Merge pull request #485 from gungorMetehan/patch-1

Update 92-appendixB.Rmd

92-appendixB.Rmd

@@ -604,7 +604,7 @@ n_nograd <- off_summ$sample_size[1]
 n_grad <- off_summ$sample_size[2]
 ```
 
-Observe that of the college graduates, a proportion of `r counts[2, 2]`/(`r counts[2, 2]` + `r counts[2, 3]`) = `r phat_grad` have no opinion on drilling. On the other hand, of the non-college graduates, a proportion of `r counts[1, 2]`/(`r counts[1, 2]` + `r counts[1, 3]`) = `r phat_nograd` have no opinion on drilling, whereas . The difference in these proportions is `r phat_grad` - `r phat_nograd` = `r obs_diff`.
+Observe that of the college graduates, a proportion of `r counts[2, 2]`/(`r counts[2, 2]` + `r counts[2, 3]`) = `r phat_grad` have no opinion on drilling. On the other hand, of the non-college graduates, a proportion of `r counts[1, 2]`/(`r counts[1, 2]` + `r counts[1, 3]`) = `r phat_nograd` have no opinion on drilling, whereas the difference in these proportions is `r phat_grad` - `r phat_nograd` = `r obs_diff`.
 
 Let's visualize these in a barchart. However, we first reverse the order of the levels in the categorical variable `response` using the `fct_rev()` function from the `forcats` package. We do this because the default ordering of levels in a factor is alphanumeric. However, we are interested in proportions that have `no opinion` and not `opinion`. Thus we need to reverse the default alphanumeric order.