Reports

This may not answer your question and apologies if I miss the plot completely here but using tidyverse functions should be preferable due to parsimony. Note how I also separate the functions onto their own lines that eases error tracking if needed.

Code:

library(tidyverse)

data %>% 
  mutate(pre_resp = 
           1 - sum(
             across(3:16)), 
         post_resp = 
           1 - sum(
             across(17:30)), 
         .by = participant) %>% 
  select(participant, pre_resp, post_resp)

Result

# A tibble: 6 × 3
  participant pre_resp post_resp
        <int>    <dbl>     <dbl>
1       39496    -3.33    -0.333
2       40008    -2.33     1.33 
3       39550     1       -1.67 
4       39530    -7.67     0.667
5       39956    -2.33     2    
6       39941     1.67     1.33

Since you did ask for any advice, I hope you will take something from the above.

PS remember to up- or downvote any answers or comments.

79077333