从理性决策到行为决策

单击此处编辑母版标题样式,单击此处编辑母版文本样式,第二级,第三级,第四级,第五级,*,从理性决策到行为决策,王刊良,西安交通大学管理学院,2003-2005,管理,是,是什,么,么，,或,或不,是,是什,么,么,管理,不,不是,科,科学?Why?,管理,既,既是,科,科学,也,也是,艺,艺术?,存惑,与,与科,学,学精,神,神,科学,家,家因,为,为对,科,科学,的,的不,懈,懈追,求,求而,被,被迫,致,致死,？,？,中国,历,历史,上,上有,多,多少,？,？,西方,历,历史,上,上有,多,多少,？,？,苏格,拉,拉底,（,（Socrates,前,前,469,前,399,古,古,希,希腊,哲,哲学,家,家）,教育,不,不是,灌,灌输,，,，教,育,育是,点,点燃,火,火焰,。,。Educationisnotthefillingofavessel,butthekindlingofaflame.,科学,始,始于,观,观察,和,和实,验,验,FrancisBacon(,1561,-,1626,),anEnglishlawyer,statesman,essayist,historian,intellectualreformer,philosopher,andchampionofmodernscience.,“知,识,识就,是,是力,量,量”,“新,工,工具,”,”,“,“,归,归纳,法,法”,管理,研,研究,的,的不,可,可知,论,论,以人,为,为研,究,究对,象,象的,难,难点,，,，人,的,的动,机,机，,无,无损,观,观测,之,之难,庄子,与,与惠,子,子的,塬,塬上,之,之争,，,，“,子,子非,鱼,鱼，,安,安知,鱼,鱼之,乐,乐乎,”,”,不可,避,避免,的,的无,知,知（unavoidableignorance）,经济,研,研究,在,在获,得,得准,确,确的,数,数据,方,方面,更,更困,难,难,决策,管理,就,就是,决,决策,（,（H.Simon）,决策,的,的本,质,质在,于,于算,计,计不,同,同方,案,案的,损,损益,并,并进,行,行权,衡,衡,例,例,如,如成,本,本效,益,益分,析,析,决策,问,问题,的,的三,个,个特,点,点：,不,不确,定,定性,、,、主,观,观偏,好,好和,序,序惯,性,性(,路,路径,依,依赖),理性,决,决策,规范,性,性(prescriptionornormative),指定,的,的程,序,序和,方,方法(procedure),决策,问,问题,的,的描,述,述：,决策,目,目标(objectives),决策,准,准则(criteria),决策,方,方案(alternatives),决策,方,方法(tools),决策,程,程序(procedure/process),期望,效,效用,理,理论,(万人),方案,状态及概率,差(0.2),一般(0.4),好(0.3),很好(0.1),甲,电视广告,22,21,19,15,乙,报纸广告,18,17,16,15,丙,街头广告牌,13,16,16,16,受众数量,天气状况对广告效果(受众数量)有明显影响,天气不好(闭门在家看电视或读报的人多),天气晴朗(外出街头的人多),期望,效,效用,理,理论,方案,期望值计算,期望值,电视广告,0.2*22+0.4*21+0.3*19+0.1*15,20.0,报纸广告,0.2*18+0.4*17+0.3*16+0.1*15,16.7,街头广告,0.2*13+0.4*16+0.3*16+0.1*16,15.4,多目,标,标/,多,多准,则,则决,策,策,准则/目标,方案,价格,(万元),使用面积,(m2),到单位距离(km),房内设施,周围环境,甲,30,100,10,7,7,乙,25,80,8,3,5,丙,18,50,20,5,11,丁,22,70,12,5,9,偏好不同,权重不同,第三方,主观评分,a,ij,客观数据,a,ij,TOPSIS,TechniqueforOrderPreferencebySimilaritytoIdeaSolution,Step1.,决,决策,矩,矩阵,归,归一,化,化(normalization),TOPSIS,方案,价格(万元),使用面积(m2),到单位距离(km),房内设施,周围环境,甲,0.621,0.648,0.376,0.674,0.421,乙,0.518,0.519,0.301,0.289,0.301,丙,0.373,0.324,0.752,0.481,0.662,丁,0.455,0.454,0.451,0.481,0.542,Step1.,决,决策,矩,矩阵,归,归一,化,化(normalization),TOPSIS,方案,价格(万元),使用面积(m2),到单位距离(km),房内设施,周围环境,甲,0.0424,0.1370,0.0442,0.1864,0.1374,乙,0.0353,0.1096,0.0354,0.0799,0.0981,丙,0.0254,0.0685,0.0885,0.1331,0.2159,丁,0.0311,0.0959,0.0531,0.1331,0.1767,Step2.,决,决策,者,者的,偏,偏好(AHP),Step3.,加,加权,归,归一,化,化,价格(万元),使用面积(m2),到单位距离(km),房内设施,周围环境,权重,0.0682,0.2113,0.1177,0.2767,0.3261,TOPSIS,方案,价格(万元),使用面积(m2),到单位距离(km),房内设施,周围环境,甲,0.0424,0.1370,0.0442,0.1864,0.1374,乙,0.0353,0.1096,0.0354,0.0799,0.0981,丙,0.0254,0.0685,0.0885,0.1331,0.2159,丁,0.0311,0.0959,0.0531,0.1331,0.1767,正,0.0254,0.1370,0.0354,0.1864,0.2159,负,0.0424,0.0685,0.0885,0.0799,0.0981,Step4.,确,确定,正,正理,想,想方,案,案和,负,负理,想,想方,案,案,TOPSIS,方案,价格(万元),使用面积(m2),到单位距离(km),房内设施,周围环境,甲,0.0424,0.1370,0.0442,0.1864,0.1374,乙,0.0353,0.1096,0.0354,0.0799,0.0981,丙,0.0254,0.0685,0.0885,0.1331,0.2159,丁,0.0311,0.0959,0.0531,0.1331,0.1767,正,0.0254,0.1370,0.0354,0.1864,0.2159,负,0.0424,0.0685,0.0885,0.0799,0.0981,Step5.,计,计算,各,各方,案,案到,正,正理,想,想方,案,案和,负,负理,想,想方,案,案的,距,距离,方案,甲,乙,丙,丁,正,0.0808,0.1614,0.1017,0.0801,负,0.1398,0.0675,0.1304,0.1055,TOPSIS,Step6.,计,计算,各,各方,案,案到,正,正理,想,想方,案,案的,相,相对,接,接近,程,程度,方案,甲,乙,丙,丁,正,0.0808,0.1614,0.1017,0.0801,负,0.1398,0.0675,0.1304,0.1055,0.3664,0.7052,0.4383,0.4315,甲 丁 丙 乙,期望,效,效用,理,理论,p,1-,p,C,1,C,2,L,期望,效,效用,值,值为,u,(,L,)=,pC,1,+(1-,p,),C,2,理性,假,假说,的,的局,限,限性,有限,理,理性(boundedrationality),人类,获,获取,决,决策,所,所需,全,全部,信,信息,的,的局,限,限性,人类,处,处理,信,信息,的,的能,力,力的,局,局限,性,性,满意,决,决策(satisfyingdecision)vs.,最,最优,决,决策(optimaldecision),有效,市,市场,假,假说(efficientmarkethypothesis),信息,不,不对,等,等,逆向,选,选择,和,和败,德,德行,为,为,理性,并,并不,理,理性,：,：BF,实,实验,Beeandfly：,把,把几,只,只蜜,蜂,蜂和,几,几只,苍,苍蝇,撞,撞在,一,一个,玻,玻璃,瓶,瓶中,，,，然,后,后将,瓶,瓶子,平,平放,，,，让,瓶,瓶底,朝,朝向,窗,窗口,，,，结,果,果会,怎,怎么,样,样？,蜜,蜜蜂,会,会不,停,停地,在,在瓶,子,子底,部,部寻,找,找出,口,口，,直,直到,累,累死,；,；而,苍,苍蝇,在,在不,到,到两,分,分钟,内,内全,部,部逃,出,出。,蜜蜂喜欢光,亮,亮而且有智,力,力，坚定地,认,认为出口在,光,光线最亮的,地,地方（实验,人,人员的臆测,）,）；苍蝇则,对,对事物的逻,辑,辑不在意，,到,到处乱飞。,理性并不理,性,性：选举,A/B/C,三,三候选人，17人投票,第一选择：A/6人，B/6人，C/5人,在首选B和C的选民中,分,分别有2人,和,和3人把A,作,作为第二选,择,择,复选制,每位选民只,投,投一票，当,无,无人得票过,半,半时，在最,高,高票的两人,中,中进行复选,A(比B多1票),假设第一次,选,选举前，A,展,展示了一件,英,英雄行为（,如,如救了一位,落,落水儿童）,，,，使B的部,分,分支持者（,即,即对A也有,好,好感的那2,个,个人）在最,后,后一分钟决,定,定改投给A,。,。结果是A,的,的8票没有,过,过半，B少,了,了2票遭到,淘,淘汰；,在复选中，A仍维持8,票,票，C则因,获,获得了B选,民,民的4票后,来,来居上，赢,得,得了最后胜,利,利。,理性并不理,性,性：阿莱悖,论,论(Allais Paradox),France Economist,1988,NobelLaureate,方案A,100%肯定能赢得100万元,方案B,10%的概率赢得150万元,89%的概率赢得100万元,1%的概率没有盈亏,u,(100)0.1,u,(150)+0.89,u,(100)+0.01,u,(0),方案A,11%肯定能赢得100万元,89%概率没有盈亏,方案B,10%的概率赢得150万元,90%的概率没有盈亏,0.1,u,(150)+0.9,u,(0)0.11,u,(100)+0.89,u,(0),0.11,u,(100)0.1,u,(150)+0.01,u,(0),0.11,u,(100)0.1,u,(150)+0.01,u,(0),Daniel Kahneman,瑞典皇家科,学,学院:Princeton Univ.的,心,心理学教授,丹,丹尼尔卡,尼,尼曼(DanielKahneman)因,为,为“,将来自心理,研,研究领域的,综,综合洞察力,应,应用在了经,济,济学当中,尤,尤其是在不,确,确定情况下,的,的人为判断,和,和决策方面,作,作出了突出,贡,贡献,”,摘得,2002,年度诺贝尔,经,经济学奖的,桂,桂冠.,for havingintegrated insightsfrompsychologicalresearch into economic science,especially concerning humanjudgment and decision-making underuncertainty.,AmosTversky,传统经济学,一,一直以“理,性,性人”为理,论,论基础,通,过,过一个个精,密,密的数学模,型,型构筑起完,美,美的理论体,系,系.而行,为,为经济学研,究,究则从实证,出,出发,从,人,人自身的心,理,理特质、行,为,为特征出发,去揭示,影,影响选择行,为,为的非理性,心,心理因素.,70年代,卡,卡