How To Make Inferences With Sequentially Updating Information:A Model Of Momentum Heuristic

People often need to make predictions about the future based on information that is constantly updating over time.For example,predicting winners and losers based on dynamic changes in game scores or predicting subsequent ups and downs based on dynamic changes in stock prices.The information is on the axis of time to form a sequence of continuously updating information,that is,Sequencially Updating Information.Such information often contains a momentum feature,which is shown as sequence fragments with the same result(e.g.,3 consecutive wins or 3 consecutive losses).In the past,researchers have found that individuals use momentum to make decisions(such as the hot hand effect or the gambler's fallacy),but there is a lack of quantitative analysis on how momentum affects decision behavior and prediction performance.In order to make up for the deficiency of previous researches,based on ecological rationality theory,this study systematically analyzed the characteristics of momentum,and constructed the mathematical model of the Momentum Heuristic.Through three studies,the use of momentum heuristic strategies by individuals and its influencing factors are investigated,as well as the impact of the use of momentum heuristic on prediction performance.In the first study,the mathematical models of the momentum heuristic and other competitive decision-making strategies are constructed.Through model fitting and comparison,the application of the momentum heuristic in the prediction of victory in snooker and the prediction of victory in quiz games is investigated.The results show that in two different prediction tasks,the vast majority of individuals will use the "Momentum-State Heuristic" to make decisions,and about 50%of individuals will statically use the momentum heuristic as the dominant strategy in the mixed heuristic.On the basis of Study 1 and combined with the behavioral experiment method,Study 2 investigated the influence of decision task difficulty and decision time pressure on the use degree of momentum heuristic strategy.The results show that time pressure is an important factor affecting the use of momentum heuristic.Compared with low time pressure,individuals use more momentum heuristics to make decisions under high time pressure.In the third study,a computer simulation method was used to investigate the relationship between individuals' use of momentum heuristic strategies and their decision performance.The results show that the degree of using the momentum heuristic strategy has a significant positive impact on individual decision performance.In addition,the third study also analyzed the relative competitive advantages of various strategies,and once again confirms that the momentum heuristic has good predictive effectiveness.To sum up,our research found that most individuals will use momentum heuristics(or mixed heuristics dominated by momentum heuristics)in the prediction task,and using momentum heuristics can produce better prediction performance.This result makes up for the defects of previous studies that failed to quantitatively analyze the influence of momentum characteristics on decision-making behavior and prediction performance,and provides a new paradigm and tool for the research of momentum feature and momentum heuristic.At the same time,the mixed heuristics revealed in this study also improve the ecological rationality theory's ability to explain individual predictive behavior,and make a contribution to the development of ecological rationality theory.In addition,the construction of the mathematical model of momentum heuristic and the analysis of prediction performance also provide research support for the use of momentum strategies to make various kinds of predictions in the era of big data,which has important practical significance.