[讨论][求助]如何检验两组数据的差异性

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

如：在一个分组调查中，分为男女两组。

这两组人都对同一个问题进行回答，比如对学校的看法：a、很好；b，好；c，一般；d，不好；e，很不好。

如何检验男女两组对该问题回答的差异性？

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：差异性 数据 检验 讨论 差异性

You can assign these five scores these response, higher score means better.

create a variable 'response': 1=很不好, 2=不好, 3=一般, 4=好, 5=很好

(a、很好；b，好；c，一般；d，不好；e，很不好。)

And then you use t-test to calculate the p value by gender group (male/female), if p < .05, you say the boys and girls are different. (One group is higher). If your response variable is not normally distributed, you can use nonparametric test, Wilcoxon test is the most common one.

SAS code:

proc means data=yourdata;

class gender;

var response;

run;

Proc npar1way data=yourdata wilcoxon;

class gender;

var response;

run;

***Any other way is to compare the percentage of each level, and use 卡方 chi-sqare test for 2 by 5 table (response is 5 levels) or 2 by 2 table (2 levels: one for good, one for not good). And use proc freq in SAS program.

proc freq data=yourdata;

tables gender*response/chisq;

run;

以下是引用zzheng722在2007-4-12 5:29:00的发言：

You can assign these five scores these response, higher score means better.

create a variable 'response': 1=很不好, 2=不好, 3=一般, 4=好, 5=很好

(a、很好；b，好；c，一般；d，不好；e，很不好。)

And then you use t-test to calculate the p value by gender group (male/female), if p < .05, you say the boys and girls are different. (One group is higher). If your response variable is not normally distributed, you can use nonparametric test, Wilcoxon test is the most common one.

SAS code:

非常感谢，高手啊。

不过，用spss应该怎么做？

关于差异性，这个P值的节点到底是0.1还是0.05啊，有些资料上说是0.1