success argument in prop_test()

[ismayc]

Reviewing https://infer.tidymodels.org/reference/prop_test.html, it strikes me that {infer} should probably have a success argument for prop_test() if testing a point estimate in much the same way that order exists in the two sample case. In the gss data below, the unique values of college are "degree" or "no degree" and it seems that "degree" is specified as a success here (maybe internally to prop.test()?), but to a beginner that wouldn't necessarily be apparent.

library(infer)

# proportion test for difference in proportions of 
# college completion by respondent sex
prop_test(gss, 
          college ~ sex,  
          order = c("female", "male"))
#> # 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          
# a one-proportion test for hypothesized null 
# proportion of college completion of .2
prop_test(gss,
          college ~ NULL,
          p = .2)
#> # A tibble: 1 x 4
#>   statistic chisq_df   p_value alternative
#>       <dbl>    <int>     <dbl> <chr>      
#> 1      636.        1 2.98e-140 two.sided  

[echasnovski]

It seems to be the same functionality as added in #344. Am I missing something here?

[ismayc]

Sorry, I didn't have the develop version of the package installed and I missed this update for some reason..

[ismayc]

@echasnovski @simonpcouch Should we also update the example to include the new success argument?

# one-sample proportion test for hypothesized null
# proportion of college completion of .2
prop_test(gss,
          college ~ NULL,
          success = "degree",
          p = .2)

[ismayc]

I know that prop.test() reports a Chi-square statistic for the point estimate test and for the difference in proportions test, but that is rarely how it is taught initially to novices. Usually, it is first taught as a z statistic (for example, from OpenIntro). Maybe having a z = FALSE argument by default could be a nice way for instructors to turn it on and then after students feel comfortable with that, they could move on to discuss how z^2 = Chi-square?

[ismayc]

This also would help prop_test() be more in alignment with the example in the Full infer Pipeline Examples vignette similar to how the other examples there match with corresponding wrapper functions like t_test():

null_distn_theoretical <- gss %>%
  specify(college ~ sex, success = "no degree") %>%
  hypothesize(null = "independence") %>%  
  calculate(stat = "z", order = c("female", "male"))

[simonpcouch]

I'm on board for both of your suggestions, I think!

Working on finals at the moment, so will need a few days, but glad to spend some time with this.

[ismayc]

Of course! No rush

[simonpcouch]

Closing in favor of #353. :-)

[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.