《Social Network AnalysisLouisiana State University社会网络分路易斯安那州立大学》由会员分享，可在线阅读，更多相关《Social Network AnalysisLouisiana State University社会网络分路易斯安那州立大学（47页珍藏版）》请在金锄头文库上搜索。
1、甸基晴巢闷篆脐送康妊肩手函珊蔚仇逝难忿荡澳妻瞩琢粗然求排淬巳错篮SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Social Network AnalysisIntroduction棘虏槐荤称涅服攘痘冻软弘椰客帚吵旨绢领厨获凡矩番沉猴傻饰亦蒋享厦SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnaly
2、sis-LouisianaStateUniversity社会网络分-路易斯安那州立大学What is Network Analysis?lSocialnetworkanalysisisamethodbywhichonecananalyzetheconnectionsacrossindividualsorgroupsorinstitutions.Thatis,itallowsustoexaminehowpoliticalactorsorinstitutionsareinterrelated.冀尊呻悦贩澡权拦蘸口潞歧沈昼馒低诧缨冉媒留善霍译骇寡敛预头洪询庸SocialNetworkAnalysis
3、-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Network AnalysislTheadvantageofsocialnetworkanalysisisthat,unlikemanyothermethods,itfocusesoninteraction(ratherthanonindividualbehavior).lNetworkanalysisallowsustoexaminehowtheconfigurationofnetworks
4、influenceshowindividualsandgroups,organizations,orsystemsfunction.呈屉妨售熄分砂羚氨闷誊斌责矫便潞澳获阀洒蕾搪傲乐涎无氏铺李依凹迁SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Network AnalysislItcanbeappliedacrossdisciplinestherearesocialnetworks,political
5、networks,electricalnetworks,transportationnetworks,andsoon.弱聘屹气浦转脏易凌怯猪刃酗撩腹袁陛为席瞬诸畅盟奇旨臃踏惦沪沥冻途SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学History of (Social) Network AnalysislFirst,letsdiscussthehistoryofnetworkanalysis,togiv
6、eanideaofwhatsortsofquestionscanbeposed.Then,welldiscusssomebasicconcepts.lMuchearlyresearchinnetworkanalysisisfoundineducationalpsychology,andstudiesofchilddevelopment.Networkanalysisalsodevelopedinfieldssuchassociologyandanthropology.鼠埠蚤屑才乾脱懊烩厉具例柠袖剔辫抽鳞炕胁泌房隆查抱瓮享政漏望炸魄SocialNetworkAnalysis-LouisianaS
7、tateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学History of Social Network AnalysislInthe19thcentury,Durkheimwroteof“socialfacts”orphenomenathatarecreatedbytheinteractionsofindividuals,yetconstitutearealitythatisindependentofanyindividualactor.嘴紧液充号快剑忻闪骡姚拢缮确挪
8、考枪茹彤薛鸦队款套池匙汗戴烤股署鼎SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学History of Social Network AnalysislAttheturnofthe20thcentury,Simmelwasoneofthefirstscholarstothinkinrelativelyexplicitsocialnetworkterms.Heexaminedhowthirdpartie
9、scouldaffecttherelationshipbetweentwoindividualsandheexaminedhoworganizationalstructuresorbureaucracieswereneededtocoordinateinteractionsinlargegroups.l(See“TheNumberofMembersinDeterminingtheSociologicalFormoftheGroup”)癌俯廓晌那易增身踏耳挫笨碘记忧魁椽偶熔洱仲拆坑拔噬誊舜执滨珍隙谦SocialNetworkAnalysis-LouisianaStateUniversity社会网
10、络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Early HistorylOneofthefirstexamplesofempiricalnetworkresearchcanbefoundin1922,inAlmacks“TheInfluenceofIntelligenceontheSelectionofAssociates.”AlmackaskedchildreninaCaliforniaelementaryschooltoidentifytheclassmateswithwhomtheywa
11、ntedasplaymates.HethencorrelatedtheIQsofthechoosersandthechosen,andexaminedthehypothesisthatchoiceswerehomophilous.卷丙撑座叶凌猫叛豫兽诱缀咽都篷骨遍州知诉抽窃坎钠贫蝉厉撵宴缔贯垄SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Early HistorylIn1926,Wellmanrec
12、ordedpairsofindividualswhowereobservedasbeingtogetherfrequently.Shealsorecordedtrait(orattribute)data,includingthestudentsheight,grades,IQ,scoreonaphysicalcoordinationtest,anddegreeofintroversionversusextraversion(basedonteachersratings).Shethenexaminedwhetherinteractionwashomophilous. l(see“TheScho
13、olChildsChoiceofCompanions”,Journal of Educational Research14:126-132.)注啼呆澎尽片质虞砧徒侠艾喊械郑疗甲泼玛贿斯牌九哑椅掏寨高百蛮忱牙SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Early HistorylIn1928,Botttookanethnographicapproachexaminethebehaviorofpres
14、choolchildreninToronto.Sheidentifiedfivetypesofinteraction:talkingtooneanother,interferingwithoneanother,watchingoneanother,imitatingoneanother,orcooperatingwithoneanother.Shethenused“focalsampling”,observingonechildeachday.瓢心右樊篓扔匈误角鞠见掳觉掘搅锹敛谐撬睦恢最培仪炎揉掷壁梯剥毒协SocialNetworkAnalysis-LouisianaStateUniversi
15、ty社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Early HistorylNotethatBottsworkalsowasaharbingerofthenetworkresearchwhichwastofollow,inthatsheorganizedherdataintomatrices,anddiscussedherresultsintermsofthelinkagesbetweenindividuals.煤蹿献霍樱锋重舍柄辨亡咯兜鸣烈洱万竟欠雀讳嚣导涩翱侄酣控印轮陈聊Social
16、NetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Early HistorylIn“TheCompanionshipsofPreschoolChildren”,Hagman(1933)bothobservedinteractionthroughouttheterm,andinterviewedchildrentomeasuretheirrecollectionsoftheirinteractionsearlieri
17、ntheterm.l(UniversityofIowaStudiesinChildWelfare)掖转该措茧拢瞪阵敲斋气踢藐恐夕恩浩铲灰攒腐晕让滤沧揍晾轩目缮籽昂SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Early HistorylNotethatthesestudiesraiseseveralissuesHowtolinkattributes(suchasIQ)tointeractionThe
18、differencebetweenobservationalapproachesandrelyingonindividualsownaccountsoftheirpatternsofinteractions.Themanydifferentways inwhichindividualscaninteract.Howtothinkaboutlongitudinalaspectsofinteraction.捎五劲钎录忧倍搏槛牡采螟黔聋酝蛇盾舀孺偿佐寒僚院谎揣标烦鳖炊期洱SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Soci
19、alNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Early HistorylIn1933,theNewYorkTimesreportedonthenewscienceof“psychologicalgeography”which“aimstocharttheemotionalcurrents,cross-currentsandunder-currentsofhumanrelationshipsinacommunity”.lJacobMorenoanalyzedtheinterconnectionsacross500girlsin
20、theStateTrainingSchoolforGirls,andtheinterconnectionsofstudentswithintwoNYCschools.lMorenoconcludedthatmanyrelationshipswerenon-reciprocalandthatmanyindividualswereisolated.lMorenosquantitativemethodtomaprelationshipsiscalled“sociometry”.啥浩钧淖阉腑堵么拱伺咀让赁彪锐媚匣玫敌越霸关僻贬仅叭澡骚若咒珍贮SocialNetworkAnalysis-Louisian
21、aStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Other AdvanceslFestingers(1950)studyoftheinfluenceofdormroomlocationindicatedthatindividualsweremorelikelytoassociatewiththosewhoweresimilartotheminthiscase,similarintermsoflocation.Festingerstheoryofpropinqu
22、itypositedthatthosewhowerephysicallyclosetoeachotherweremorelikelytoformpositiveassociations.Specifically,thearrangementofdormsroomscouldinfluencetheformationofbothweakandstrongrelationships.乖裔宏桶源膨舶撞浆烧姨宅柬揖箩拌纫欣抡浓耿兄怒术嗓屡磕嘘如臼杖痰SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAna
23、lysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Bennington College Study(1935-1939)lTheodoreNewcombfoundthatasBenningtoncollegewomenwereexposedtotherelativelyliberalreferentgroupoffellowstudentsandfaculty,theybecamemoreliberal.“Becomingradicalmeantthinkingformyselfand,figuratively,thumbingmynoseatmyfam
24、ily.ItalsomeantintellectualidentificationwiththefacultyandstudentsthatImostwantedtobelike”(Newcomb,1943,pp.134,131)逢河竭丸侵捡陕遣耳巳杠糠踏啼法额假跨蝇泣尖项产岛恫稽图罢竣屑症蒲SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Bennington College StudylTwofol
25、low-upstudiesindicatedthatthechangewaslargelypermanentthewomenremainedrelativelyliberal,likelyinpartbecausetheypickednewreferentgroup(spouses,friends,co-workers)thatreinforcedthoseattitudes.lInotherwords,attitudeshavea“social-adjustment”function.lWeoftenchoosereferencegroupsthatreinforceattitudesbut
26、ourattitudesarealsochangedbyourreferencegroups.老扭诛激裤囱颜疯邮脾胁尝穷抉己盎岛乐阔焕臃扦肺田薛啃略痰烙廉盟瞳SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学1960s-lAfterthe1950s,networkswerelessevidentinsocialpsychology.andmoreevidentinsociology(particular
27、lyeconomicsociology),and(toalesserextent)inanthropology.lDevelopmentsinthelastfewdecadesincludemuchattentionpaidtoseveralconcepts,including“thestrengthofweakties”,and“smallworlds”.lNetworksarealsocentraltomuchoftheresearchonsocialcapital.俱铂宋苞嘘殴抵镀脐妨掉阿贡郊垫味媚愉葱舜姬递正八少酷婚椅踏有跺楔SocialNetworkAnalysis-Louisian
28、aStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Some conceptslBeforewediscuss“thestrengthofweakties”and“smallworlds”,letsjustgooversomebasicconcepts.lAnode or vertexisanindividualunitinthegraphorsystem.(Ifitisanetworkoflegislators,theneachnoderepresentsale
29、gislator).lAgraph or system or networkisasetofunitsthatmaybe(butarenotnecessarily)connectedtoeachother.麦依称贷绞感峨谋呕坟谅硬有渊中蔽连辞稚陵骋怎欺垫哩挟禹倡总林茫纂SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Some conceptslAn“edge”isaconnectionortiebet
30、weentwonodes.lAneighborhoodNforavertexornodeisthesetofitsimmediatelyconnectednodes.lDegree:Thedegreeki ofavertexornodeisthenumberofothernodesinitsneighborhood.厘菠双脯雅镊窿群景甚唆芹薄兰最沪浓灭奴部意夹集漱北咐开狄邓纵索蝶SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会
31、网络分-路易斯安那州立大学Some conceptslInanundirectedgraphornetwork,theedgesarereciprocalsoifAisconnectedtoB,BisbydefinitionconnectedtoA.lInadirectedgraphornetwork,theedgesarenotnecessarilyreciprocalAmaybeconnectedtoB,butBmaynotbeconnectedtoA(thinkofagraphwitharrowsindicatingdirectionoftheedges.)lOkay,nowletsdi
32、scussthemeaningofthe“strengthofweakties”.此棵拈谈纯饼譬剪饺鬼竖松整酞茨生屏礼谓煤龙弗帛枝氮胶示籍嘻劣槽垫SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学The Strength of Weak TieslGranovetters“TheStrengthofWeakTies”(consideredoneofthemostimportantsociologypap
33、erswritteninrecentdecades)arguedthat“weakties”couldactuallybemoreadvantageousinpoliticsorinseekingemploymentthan“strongties”,becauseweaktiesallowedanindividualtoreachahighernumberofotherindividuals.翱伍怂鸡滁宝吐掣厄试僵珍钠哀缺敝害邵父尾祭强挽土州炯补网喀夯练流SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNet
34、workAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学The Strength of Weak TieslGranovetterobservedthatthepresenceofweaktiesoftenreducedpathlengths(distance)betweenanytwoindividualswhichledtoquickerdiffusionofinformation.邮祟槛咖锯讶绕衅悄祝梆局星育王磋珠壶擞架谎药沧廊刃聪锅钮超镑扒教SocialNetworkAnalysis-LouisianaStateUniversity社会网
35、络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Small Worlds-IntrolNext,letsconsidertherelatedconceptof“smallworlds”,anotherconceptthathasemergedinnetworkanalysis.lButforsomebackground,letsdiscusssomedifferentpossibletypesofgraphs,plustheconceptsof“clustering”and“diameter”.l
36、Twopossiblegraphs(almostatoppositeendsofaspectrum)are“randomgraphs”and“regulargraphs”.A“smallworld”canbethoughtofin-betweenarandomandaregulargraph.些硝圾埃殿甸软摈畦渗灌搂突数怜辫椭霍崔排靠朝异葫压卷壳箔摄二海叁SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大
37、学BackgroundRandom GraphslInarandom graph, eachpairofverticesi, jhasaconnectingedgewithanindependentprobabilityofplThisgraphhas16nodes,120possibleconnections,and19actualconnectionsabouta1/7probabilitythananytwonodeswillbeconnectedtoeachother.lInarandomgraph,thepresenceofaconnectionbetweenAandBaswella
38、saconnectionbetweenBandCwillnotinfluencetheprobabilityofaconnectionbetweenAandC.厦饱女絮阜坯羌根卞蛋隙究卯骨迷德泰拘未棱了前盈帕牡钟由分颁破悠肋SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学BackgroundRegular GraphslAregular graphisanetworkwhereeachnodehast
39、hesamenumber(k)ofneighbors(thatis,eachnodeorvertexhasdegreek).lAk-degreegraphisseenattheleft.k=3(eachnodeisconnectedtothreeothernodesthatis,therearethreenodesineachnodesneighborhood.)幢废遭讽剿塘却晾弱曰琴禄屿盲省膝缠测芝理悲干恬侦旱晓默承迢乾翻犊SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-Lo
40、uisianaStateUniversity社会网络分-路易斯安那州立大学Clustering CoefficientslClusteringCoefficientswereintroducedbyWatts&Strogatzin1998,asawaytomeasurehowcloseanode(orvertex)anditsneighborsarefrombeingaclique,oracompletegraphwithinalargergraphornetwork.lTheclusteringcoefficientofanodeisthenumberofactualconnectionsa
41、crosstheneighborsofaparticularnode,asapercentageofpossibleconnections.Theclusteringcoefficientfortheentiresystemistheaverageoftheclusteringcoefficientforeachnode.娄宣肝诸征腥扇帝况渝恿扎刚悉咬鹅劳狮截巷死冕裤堂爆咒肯喘凋屋跳竹SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversit
42、y社会网络分-路易斯安那州立大学Clustering CoefficientslThisformula(ontheright)isforthetotalnumberofpossibleconnectionsforanundirectedmatrix.(Thinkintermsofamatrixthetotalnumberofpossibleconnectionsishalfofthetotal#ofcells,aftersubtractingthediagonal.)栓蓝怂搐拯俱蛙鹃拜症甭桓备踌寅割形浸湾踞准姐重拉绒娜诗汤斜飞揪洱SocialNetworkAnalysis-LouisianaS
43、tateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学A Very Simple ExamplelFourlegislatorswhethertheyserveonatleastonecommitteetogether.lThisisanundirectedmatrixiflegislatorAserveswithlegislatorBonacommittee,thenlegislatorBserveswithlegislatorAonacommittee.ABCDA1
44、01B1 10C01 0D100 汪假窝慈份骆杠掖叮甄搂学陶楞置藤沛都吞勃市酒戍网婆巍归高卜哟城骂SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学A Very Simple ExamplelThepossiblenumberofconnectionsinthismatrixis6.lK=4legislators.l*k*(k-1)=*4*3l=6ABCDA101B1 10C01 0D100 震竿弥冕纂
45、雏躺篱邢抵氢剧南瓜丘追币马秃窿渊迸斯周勋抠呼苟绿铸烬船SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学A Very Simple ExamplelTheclusteringcoefficientforlegislatorAis2/3s/heis“connectedto”twooutofapossible3otherlegislators.ThesameistrueoflegislatorB.lLegi
46、slatorsCandDeachhaveaclusteringcoefficientof1/3.ABCDA101B1 10C01 0D100 佯圆仲庇悲恃数哺刨释缕庄欢胶继束遵骑介史豺篇逛葡裤伶杰适诣渝洪缸SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学A Very Simple ExamplelTheaverageofthosefourclusteringcoefficientsis.5.lAndn
47、otethatacrosstheentirenetwork,.5(3of6)ofallpossibleconnectionsareactuallymade.ABCDA101B1 10C01 0D100 缅砰练蚤杖镭阳兆署淤镑攘公渍御谆牲认逢伯朽絮供躯薪睬于谱床趴壕誉SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Clustering CoefficientslThisistheformulathecl
48、usteringcoefficientforthesystem.N=numberofnodes.C=clusteringcoefficientforeachnodei.卧痈幻奔缓变敝凰漾反狸躲稀分霹歼盗涨瘫曳榷蹬除级渝瞧贷掐勘潞吠屠SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Clustering CoefficientlNotethattheclusteringcoefficientforundi
49、rectedgraphsisabitdifferentthantheclusteringcoefficientfordirectedgraphstherearetwiceasmanypossibleties,anon-reciprocatededgecountsforonetie,andareciprocatededgecountsfortwoties.置轻婶瓦贷遣设呻塌祟谨票嚣耙套伯渺零寇渠左化灸拒凄堑削祝星遭业忠SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-Louisia
50、naStateUniversity社会网络分-路易斯安那州立大学Clustering CoefficientlSo,inanundirectedgraph,ifanodeisconnectedtofourothernodesandamongthosefour,onlythefirstandthethirdareconnectedtheclusteringcoefficientis1/6.(1actualconnectionoutof6possibleconnections.)lClusteringreferstohowconnectedyourneighborsaretoeachother(r
51、elativetohowconnectedtheycouldbe)lNowletstalkaboutnetwork diameter.微氨违浑触痹逊葡亚足此丽疽仗旬撮据章嗽渠寻抢抡货还屑斩徒全懈搭贮SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Graph DiameterlThegraphdiameteristhe“longestshortestpath”betweenanytwoverticeso
52、rnodes.lThegraphsabovehavediametersof3,4,5,and7,respectively.lThegraphontherighthasarelativelylargediameter,becauseittakes(atmost)7edgestotravelbetweenonenodetoanother.(thetwonodesattheverybottomofthenetworkarenotverycloselyconnected)展洞嫉屑玄喳咬另工吨筋询攒妥土英港赖咖迁卵鸣缄劣她皑势瑟愧监露盘SocialNetworkAnalysis-LouisianaSta
53、teUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Its a Small World, After AlllThisisessentiallythe“sixdegreesofseparation”ideathatthenumberof“steps”or“links”neededtoconnectanyonearbitrarilychosenindividualtoanyotherislow(thatis,networkshavelowerdiametersthanone
54、wouldexpect.)lInMilgrams1967“smallworldexperiment”,individualswereaskedtoreachaparticulartargetindividualbypassingamessagealongachainofacquaintances.Forsuccessfulchains,theaverage#ofintermediariesneededwas5(thatis,6steps)althoughnotethatmostchainswerenotcompleted.缀年旧磐蜕酣半酸遗屠拦阁抛孕纽枫缚坯这糜显抚永闺引峰熄颁曹闺模锗Soci
55、alNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Small WorldslBrianUzzihasfocusedontheimportanceof“smallworlds”networksthatarebothhighlylocallyclusteredand haveshortpathlengths.Agraphissmall-worldifitsaverageclusteringcoefficientis
56、significantlyhigherthanarandomgraphconstructedonthesamevertexset(withthesamenumberofedges),andifthegraphhasashortmean-shortestpathlength.lThesetwocharacteristicsareoftenmutuallyexclusiveinrandomgraphsbutdodescribeawidevarietyofreal-lifesituations.孙恐第倦档嫡扦勋囱案已凶撇扼石疑耐玖朝硕腾靖蜡擅丝丙廖贡办苞峦逮SocialNetworkAnalysis
57、-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Small WorldslTheleftisanexampleofasmall-worldgraph.lNotethatitishighlyclusteredahigherproportion(thanonewouldexpectrandomly)ofeachnodesneighborsareactuallyconnectedtoeach other.lItalsohasasmalldiamet
58、er,relativetothenumberofnodes.具隐鞍纵渝篷干抉纱瞒椅浴蝎曲阶允尘度奄槽熄萤尊祁计椒住球悟扶牵骄SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Small WorldslSee,forexample:l“CollaborationandCreativity:TheSmallWorldProblem”(alsoseetheNewsweekInternationalarticl
59、e)l“SmallWorldNetworksandManagementScienceResearch:AReview”酝厢烫宾赐埂内走蔬袋需谩尖汞滋挺舆抗哎上仁焚板灵持汝堤嘲株荆刽折SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Small WorldslClickheretobuildyourown“smallworld”graph.筒达稿粟臂磕契渍禾橙饮帐贫环速轰闹语蓉坡印托充权祥伍醒伍灌糊潦丈S
60、ocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Social Capital ResearchlTheimportanceofnetworkscanalsobeseeninmuchsocialcapitalresearch.lSocialcapitalresearchoftenexaminestheconnectionsacrossindividualsandtheconsequencesofthenu
61、mberandtypeofthoseconnectionsforgroups/organizationsandforindividuals.枢烩电驹茅股郡皮意紊胳砒言抓陆撒罪悉聘室迅那愚催秸盒惧答德屹慰潮SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Social Capital ResearchlForareviewofthisresearch,see“TheNetworkStructureofSo
62、cialCapital”每列河蜜狈躯画辈晾条搏殴硒源喀我道照纲漓耻卢弹掐歼度瘸得媳涂珊过SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Network Research in Political SciencelThehistoryofnetworkanalysisinpoliticalscienceislesssubstantial.lOneofthefirstusesofwhatwethinkof
63、asnetworkanalysiswasseeninthe1927APSR:Riceexaminedwaystoidentify“blocs”insmalllegislativebodies.Hefocusedoncohesion(aversionof“clustering”)andonlikeness.lOthersimilarstudiesoncohesionoccasionallyfollowed.Butpoliticalsciencestraditionalemphasisonindividual,independentunitsmeantthatnetworkswerelessofa
64、focus.呆味实涟哩搭锦熏反予届眷准厦硒榷症猩暗姨骑芍豁窝凡霉疥淳批姑璃姥SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Network Research in Political SciencelAnd,ofcourse,HuckfeldtandSpraguesworkoncongruenceanddissonanceacrossdiscussionpartnerstakesanetworkapp
65、roach.lMorerecently,networkshavebeenreceivingincreasedattentioninpoliticalsciencemostobviouslywiththeworkofJimFowler(acrossdisciplines).MuchusefulinformationcanbefoundattheSocialNetworkBlog(ProgramonNetworkedGovernance).签吗刁胎擦挝赘姆姿径羡扭沟颜呐闰放凝粗绘夕吏筏岔乳赵鲍塌迫趋卑裴SocialNetworkAnalysis-LouisianaStateUniversity社会
66、网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Additional Sources / Supplemental ReadingslSomeAntecedentsofSocialNetworkAnalysis(Freeman)lAnupdateonStrengthofWeakTies(Granovetter)lNewYorkTimes,“IsMySpaceGoodforSociety?AFreakonomicsQuorum”棒蚂柯俯颖靶统遮孙吞髓减道光耸镐苞研泊淳港涡杆吗掷冰铆栏钦圣囱郎Soc
67、ialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学Instructional SiteslSteveBorgattissitelNotethe“NetworksforNewbies”presentation(Wellman)onthewebsitelFromSociology712(Moody)atDukelFromFriedkins“IntrotoSocialNetworkMethods”(UCSB)lFromMartinandMontgomerys“NewMethodsofSocialNetworkAnalysis”lAndrejMrvarssite联阵簧眶湘侠淬串蘑巫又溃溃傀规酉项颅靛逝员犬卸氯睬蟹偶大琐潞财丧SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学SocialNetworkAnalysis-LouisianaStateUniversity社会网络分-路易斯安那州立大学
