簡易流行病學與生物統計學 二十二、SAS邏輯式迴歸模式(logistic regression model) 作卡方檢定

簡易流行病學與生物統計學
二十二SAS邏輯式迴歸模式(logistic regression model
作卡方檢定
作者吳聰賢醫師

SAS邏輯式迴歸模式(logistic regression model)作卡方檢定
  一樣用有喝可樂、與中毒的例子來說明,
          

中毒
沒中毒

有喝可樂
10
2

沒喝可樂
2
8






               Program editor(編輯視窗)
data intoxi;
input cola poison count;
cards;
0 0 10
0 1 2
1 0 2
1 1 8
;
proc freq;
weight count;
tables cola*poison/chisq;
run;

0 0 10】第一個數字代表喝可樂,0是有喝可樂,1是沒喝可樂;第二個數字代表中毒,0是有中毒,1是沒中毒;後面數字是累積人數,運算結果如下

      Output(運算結果視窗)
Statistics for table of cola by poison
Statistic                    DF     Value         Prob
Chi-square                   1      8.824         0.003
Likelihood ratio Chi-square  1      9.495         0.002
Continuity adj. Chi-square   1      6.455         0.011
Mantel-Haenszel Chu-square   1      8.423         0.004
Fishers exact testleft                       1.000
              right                          4.78E-03
              2-tail                         8.28E-03

運算程式tables cola*poison/chisq、與運算程式tables poison *cola /chisq結果一樣。SAS邏輯式迴歸模式(logistic regression model)運算程式書寫如下:

Program editor(編輯視窗)
data intoxi;
input cola poison;
cards;
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
1 0
1 0
0 1
0 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
;
proc sort;
by descending poison;
proc logistic order=data;
model poison=cola/risklimits;
run;

00】第一個數字代表喝可樂,0是有喝可樂,1是沒喝可樂;第二個數字代表中毒,0是有中毒,1是沒中毒。運算結果如下:

Output(運算結果視窗)
Variable DF  parameter  Wald       Pr     odds ratio    Wald
             Estimate  chi-square  chi-square        confidence limits
                                                     Lower    upper 
Intercept 1  -1.6094   4.3176    0.0377       
Cola    1    2.9957    7.3264    0.0068   20.000    2.285    175.030 
斜率(parameter estimateslope)是2.9957
卡方值(Wald chi-square)是0.0068
勝算比(odds ratio)是20.000
勝算比95%信賴區間(Wald confidence limits)下限2.285,上限175.030。(900321日完稿
參考書目
1.  衛生統計作業手冊 中華民國七十一年 行政院衛生署編印
2.  生物統計學新論 楊志良著 巨流圖書公司印行
3.  Excel 於資料分析與統計學上的應用 王文中著 博碩文化股份有限公司
4.  疫情報導 行政院衛生署
5.  結核病防治年報 民國86年 台灣省慢性病防治局編印
回應
行政院衛生署疾病管制局特殊疾病組 姜OO科長吳醫師您的大作令人嘆為觀止,好多我不懂的地方,都一一獲得清楚的說明。而且文字如同行雲流水,讀來倍感親切。這些作品,不該只是泛讀而已,過些日子等我精讀完畢,如果還有不懂的地方,再向您請益。至於您黑體字所列的部分,大概也是目前科學還沒辦法解釋的問題,而且您問得也真深入,完全抓到了重點所在。可是,我也沒有辦法回答,真是抱歉!再次感謝您寄這些文章給我,謝謝!90414

0 意見:

張貼留言