I am new in R (statistic packet) programming and I would like to make an 1way anova.
My frame of data is like that
q1 sex
1 N M
2 Y F
3 U F
...
1000 Y M
开发者_Go百科Could you help me please ?
Following up on @Thilo's and @DWin's comments above: if you really want to analyze the relationship between two categorical variables, you might try something like this:
## make up random data (no real pattern)
dat <- data.frame(q1=sample(c("N","Y","U"),size=1000,replace=TRUE),
sex=sample(c("M","F"),size=1000,replace=TRUE))
dtab <- with(dat,table(q1,sex))
chisq.test(dtab)
mosaicplot(dtab)
It would be really helpful from the point of view of answering your questions to have some more context: what question are you trying to answer? Also, as always, it's nice to have a reproducible example (just to save the time of coming up with my own way of generating some fake data to play with).
One small point is that the Pearson chi-square test is testing for association; it doesn't distinguish between response and predictor variables as the ANOVA framework does.
Of course, if you really have a continuous response variable that you're neglecting to tell us about, then this won't be too useful ...
精彩评论