Table 102

The attributes used to characterise the individual members of two gangs, the Jets and Sharks, from McClelland (1981)

Person

Name

Age

Education

Marital status

Job

Gang

_Art

Art

40s

junior high

single

pusher

Jets

_Al

Al

30s

junior high

married

burglar

Jets

_Sam

Sam

20s

college

single

bookie

Jets

_Clyde

Clyde

40s

junior high

single

bookie

Jets

_Mike

Mike

30s

junior high

single

bookie

Jets

Jim

Jim

20s

junior high

divorced

burglar

Jets

_Greg

Greg

20s

high school

married

pusher

Jets

The attributes used to characterise the individual members of two gangs, the Jets and Sharks, from McClelland (1981)

Person

Name

Age

Education

Marital status

Job

Gang

John

John

20s

junior high

married

burglar

Jets

Doug

Doug

30s

high school

single

bookie

Jets

Lance

Lance

20s

junior high

married

burglar

Jets

George

George

20s

junior high

divorced

burglar

Jets

_Pete

Pete

20s

high school

single

bookie

Jets

Fred

Fred

20s

high school

single

pusher

Jets

Gene

Gene

20s

college

single

pusher

Jets

_Ralph

Ralph

30s

junior high

single

pusher

Jets

_Phil

Phil

30s

college

married

pusher

Sharks

_lke

Ike

30s

junior high

single

bookie

Sharks

_Nick

Nick

30s

high school

single

pusher

Sharks

_Don

Don

30s

college

married

burglar

Sharks

_Ned

Ned

30s

college

married

bookie

Sharks

_Karl

Karl

40s

high school

married

bookie

Sharks

_Ken

Ken

20s

high school

single

burglar

Sharks

Earl

Earl

40s

high school

married

burglar

Sharks

_Rick

Rick

30s

high school

divorced

burglar

Sharks

_Ol

Ol

30s

college

married

pusher

Sharks

_Neal

Neal

30s

high school

single

bookie

Sharks

Dave

Dave

30s

high school

divorced

pusher

Sharks

can "clamp" a particular person-node (e.g., _Art); clamping is a term used to mean that you keep the activation in this node at a constant level on each cycle—you fix its activation at a given level so it does not change. When this is done, after a number of cycles all of Art's attributes will have high activations, indicating that these are his attributes. Furthermore, because activation passes out from some of these nodes to other instance nodes, all the individuals that share key attributes with Art will be activated to varying degrees. So, for example, the person nodes for _Clyde and _Ralph will also have high activations reflecting their similarity to _Art. More interestingly, the network can form a "generalisation" about the key attributes of different classes of individuals. If we clamp the Jets node, then when the network settles we have a weighted attribute description of the typical attributes of Jets gang-members; namely, that they tend to be single, are in their 20s, and have only been to junior high-school. A generalisation for the Sharks gang can also be determined in this way.

The important point about this model is that you do not have to explicitly encode general knowledge, rather this knowledge emerges from a collection of instances over which certain computations are carried out. Thus, this is one way of implementing exemplar-based models of categorisation reviewed earlier in this chapter (see e.g., Hintzman, 1986; Kruschke, 1992; Nosofsky, 1991).

Business Correspondence

Business Correspondence

24 chapters on preparing to write the letter and finding the proper viewpoint how to open the letter, present the proposition convincingly, make an effective close how to acquire a forceful style and inject originality how to adapt selling appeal to different prospects and get orders by letter proved principles and practical schemes illustrated by extracts from 217 actual letter.

Get My Free Ebook


Post a comment