success argument for prop_test

[AmeliaMN]

It would be nice to have a success argument for prop_test, analogous to the one from mosaic::prop.test. Here's how it could be used:

gss %>%
    prop_test(response = college, p = 0.6, success = "no degree")

[mine-cetinkaya-rundel]

I agree, especially since specify() also has that argument

[simonpcouch]

Ah, agreed! Giving this a go right now. :-)

[rpruim]

Note about how mosaic processes this: If you specify success, all other values are collectively "failure":

prop.test( ~ resample(LETTERS[1:4], size = 100), success = "A", p = 0.25)
## 
## 	1-sample proportions test with continuity correction
## 
## data:  $resample(LETTERS[1:4], size = 100)  [with success = A]
## X-squared = 0.12, df = 1, p-value = 0.729
## alternative hypothesis: true p is not equal to 0.25
## 95 percent confidence interval:
##  0.1883648 0.3696071
## sample estimates:
##    p 
## 0.27 

Is that how specify() works as well?

[simonpcouch]

In specify(), the success argument is only used for binary response variables. It works similarly in that the "not success" doesn't have to be explicitly stated, but "all other values" in that case would only be one value regarded as failure. :-)

I usually think of these wrappers as just providing a tidy interface to the analogous function from stats, in which case success would be ignored (as the factor levels are in stats::prop.test, I think) in the case that the response has more than 2 levels.

Just put in a PR (linked above)! A demo of how this could look:

library(infer)

# one-sample (note that p != .5)
prop_test(gss,
          college ~ NULL,
          p = .2,
          success = "degree")
#> # A tibble: 1 x 4
#>   statistic chisq_df  p_value alternative
#>       <dbl>    <int>    <dbl> <chr>      
#> 1      67.5        1 2.08e-16 two.sided

prop_test(gss,
          college ~ NULL,
          p = .2,
          success = "no degree")
#> # A tibble: 1 x 4
#>   statistic chisq_df   p_value alternative
#>       <dbl>    <int>     <dbl> <chr>      
#> 1      636.        1 2.98e-140 two.sided

prop_test(gss,
          college ~ NULL,
          p = .8,
          success = "degree")
#> # A tibble: 1 x 4
#>   statistic chisq_df   p_value alternative
#>       <dbl>    <int>     <dbl> <chr>      
#> 1      636.        1 2.98e-140 two.sided

# two-sample
prop_test(gss, 
          college ~ sex,  
          order = c("female", "male"),
          success = "degree")
#> # A tibble: 1 x 6
#>   statistic chisq_df p_value alternative lower_ci upper_ci
#>       <dbl>    <dbl>   <dbl> <chr>          <dbl>    <dbl>
#> 1 0.0000204        1   0.996 two.sided    -0.0917    0.101

prop_test(gss, 
          college ~ sex,  
          order = c("female", "male"),
          success = "no degree")
#> # A tibble: 1 x 6
#>   statistic chisq_df p_value alternative lower_ci upper_ci
#>       <dbl>    <dbl>   <dbl> <chr>          <dbl>    <dbl>
#> 1 0.0000204        1   0.996 two.sided     -0.101   0.0917

prop_test(gss, 
          college ~ sex,  
          order = c("male", "female"),
          success = "degree")
#> # A tibble: 1 x 6
#>   statistic chisq_df p_value alternative lower_ci upper_ci
#>       <dbl>    <dbl>   <dbl> <chr>          <dbl>    <dbl>
#> 1 0.0000204        1   0.996 two.sided     -0.101   0.0917

^{Created on 2020-12-03 by the reprex package (v0.3.0.9001)}

[github-actions]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.