R Language Note Independence test

> Mytable1 <-xtabs (~ treatment+improved,data=arthritis)

> Mytable2 <-xtabs (~ improved+sex,data=arthritis)

> Mytable1

Improved

Treatment None Some Marked

Placebo 29 7 7

Treated 13 7 21

> Mytable2

Sexa

Improved Female Male

None 25 17

Some 12 2

Marked 22 6

> Chisq.test (MYTABLE1)

Pearson ' s chi-squared test

Data:mytable1

x-squared = 13.055, df = 2, P-value = 0.001463

> Chisq.test (mytable2)

Pearson ' s chi-squared test

Data:mytable2

x-squared = 4.8407, df = 2, P-value = 0.08889

Warning message:

In Chisq.test (mytable2): chi-squared approximate algorithm may not be allowed

In the results, the level of treatment and improvement that patients received seemed to have a relationship (P < 0.01). and patient sex

There is no relationship between the improvement situation (P > 0.05). The P-Value here indicates that the sample row variables and columns are extracted from the population

Quantity is the probability of mutual independence. Because the probability value is very small, so you reject the treatment type and the outcome of the treatment is independent of the original false

Set Because the probability is not small enough, there is not enough reason to explain that the results of treatment and sex are not independent.

The reason for the warning message is that one of the 6 cells in the table (male-to some extent) has a value less than 5,

This may cause the card-side approximation to be invalid.

R Language Note Independence test