终于会用SPSS做卡方检验了,结果发现双侧检验P=0.096,单侧检验P=0.046,结果如何解释?

请不吝指教

如果取a为0.05即置信度为95%的话,由于p值已经大于a,则拒绝零假设,治疗方式和结果显著相关,也就是说采取不同的治疗方式的治疗后果有显著差异.

如果是比较大于小于之类的就用单侧检验,等于之类的问题就用双侧检验.主要在t检验中.

在crosstabs里面零假设是行列变量不相关,用双侧检验,即直接用p值与a比较，而不是用p/2与a值比较.

Example

CROSSTABS TABLES=FEAR BY SEX
/CELLS=ROW COLUMN EXPECTED RESIDUALS
/STATISTICS=CHISQ.

• CROSSTABS generates a Case Processing Summary table, a Crosstabulation table, and a
Chi-Square Tests table.

• The variable FEAR defines the rows and the variable SEX defines the columns of the
Crosstabulation table. CELLS requests row and column percentages, expected cell
frequencies, and residuals.

• STATISTICS requests the chi-square statistics displayed in the Chi-Square Tests table.Display the Chi-Square Test table. Chi-square statistics include Pearson chisquare,
likelihood-ratio chi-square, and Mantel-Haenszel chi-square (linearby-
linear association). Mantel-Haenszel is valid only if both variables are
numeric. Fisher’s exact test and Yates’ corrected chi-square are computed for
all 2 × 2 tables. This is the default if STATISTICS is specified with no
keywords.

怎样确定采用双侧检验还是单侧检验？有兄弟姐妹碰到这种情况的吗？双侧单侧显著性不同！
再问大家一次这个问题

很傻很天真？

请大师们能否详细告之如何用

spss软件进行卡方检验

如正常体重血脂高17正常361人

 肥胖者血脂高27正常138人

如何用spss软件进行卡方检验

请教具体做法

还有算出的结果，多谢