1、2020/6/21,Xian Jiaotong University-Medical School,第二章 計(jì)量資料的統(tǒng)計(jì)描述,集中趨勢(shì)描述 均數(shù):對(duì)稱分布,特別是正態(tài)分布資料 幾何均數(shù):等比級(jí)數(shù)資料,對(duì)數(shù)正態(tài)分布資料 中位數(shù):不適合用幾何均數(shù)描述的偏態(tài)分布資料 末端有不確定數(shù)據(jù)的正態(tài)分布資料 離散趨勢(shì)描述 極差:不穩(wěn)定 四分位數(shù)間距:用于偏態(tài)分布資料 方差與標(biāo)準(zhǔn)差:用于正態(tài)分布資料,描述平均變異絕對(duì)變異 變異系數(shù):用于正態(tài)分布資料變異度對(duì)比相對(duì)變異,2020/6/21,Xian Jiaotong University-Medical School,例2-1 從某單位1999年的職工體檢資料中

2、獲得101名正常成年女子的血清總膽固醇的測(cè)量結(jié)果如下，試編制頻數(shù)分布表。,2020/6/21,Xian Jiaotong University-Medical School,編制步驟,求極差 極差（range）也稱全距 確定組段數(shù)和組距 根據(jù)組距寫出組段 分組劃記并統(tǒng)計(jì)頻數(shù),2020/6/21,Xian Jiaotong University-Medical School,表2-2 115名正常成年女子血清轉(zhuǎn)氨酶（mmol/L）含量分布,2020/6/21,Xian Jiaotong University-Medical School,2020/6/21,Xian Jiaotong Unive

3、rsity-Medical School,頻數(shù)表和頻數(shù)分布圖的用途,描述頻數(shù)分布的類型 描述頻數(shù)分布的特征 便于發(fā)現(xiàn)一些特大或特小的可疑值 便于進(jìn)一步做統(tǒng)計(jì)分析和處理,2020/6/21,Xian Jiaotong University-Medical School,2020/6/21,Xian Jiaotong University-Medical School,集中趨勢(shì)的描述 均數(shù),直接計(jì)算法 頻數(shù)表法,2020/6/21,Xian Jiaotong University-Medical School,集中趨勢(shì)的描述 幾何均數(shù),公式,2020/6/21,Xian Jiaotong Univ

4、ersity-Medical School,集中趨勢(shì)的描述 中位數(shù),n為奇數(shù): n為偶數(shù):,2020/6/21,Xian Jiaotong University-Medical School,離散趨勢(shì)描述-極差四分位距,極差R: 最大值與最小值之差 四分位距 = -,2020/6/21,Xian Jiaotong University-Medical School,離散趨勢(shì)描述-方差標(biāo)準(zhǔn)差,總體方差與總體標(biāo)準(zhǔn)差 樣本標(biāo)準(zhǔn)差,2020/6/21,Xian Jiaotong University-Medical School,離散趨勢(shì)描述-變異系數(shù),不同均數(shù)資料的變異度比較 不同度量單位資料的變異

5、度比較,2020/6/21,Xian Jiaotong University-Medical School,一、常用相對(duì)數(shù),強(qiáng)度相對(duì)數(shù)：率 某時(shí)期內(nèi)發(fā)生某現(xiàn)象的觀察單位數(shù) 同期可能發(fā)生某現(xiàn)象的觀察單位總數(shù) 結(jié)構(gòu)相對(duì)數(shù) ：構(gòu)成比 某一組成部分的觀察單位數(shù) 同一事物各組成部分的觀察單位總數(shù) 相對(duì)比 甲指標(biāo)/乙指標(biāo),2020/6/21,Xian Jiaotong University-Medical School,二、計(jì)算相對(duì)數(shù)的注意事項(xiàng),結(jié)構(gòu)相對(duì)數(shù)不能代替強(qiáng)度相對(duì)數(shù)（以比代率） 計(jì)算相對(duì)數(shù)應(yīng)有足夠數(shù)量 正確計(jì)算合并率 注意資料的可比性 相對(duì)數(shù)的抽樣誤差,2020/6/21,Xian Jiaoton

6、g University-Medical School,SPSS軟件簡(jiǎn)介,常用窗口:數(shù)據(jù)窗程序窗結(jié)果輸出窗口 變量定義: 變量名:不超過8個(gè)字符,最好以英文字母加數(shù)字組成,避免用中文 變量類型:數(shù)值型日期型字符型 小數(shù)位數(shù) 缺失值定義 變量標(biāo)簽:可用中文 值標(biāo)簽:分類變量需要定義值標(biāo)簽 文件類型: 數(shù)據(jù)文件.sav 程序文件.sps 結(jié)果文件.spo,2020/6/21,Xian Jiaotong University-Medical School,產(chǎn)生新變量,數(shù)據(jù)排序:sort,2020/6/21,Xian Jiaotong University-Medical School,數(shù)據(jù)庫(kù)建立,設(shè)

7、計(jì)正確的數(shù)據(jù)庫(kù)，保證統(tǒng)計(jì)軟件能識(shí)別 數(shù)據(jù)庫(kù)一般為行列表 第一行為變量名 一個(gè)case 占用一行 一個(gè)variable占一列 盡可能采用數(shù)據(jù)形式錄入,2020/6/21,Xian Jiaotong University-Medical School,數(shù)據(jù)的前處理,數(shù)據(jù)庫(kù):數(shù)據(jù)的儲(chǔ)存?zhèn)浞菔褂脵?quán)限 數(shù)據(jù)雙輸?shù)谋匾?數(shù)據(jù)庫(kù)名稱,變量定義,必要的說明 原始數(shù)據(jù)的重要性!電子數(shù)據(jù)庫(kù)紙質(zhì)數(shù)據(jù)庫(kù) 數(shù)據(jù)考察:數(shù)據(jù)類型? 分布? 異常值? 缺失數(shù)據(jù)的處理? 數(shù)據(jù)清理與鎖定 數(shù)據(jù)庫(kù)軟件: Access, EpiData, Oracal 統(tǒng)計(jì)軟件中的數(shù)據(jù)庫(kù),2020/6/21,Xian Jiaotong Univer

8、sity-Medical School,描述性統(tǒng)計(jì),計(jì)量資料 圖表形式：箱式圖、直方圖 指標(biāo)形式：集中趨勢(shì)：均數(shù)、中位數(shù)、幾何均數(shù) 離散趨勢(shì)：極差、四分位距、方差、 標(biāo)準(zhǔn)差、變異系數(shù) 分類資料 圖標(biāo)形式：直方圖、線圖 指標(biāo)形式：率、比,2020/6/21,Xian Jiaotong University-Medical School,數(shù)據(jù)庫(kù)一般操作,數(shù)據(jù)庫(kù)合并：merge 追加變量、追加病例 數(shù)據(jù)庫(kù)拆分：split 衍生數(shù)據(jù)：產(chǎn)生新的變量,2020/6/21,Xian Jiaotong University-Medical School,常用命令_衍生數(shù)據(jù)命令,產(chǎn)生新變量 (transform

9、) Recode: numeric - string Compute: constant functions Date/time:,2020/6/21,Xian Jiaotong University-Medical School,常用命令_數(shù)據(jù)步命令1,數(shù)據(jù)步 Insert variables Insert case Go to case Sort cases,Select cases: give some conditions Weight cases: just used for grouped database.,2020/6/21,Xian Jiaotong University-M

10、edical School,常用命令_數(shù)據(jù)步命令2,File commands Split file: split file into sub-database according to the categorical variables Merge files: Add cases: the files should be have same structures. Add variables: Key variable (unique identification number for the observations),2020/6/21,Xian Jiaotong University

11、-Medical School,常用命令_數(shù)據(jù)步命令2,數(shù)據(jù)庫(kù)架構(gòu)改變 Transpose: row - column (transpose rows and columns ) Restructure: long - short (repeated measurement),2020/6/21,Xian Jiaotong University-Medical School,描述性統(tǒng)計(jì)過程,Descriptive statistics Frequencies: continuous variables categorical variables Descriptive: continuous

12、variables (Normal distributed) Explore: continuous variables Crosstabs: categorical variables,2020/6/21,Xian Jiaotong University-Medical School,基本統(tǒng)計(jì)圖,Graphs Bar graph: continuous/categorical variables (comparison) Pie graph: categorical variables (constituent ratio) Line graph: proportions and time

13、variables (time trend) Histogram: continuous variables (frequency distribution) Scatter plot: continuous variables (correlation analysis ) Box plot: continuous variables (check median, quartiles, maximum, minimum),2020/6/21,Xian Jiaotong University-Medical School,統(tǒng)計(jì)圖_連續(xù)型變量,measures that are recorded on a naturally occurring number scale. Box plots Histogram,2020/6/21,Xian Jiaotong University-Medical School,Graph commands,Histogram Horizontal s