SPSS_1.ppt

Diagram of descriptive statistics for categorical variables Confidence interval Hypotheses test 单个样本 explore 配对样本差值均数（或率）的可信区间 paired t-test 完全随机设计量样本差值均数（或率）的可信区间 Independent t-test 二项分布指标以0，1 方式输入 数据的正态性检验 (explore) 单样本的t检验 配对设计的t检验 成组设计的t检验 * The rows are the “true” status of the women as determined by the “GOLD STANDARD”. The columns are the results of the test. * This is a case-control study looking at reasons for high rates of lung cancer among male residents of costal Georgia. CONTROLS: were selected from admissions to the four hospitals and from death certificates in the same period for diagnoses other than lung cancer or bladder cancer. CASES: Identified from 3 sources: males diagnosed since 1970 at a large hospital in Brunswick Diagnoses during 1975-1976 at 3 major hospitals in Savannah Death certificates for the period of 1970-1974 in the area EXPOSURE: smoking and working in a shipyard during WWII. Shipbuilding is thought to be a confounder. 西安交通大学医学院 SAS Statistics analysis system 　－University of North Carolina, USA 　－ＳＡＳ９．０，１Ｇ 　－Designed for professional users 　－Operating mainly in programming, human-machine dialogue interface unfriendly 　－Too Expensive Stata Statistics/Data Analysis － Development Center for Computer Resource, USA 　－Ｓtata 10.0，２０Ｍ 　－ Designed for professional users －Procedures Statements are simple and flexible, easy to study, and has powerful features to help 　 SPSS Statistical package for the social sciences 　－Stanford University, USA 　－ＳＰＳＳ１7.０, 700 M 　－Mainly menu-driven, user-friendly 　－Non-professional users preferred software 　－Only include the sophisticated statistical methods, the results can not be opened directly by commonly used word processing software 结构合理的数据库能保证统计分析软件正确识别 变量名放在第一行 每一行------一个观察对象 每一列------一个观察变量 No. Weight Height Gender No. Weight Height Gender No. Weight Height Gender No. Weight Height Gender 1 68 170 M 16 65 144 M 31 12 115 F 46 22 122 M 2 63 150 M 17 43 133 M