儿童抑郁症调查中数据缺失情形下logistic回归模型参数的最大似然预计.pdfVIP

Abstract existsmostmedical Becauseofvarious in reasons，data missing surveys．We notuse the We could themethodto such dataaS handlethe analyzemissing way observeddata．Howtohandlethese datais oneofthe complete missing becoming focusesofthestatisticiansinrecent years． The methodof this isthe ofthe rejection data，it original solving missing problem and ofthe means estimationobserved onlynlodelingparameter data．Obviously,this will in ofthe methodresultthe ofinformation statisticianshave losing model．Many varjonsmethodstothe Little(1987) suggested problem，Rubin(1974，1976，1978)and the the basic of of presented approaches analysisincompletedata．Dempster,Laird and describedtheEM under Rubin(1977)systematically algorithmincomplete situation．Theseall the of dataforUS． solvingmissing appliedthought isbasedOilthe This dataof Children paper Shanghai DepressionSurvey．The ofthemodelis outcome intendtouseaimeddatatoresetthe missing．We data,and the model