The Multivarite Statistical Case Study Of College Students Participate In Science And Technology Competition Demand And Benefit

In this paper, we mainly use the stepwise regression of multivariate statisticalanalysis and Logistic regression as the research method, based on the students whoparticipated in the National Mathematical Modeling Contest from2008-2010threegrades. Through data collection, missing value analysis (EM Algorithm), variableselection (factor analysis and stepwise regression), data analysis, statistical modeling(logistic regression model), model test (goodness-of-fit testing and Wald significanttest) and many other steps to analyze the demand and benefit model of the studentswho took part in the Mathematical modeling contest.By analyzing the actual effects from the every demand of the students in theContest, we get the key factors and the non-significant factors. Then we develop thedemanding factors which should be mainly enhanced to enlarge the benefit scale ofthe Contest, namely the main safeguard could be provided and theappropriateinvestment could be reduced.Finally, we obtain that modeling training, team major, personal comprehensiveability, personal learning ability and mentor guidance have dominant effect to enhancethe benefit rate of the participants. Hence, we propose the university should intensifythe training extent of the students participating in the contest, and equip excellentmentor for guidance; the students themselves should also pay more attention to theircomprehensive ability and individual learning ability. Meanwhile, we encourage thegroup to meet professional diversification and strengthen team cooperation, to be getmore benefit from the contest, such as the contest prize and their own development.