entities, not broken down to elementary components. When
comparing to stimulus look for a close match.
1. Point by point match literal point by point match
2. Pick highest matching score template that is closest to stimulus is chosen.
Highest score, with the most matches.
3. Consider no other information don’t do anything else, use only point by point
match.

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

IV. Major problems
1. Too sensitive to transformations, process that is too sensitive. Ex. A large E
compared to a small E.
2. Small details can have large effects. Some details are important, and some are not
important. Ex. O’s that are over drawn, missing a little section, a line in a place
that makes it look like a Q. The little line makes a big difference. Each cell in the
matrix is equally important to the other cells
3. Does not allow for alternative interpretations. Ex. Is mind twisters (2 images
within one).
4. No role of context really important, the perception of a stimulus depends on its
surroundings. Perception not made in isolation, but in a context.
10 56 47 13
D G F B
V. Possible solutions
1. Standardization (preprocessing), making the stimulus similar to the template.
Standardize the stimulus. A great big E cant be too big, unless you already know
what it is. If you know what it is, you don’t need to standardize it.
2. Multiple templates templates for every kind of stimulus
3. All for context compare representations of stimulus, unless there is contextual
interpretations saying otherwise. Context determines everything.
VI. Ironic that template theories have been developed as descriptors of human perception.
Not very good descriptors of human perception. Very good representation of machine
perception. Ex. Scantron grading machine with a master template for grading the right
answers. All machine perception requires stimulus to be controlled very precisely.
VII. PATIO
VIII. Feature theories allow us to describe patterns by listing their elementary properties.
Ex property list- listing of or decomposition of the stimulus into more elementary
feautres. When children learn what something is, they are simultaneously learning
what it is not. Learn that an F is not an E and an E is not an F. another example of this
is P vs R. the bottom line of the letter is the DISTINCTIVE FEATURE. You have to
learn what something is not, in order to learn what it is. In study with little kids, if
distinctive feature is highlighted, kids do a lot better versus no highlighted. If kids are
trained on highlighted, they do better on non-highlighted. If they are trained on non- h
and tested on non-h they don’t do that well.
1. Property list helps identify stimuli with stored descriptions in long term memory.
You chose the one that matches with the most properties. Details that correspond
to the features are very important, and vice versa.

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

2. Cognitive psychologists are very interested in creating process models- different
cognitive operations that occur over time. Process model- how the mind thinks.
Input properties matching description decision stage. Ex. A angles, lines
which letters have properties that match which letter has most matches? A.
whole process is unconscious. Stimuli are not single things, but collections of
features.
3. Physiological evidence. 2 researchers (Canadian) hugo weetzlow (sp?) won nobel
prize for discovery of cells in the visual cortex of cats that do not fire in response
to the direct stimulation of light. But do fire in response to patterns of light. The
patterns of light correspond to lines, angles, edges, ect. Perception of properties
are suggested to be hardwired into your system.
4. Nature of errors evidence. people can mistake objects for other things by
missing the distinct feature. When exposed to a letter for a few miliseconds, most
common mistakes are letters that are missing or have a distinctive feature from
one another. The more features stimuli have in common, the more likely they will
be mistaken for each other, and vice versa.
5. Stabilized image evidence experiment of measuring perception of distance of
light in a really dark room. Light never moved, but subject thought that light
moved. In book something phenomenon there are muscles that hold your eyes in
the sockets, and they are always vibrating. Images that you see are always
vibrating as well. There are drugs that will paralyze those muscles and cells can
get over stimulated with light, and you will not see the object anymore. The image
does not disappear all at once, but part by part. Proves that a bike is a collection of
attributes (features). That objects are made up of more than one part or pattern.
Stimuli are bundles of features.
6. Effects of similarity of nontargets research show a person a letter, then a 15
second pause, then show a group of letters and ask if it is there. can determine
how long it takes to search through one item in the display to get a match. When
looking for letters among like letters, it take a lot longer to do. When looking for a
letter among different letters, then its pretty fast. Different letters- round or
angular. Takes a lot longer because each letter is a bundle of features, especially
among like letters. Which is why the time is really short with different kinds of
letters.
7. All evidence points in the same direction. eyes pick out certain features. Can
predict mistakes by how similar objects are. Oversaturated cells make stimuli
disappear. When looking for one thing in context of similarity, it takes a long
time.
8. When we know something about an object, we don’t only know about the
features, but how the features are related to one another. The features must have
the proper relation to each other.
You're Reading a Preview

Unlock to view full version