 |
A Computational Model of Political Cognition: The Dynamics of Candidate Evaluation
| |
| | Unformatted Document Text:
occur automatically, except for the contents of the goal buffer (Anderson, 2002). Thatis, while John Q. Public reads and processes a sentence ”Democrat Soandso supports pro-abortion”, what it is consciously aware of are only the concepts ”Democrat”, ”Soandso”,”supports”, and ”pro-abortion”. All other processes including the spreading of activationand the on-line updating process occur automatically.
The dual-process paradigm can be also modeled, following Overwalle & Siebler (2003),
by varying the attention levels given to incoming information. That is, when John Q. Publicprocesses the information that ”Democrat Soandso supports pro-abortion”, the activationlevels of corresponding chunks may be either low (low elaboration likelihood or peripheralprocessing) or high (high elaboration likelihood or central processing).
A Mechanism for Reading and Parsing Sentences via Procedural KnowledgeJohn Q. Public receives and processes external information by reading and parsing sen-tences (text strings). For example, the information ’Republicans oppose gun-control.’ ispresented to John Q. Public as a text string ”Republicans oppose gun control.” The modelthen parses the sentence word by word, retrieves associated concepts from memory, andupdates its knowledge and attitudes accordingly. As the model processes the sentences, allthe mechanisms discussed above are working in parallel. This reading mechanism is basedon a set of procedures which are essentially if-then rules. John Q. Public has about 40such built in procedures.
Computational Experiment - The Dynamics of Candi-date Evaluation
We conducted a series of computational experiments focusing the evaluation of politicalcandidate to examine how John Q. Public behaves as a voter. In this first set of tests,we focused on whether the model exhibits the following well-known empirical regularities;1) a cognitive and attitude priming effect (Neely, 1977; Fazio, 1986, 1990), where a unitof information in memory becomes more accessible and thus is retrieved faster when itis semantically associated or attitudinally congruent with a prime; 2) question order andwording effect in survey response (Tourangeau, Rips, & Rasinski, 2000), as when the orderand wording of questions affect survey responses in a systematic way; 3) on-line processing(Lodge et al, 1995; Lodge and Taber, 2000), where ordinary citizens can meaningfullyintegrate previously evaluated information into their current judgments without necessarilybeing able to recall it; 4) a dual-process model in which under some well-defined conditionssubjects sometimes deliberate carefully to arrive at an evaluation, but at other times givebut a cursory look at the information.
Experimental Setup
The computational experiments testing internal validity of of JQP consisted of three ex-periments that are identical except for the model’s initial belief system. Each of the three
9
|
| | Authors: Kim, Sung-youn., Lodge, Milton. and Taber, Charles. |
|
| |
|
|
occur automatically, except for the contents of the goal buffer (Anderson, 2002). That
is, while John Q. Public reads and processes a sentence ”Democrat Soandso supports pro-
abortion”, what it is consciously aware of are only the concepts ”Democrat”, ”Soandso”,
”supports”, and ”pro-abortion”. All other processes including the spreading of activation
and the on-line updating process occur automatically.
The dual-process paradigm can be also modeled, following Overwalle & Siebler (2003),
by varying the attention levels given to incoming information. That is, when John Q. Public
processes the information that ”Democrat Soandso supports pro-abortion”, the activation
levels of corresponding chunks may be either low (low elaboration likelihood or peripheral
processing) or high (high elaboration likelihood or central processing).
A Mechanism for Reading and Parsing Sentences via Procedural Knowledge
John Q. Public receives and processes external information by reading and parsing sen-
tences (text strings). For example, the information ’Republicans oppose gun-control.’ is
presented to John Q. Public as a text string ”Republicans oppose gun control.” The model
then parses the sentence word by word, retrieves associated concepts from memory, and
updates its knowledge and attitudes accordingly. As the model processes the sentences, all
the mechanisms discussed above are working in parallel. This reading mechanism is based
on a set of procedures which are essentially if-then rules. John Q. Public has about 40
such built in procedures.
Computational Experiment - The Dynamics of Candi-
date Evaluation
We conducted a series of computational experiments focusing the evaluation of political
candidate to examine how John Q. Public behaves as a voter. In this first set of tests,
we focused on whether the model exhibits the following well-known empirical regularities;
1) a cognitive and attitude priming effect (Neely, 1977; Fazio, 1986, 1990), where a unit
of information in memory becomes more accessible and thus is retrieved faster when it
is semantically associated or attitudinally congruent with a prime; 2) question order and
wording effect in survey response (Tourangeau, Rips, & Rasinski, 2000), as when the order
and wording of questions affect survey responses in a systematic way; 3) on-line processing
(Lodge et al, 1995; Lodge and Taber, 2000), where ordinary citizens can meaningfully
integrate previously evaluated information into their current judgments without necessarily
being able to recall it; 4) a dual-process model in which under some well-defined conditions
subjects sometimes deliberate carefully to arrive at an evaluation, but at other times give
but a cursory look at the information.
Experimental Setup
The computational experiments testing internal validity of of JQP consisted of three ex-
periments that are identical except for the model’s initial belief system. Each of the three
9
|
|
 Convention | | All Academic Convention makes running your annual conference simple and cost effective. It is your online solution for abstract management, peer review, and scheduling for your annual meeting or convention. | | Submission - Custom fields, multiple submission types, tracks, audio visual, multiple upload formats, automatic conversion to pdf. | | Review - Peer Review, Bulk reviewer assignment, bulk emails, ranking, z-score statistics, and multiple worksheets! | | Reports - Many standard and custom reports generated while you wait. Print programs with participant indexes, event grids, and more! | | Scheduling - Flexible and convenient grid scheduling within rooms and buildings. Conflict checking and advanced filtering. | | Communication - Bulk email tools to help your administrators send reminders and responses. Use form letters, a message center, and much more! | | Management - Search tools, duplicate people management, editing tools, submission transfers, many tools to manage a variety of conference management headaches! | | Click here for more information. |
|
|
|
| |
|
|
|
