Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
Categorization
Lecture 12 & Smith Reading
| Question | Answer |
|---|---|
| we code complex instances as | simple which reduces a wealth of particulars to a simple relation & frees mental capacities for other tasks |
| a category is | a class of objects that we believe belong together |
| there are an indefinite # of classes of objects whose members | don't seem to belong together where the class of objects has some properties in common yet the class is not treated as a category |
| taxonomy | a hierarchy in which successive levels refer to increasingly more specific objects |
| drawing an inference is | using existing beliefs to generate new ones |
| if an inference is deductive it is | impossible for it to be false if the old ones are true |
| if an inference is inductive it is | improbable for it to be false if the old ones are true |
| coding by category greatly reduces | the demands on perceptual processes, storage space, & reasoning processes |
| frequently used codes are associated with | brief descriptions |
| categories are often structures into a | taxonomy |
| categories at intermediate level are more likely to | be used that lower/higher levels |
| categorization of an object licenses | inductive inferences about the object |
| experimental demonstration by German & Marksman with subjects presented 3 pictures where | 3rd pic looks like one of first 2 but is from the same category as the other pic & new info given for 1st 2 pics and & ?'s about the 3rd subjects responded basing their decision on common category membership over physical similarity |
| different kind of categories differ in the extent | to which they support inductive differences |
| basic/subordinate categories support | more inferences because there is little differences between # of inductive inferences supported by basic/subordinate categories |
| natural kinds of categories deal with | naturally occurring species of flora/fauna |
| natural categories support more | inductive inferences about invisible properties |
| artificial kinds categories deal with | person-made objects |
| category members tend to be | physically similar to one another while being physically dissimilar from members of contrasting categories |
| at the superordinate level | members don't need to resemble eachother |
| at the subordinate level | members closely resemble each other but also resemble members of contrasting categories |
| at the basic level | members resemble each other & look different from members of contrasting categories |
| a basic category... | is often used to code experience, affords numerous inductive differences, & tends to maximize within-category similarity while minimizing between-category similarity |
| although a subordinate category may be used to code experiences in some contexts & it supports numerous inferences | it maximizes within category similarity at the cost of substantial between-category similarity |
| a superordinate category may be used to code experience but it | promotes few inductive inferences & doesn't maximize within category similarity |
| the 2 general approaches to measurement of similarity are | geometrical & featural |
| geometric approach is when | objects/items are represented as points in some multidimensional space such that the metric distance between 2 points corresponds to the dissimilarity between the 2 items |
| Shepard developed a systematic procedure where | a group of subjects rated the similarity between pairs of fruits & ratings were inputed into a computer that used an iterative procedure to position the items in a space with distance corresponding to the judged similarity |
| what are the 3 axioms to the geometrical approach | minimality, symmetry, & triangle inequality |
| the minimality axiom is that | the dissimilarity (distance) between any item & itself is identical for all items & is the minimum possible |
| the symmetry axiom is that | the distance between 2 items regardless of which item we start at is the same |
| the triangle inequality axiom is that | the shortest distance between 2 point is a straight line that can be represented by d(a,b)+d(b,c) >= d(a,c) if one concept is similar to a second concept, and the second concept is similar to the third, then the first and third must be reasonably similar |
| the geometrical approach has a history of success in representing | perceptual objects & less in conceptual items |
| Tversky produced evidence against | all 3 metric axioms for conceptual items |
| Tversky found that minimality axiom was compromised by | the fact that the more that we know about an item the more similar it is judged to itself |
| Tversky found that the symmetry axiom was undermined by | finding that an unfamiliar category is judged more similar to a familiar/prominent category than vice versa |
| another problem with the metric axioms is the notion of "nearest neighbor" | it is impossible for 1 item in a matrix space to be a nearest neighbor to so many other items as long as the space is of relatively low dimensionality & in a 2-D space the max # of items an item can serve as nearest neighbor is 5 |
| notion of a "nearest neighbor" is where | an item rated most similar to another item is its nearest neighbor |
| defenders of metric axioms argue that | violations of symmetry/triangle inequality arise more often when similarity is judged directly than indirectly which suggests that direct judgements require complex decision processes that are the source of asymmetries |
| the featural approach is when | an item is represented as a set of discrete features & similarity between 2 items is assumed to be an increasing function of the features they have in common & a decreasing function of the features they differ on |
| Tverskys contrast model is that | the similarity between sets of features characterizing item I & item j is given by sim(I,j)=af(I n j)- bf(I-j)-cf(j-i) |
| the violation of the triangle inequality model will be pronounced whenever | the weight given to common features a, exceeds that given to either set of distinctive features b or c because then similarity will be relatively large of r the first 2 pairs but not the 3rd |
| the contrast model is compatible with | the fact that x is more similar to y than vice versa as long as parameter b exceeds c & that a category can serve as a nearest neighbor to numerous instances |
| the limitations of the contrast model are that | it doesn't tell us what an items features are & doesn't offer any theory of the function that measures the salience of each set of features |
| people can reliable order | the instances of any category with respect to how typical/phototypical/representative they are of a category |
| evidence that categorization depends on | typicality in more naturalistic settings |
| the typicality of an instance is a measure of | its similarity to its category |
| the contrast model should predict typicality ratings by | 1. select domain of instances 2. estimate the features of the instances & the category 3. apply model to each instance-category pair 4.see whether this estimate of instance-category similarity correlates w/ rated typicality of the instance in category |
| Malt & Smith study where subjects had 90 seconds to list all the features they could think of for each instance found that the | success of contrast model in predicting typicality does not depend on whether a category is taken to be an abstraction or a set of instances |
| spreading activation is | when an item & category are presented, activation from these 2 sources begins to spread to the features associated with them, with the activation from each source being subdivided among its features |
| if the 2 sources of activation intersect at some (common) features then | further processing is undertaken to determine that an instances-category relation holds |
| there are more opportunities for an intersection with typical than atypical instances which leads to | more opportunities for an early termination of the process |
| 2 studies by Rips show that | categorization can be based on something other than similarity like variability & reasoning inductively |
| fuzzy set theory is | a generalization of traditional set theory that provides functions for relating membership in conjunctive sets (categories) to membership in simpler ones |
| even children notice there are essential features and children base categories on that. not just simple perceptual features & this is due to | essentialism |
| probabilistic categories are when properties/features are characteristic, not defining & | something belongs to a category if it is similar to members of that category & some members have more characteristic properties than others |
| people are faster to verify more | typical exemplars that less typical exemplars |
| how do we categorize using similarity? | if we agree that our categorization is probabilistic & some members are better members of a category than others than we decide using exemplars & prototypes |
| categorizing using exemplars involves | a stored example of a category in memory & categorizing new things based on similarity to stored exemplars |
Created by:
kzegelien2005
Popular Psychology sets