When using the nops_eval
function in an exam with only schoice
items (5 possible answers), we want the wrong answers to be evaluated like 1/n number of incorrect options. E.g. an item with 3 expoints with a wrong answer should receive: 1/-4 = -.25
. However, if a student answers two options in an exam:
It must not give any positive points... it is a schoice
question, so it should evaluate as a wrong answer (i.e. 1/(-4)
) even if one of the answers is the right one.
The used options were:
eval_v1 <- nops_eval(register = "nops_eval/pauta_II.csv",
solutions = "nops_eval/Ex_AEII_MTI_v1_.rds",
scans = "nops_eval/nops_scan_20210120211740_v1.zip", language = "pt",
eval = exams_eval(partial = T, negative = T, rule = "false"),
dir = "eval",
interactive = T)
If the argument rule = "false2"
the final result is the same.
We found the solution... we should set partial = F, negative = -.25
. With partial = F
, we assure that only full right answers are accepted (as schoice
presumes) and with negative = -.25
we assure that all wrong answers are penalized with -.25
times the expoints
of each exercise.