A Study On The Additivity Of The Support Theory Under The Effect Of Different Factors

Amos Tversky and Derek J. Koehler developed the Support Theory in 1994. The theory indicated that probability judgments depended not on events but on descriptions of events.This study is to investigate the probability judgment under uncertainty conditions based on the theory. The study used Mixed 3(unpacking) × 3 (question) × 3 (situation) × 2(subject) factorial design. The four independent variables are as follows: the unpacking manners consist of three levels, including packed, typical unpacking, atypical unpacking; the question categories consist of three levels, including professional questions, shopping questions, entertainment questions; the situation categories consist of three levels, including professional situation, shop situation, park situation; and the subjects consist of two levels, including office clerk and student. The packing methods and the question categories use within-subject design, and, the situation categories and the subjects categories use between-subject design. The dependent variable is the probability estimation for different experiment situations.By questionnaire surveying and data analysis, we get the results as follows:1. The main effect of unpacking manner is significant. The probability estimation for typical unpacking situation is higher than those for packed situation, and the probability estimation for packed situation is higher than those for atypical unpacking situation.2. The main effect of question categories is significant. The probability estimation for professional question is higher than those for entertainment questions, and the probability estimation for entertainment question is higher than those for shopping question.3. The main effect of situation categories is significant. The probability estimation for professional situation is higher than those for park situation, and the probability estimation for shop situation is higher than those for park situation.4. The main effect of subject categories is significant. The probability estimation for office clerk is higher than those for student under the same conditions.5. The following interactive effects are significant.a) The interactive effect between the question categories and the situation categories is significant. The probability estimation for the professional question under professional situation is the highest (M=53.51), and those for shopping question under park situation is the lowest (M=37.34).b) The interactive effect between the situation categories and the subject categories is significant. The probability estimation for office clerk under shop situation is the highest (M=50.56), and those for student under park situation is the lowest (M=41.49).c) The interactive effect between the unpacking manner and the question categories is significant. The probability estimation for professional question using typical unpacking manner is the highest (M=62.41), and those for shopping question using atypical unpacking manner is the lowest (M=32.4).d) The interactive effect between the unpacking manners and the situation categories is significant. The probability estimation for typical unpacking manner under professional situation is the highest (M=57.29), and those for atypical unpacking manner under park situation is the lowest (M=34.98).e) The interactive effect between the unpacking manners and the subject categories is significant. The probability estimation for office clerk using typical unpacking manner is the highest (M=57.56), and those for student using atypical unpacking is the lowest (M=33.31).