Package: base

Function: rowMeans()

1. Calculate the percent agreement across all variables for each teacher/pair of raters

Review the data (rater_data)

# A tibble: 6 x 5
  tch_id rater_id item1 item2 item3
   <dbl>    <dbl> <dbl> <dbl> <dbl>
1    100        5     1     0     3
2    100       10     1     1     3
3    101        5     0     1     2
4    101        8     0     0     2
5    102        6     1     1     1
6    102        8     1     1     1

First get an agreement score for each variable. 0 = no agreement, 1 = agreement

We can do this by
1. First grouping by tch_id using dplyr::group_by()
2. Then summarizing the differences between raters across all variables using dplyr::across() and base::diff()
3. Then using dplyr::mutate() to recode the items to values of 0 (no agreement) or 1 (agreement)

rater_summary <- rater_data %>% 
  group_by(tch_id) %>%
  summarise(across(item1:item3, diff)) %>%
  mutate(across(item1:item3,
                       ~ case_when(
                         . == 0 ~ 1,
                         .default = 0
                       )))

rater_summary
# A tibble: 3 x 4
  tch_id item1 item2 item3
   <dbl> <dbl> <dbl> <dbl>
1    100     1     0     1
2    101     1     0     1
3    102     1     1     1

Next we can calculate a row mean (using base::rowMeans()) to get a percent agreement per teacher/rater pair.

rater_summary %>%
  mutate(percent_agree =
           round(rowMeans(across(item1:item3), na.rm = TRUE),
                       digits = 2))
# A tibble: 3 x 5
  tch_id item1 item2 item3 percent_agree
   <dbl> <dbl> <dbl> <dbl>         <dbl>
1    100     1     0     1          0.67
2    101     1     0     1          0.67
3    102     1     1     1          1

Return to Calculate Sums and Means