The work considers modification of the Best–Worst Scaling(BWS)to the so-called System 1(S1)approach.S1 was described by D.Kahneman as a spontaneous and automatic reaction by an unconsciousway in which human decision-...The work considers modification of the Best–Worst Scaling(BWS)to the so-called System 1(S1)approach.S1 was described by D.Kahneman as a spontaneous and automatic reaction by an unconsciousway in which human decision-makers choose among multiple alternatives.Application of S1 can be seen as a simplified BWS for data eliciting and express analysis of prioritization between many compared items.In S1,a respondent picks the items with which she feels“happy”,and those can be one,several,all,or none items in a task.Estimation of utilities is performed by multinomial-logit modeling with different optimization criteria which produce parameters of the models and choice probabilities of the items.Numerical examples by marketing research data are encouraging and demonstrating that spontaneous choice decisions can make S1 approach very fast,efficient,and convenient for express analysis of items prioritization,especially for big data.展开更多
文摘The work considers modification of the Best–Worst Scaling(BWS)to the so-called System 1(S1)approach.S1 was described by D.Kahneman as a spontaneous and automatic reaction by an unconsciousway in which human decision-makers choose among multiple alternatives.Application of S1 can be seen as a simplified BWS for data eliciting and express analysis of prioritization between many compared items.In S1,a respondent picks the items with which she feels“happy”,and those can be one,several,all,or none items in a task.Estimation of utilities is performed by multinomial-logit modeling with different optimization criteria which produce parameters of the models and choice probabilities of the items.Numerical examples by marketing research data are encouraging and demonstrating that spontaneous choice decisions can make S1 approach very fast,efficient,and convenient for express analysis of items prioritization,especially for big data.
