chi square test - R dyplr-CodePudding

I have the following dataframe :

data = data.frame(group = c("conse", "dem", "schoo"),
                  f = c(43, 30, 36),
                  m = c(rep(29, 17, 36))

I would like to apply a chi square test to check whether the proportion of females and males is comparable in the three groups.

But when I write the following code :

data_chi = data%>%
  chisq.test()

I get an error message saying "Error in sum(x) : invalid 'type' (character) of argument"

I have tried to transform my table as a tibble and nothing happened then (not even an error message).

Could you help me find a solution ?

Thanks a lot!

CodePudding user response：

You need to drop the group column, since the Chi square test will only work on numeric data, and doesn't know what to do with a character column

data = data.frame(group = c("conse", "dem", "schoo"),
                  f = c(43, 30, 36),
                  m = c(rep(29, 17, 36)))

library(dplyr)

data %>%
  select(-group) %>%
  chisq.test()
#> 
#>  Pearson's Chi-squared test
#> 
#> data:  .
#> X-squared = 12.44, df = 35, p-value = 0.9998

^{Created on 2023-01-25 with reprex v2.0.2}

CodePudding user response：

another option

library(magrittr)
data <- data.frame(
  group = c("conse", "dem", "schoo"),
  f = c(43, 30, 36),
  m = c(29, 17, 36)
)

data %$% chisq.test(data[-1])
#> 
#>  Pearson's Chi-squared test
#> 
#> data:  data[-1]
#> X-squared = 2.5522, df = 2, p-value = 0.2791

^{Created on 2023-01-25 with reprex v2.0.2}