| title | output | editor_options | ||||
|---|---|---|---|---|---|---|
Semi-automated grading workflow |
html_document |
|
This is not fully developed R package, just some utility functions and this Rmd file.
The intended use is
- install the package locally:
remotes::install_github("fabian-s/credentialism") - download this Rmd file:
utils::download.file(url = "https://raw.githubusercontent.com/fabian-s/credentialism/main/README.Rmd", destfile = <Your_File>) - modify it so it fits your exam and run the chunks interactively, and copy-paste the relevant outputs into your lecture homepage or the template spreadsheet "Notenliste"-files for the grades.
Assumes that student info was collected from moodle via the standard form for "Klausuranmeldung" and that the resulting sheet as exported from moodle also contains the points (or grades) the students achieved.
For processing further, your results-table needs at least the following columns:
matriculation: "Matrikelnummer"vn,nn: surname, family namepo,anderePO: subjects, as exported from moodle's "Klausuranmeldung"points: achieved points --
NB: 0 points are assumed to mean invalidated / "entwertet" or no-show if you have participants that actually got 0 points, just set them to 0.1, e.g...- (
grade, if not generated automatically frompoints, see below)
E.g. for the fake spreadsheet included in the package:
results <-
rio::import(system.file("extdata", "testtest.csv", package = "credentialism"))
head(results)## Matrikelnummer vn nn po Gesamt
## 1 120798 Goy Xukuraqus Ethnobotanik 28
## 2 112259 Bir Hahogumeqafub Ethnobotanik 35
## 3 154577 Kunu Hamahenanidajaj Finnougristik 57
## 4 141328 Hej Dekihuqecoy Agrartheologie 25
## 5 106914 Guhas Xabelesinetu Agrartheologie 64
## 6 140726 Faxa Yonayeyuq Agrartheologie 48
## andere PO
## 1 NA
## 2 NA
## 3 NA
## 4 NA
## 5 NA
## 6 NA
results <- results |>
dplyr::select(c(Matrikelnummer, vn, nn, po, Gesamt, `andere PO`)) |>
dplyr::rename("matriculation" = Matrikelnummer,
"points" = Gesamt)
pillar::glimpse(results)## Rows: 21
## Columns: 6
## $ matriculation <int> 120798, 112259, 154577, 141328, 106914…
## $ vn <chr> "Goy", "Bir", "Kunu", "Hej", "Guhas", …
## $ nn <chr> "Xukuraqus", "Hahogumeqafub", "Hamahen…
## $ po <chr> "Ethnobotanik", "Ethnobotanik", "Finno…
## $ points <int> 28, 35, 57, 25, 64, 48, 47, 63, 24, 0,…
## $ `andere PO` <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
Set maximum number of points and adjust recommended grading scheme
(Notenstufen) if necessary -- e.g. niceness = 0.01 would decrease the
point percentages needed for a given grade by 1 %.
MAX_POINTS <- 75 #! change this if you rerun for a different exam
(scheme <- set_grading_scheme(max_points = MAX_POINTS, niceness = .0))## 4.0 3.7 3.3 3.0 2.7 2.3 2.0 1.7 1.3 1.0
## 32.5 34.5 36.5 41.5 45.5 50.5 54.5 59.5 62.5 66.5
Then compute grades and identify borderline cases
(skip this if your table already contains grades and not just points, but make sure
scheme and MAX_POINTS are set correctly):
results <- results |>
dplyr::mutate(
grade = grade_points(points, scheme = scheme),
# who failed by 2 pts or less?
almost_but_no_cookie = (points < scheme[1]) & (points > (scheme[1] - 2)))
pillar::glimpse(results)## Rows: 21
## Columns: 8
## $ matriculation <int> 120798, 112259, 154577, 141328,…
## $ vn <chr> "Goy", "Bir", "Kunu", "Hej", "G…
## $ nn <chr> "Xukuraqus", "Hahogumeqafub", "…
## $ po <chr> "Ethnobotanik", "Ethnobotanik",…
## $ points <int> 28, 35, 57, 25, 64, 48, 47, 63,…
## $ `andere PO` <lgl> NA, NA, NA, NA, NA, NA, NA, NA,…
## $ grade <fct> 5.0, 3.7, 2.0, 5.0, 1.3, 2.7, 2…
## $ almost_but_no_cookie <lgl> FALSE, FALSE, FALSE, FALSE, FAL…
# be merciful to these folks who just missed the cut-off...?
dplyr::filter(results, almost_but_no_cookie)## matriculation vn nn po points
## 1 116242 Kofuru Pipavehavavasiy Ethnobotanik 31
## andere PO grade almost_but_no_cookie
## 1 NA 5.0 TRUE
See utils-grading.R for function details.
summarize_grades and publish_grades both put their output on the
clipboard if you run them in interactive()-mode.
These outputs are formatted as an HTML table,
so you can paste them directly into Moodle's HTML editor... :)
# without invalid / no-show exams (by default):
summarized <-
summarize_grades(results, scheme = scheme,
max_points = MAX_POINTS) #, exclude_invalid = TRUE)## Joining, by = "grade"
## HTML table copied to your clipboard.
summarized |> knitr::kable()| Grade | % | Points |
|---|---|---|
| 1.0 | 5.3 | 66.5-75 |
| 1.3 | 15.8 | 62.5-66 |
| 2.0 | 5.3 | 54.5-59 |
| 2.3 | 5.3 | 50.5-54 |
| 2.7 | 10.5 | 45.5-50 |
| 3.0 | 5.3 | 41.5-45 |
| 3.3 | 5.3 | 36.5-41 |
| 3.7 | 10.5 | 34.5-36 |
| 4.0 | 5.3 | 32.5-34 |
| 5.0 | 31.6 | 0-32 |
# all:
summarized_all <-
summarize_grades(results, scheme = scheme, max_points = MAX_POINTS, exclude_invalid = FALSE)## Joining, by = "grade"
## HTML table copied to your clipboard.
summarized_all |> knitr::kable()| Grade | % | Points |
|---|---|---|
| 1.0 | 4.8 | 66.5-75 |
| 1.3 | 14.3 | 62.5-66 |
| 2.0 | 4.8 | 54.5-59 |
| 2.3 | 4.8 | 50.5-54 |
| 2.7 | 9.5 | 45.5-50 |
| 3.0 | 4.8 | 41.5-45 |
| 3.3 | 4.8 | 36.5-41 |
| 3.7 | 9.5 | 34.5-36 |
| 4.0 | 4.8 | 32.5-34 |
| 5.0 | 28.6 | 0-32 |
| invalid | 9.5 | - |
Pseudonymous results with matriculation numbers for your moodle page or public notices:
public <- publish_grades(results, scheme = scheme, max_points = MAX_POINTS)## HTML table copied to your clipboard.
public |> knitr::kable()| Matrikelnummer | Grade | Points |
|---|---|---|
| 105621 | 3.7 | 35 |
| 106914 | 1.3 | 64 |
| 107662 | 4.0 | 34 |
| 109735 | 5.0 | 27 |
| 110417 | 3.3 | 38 |
| 112259 | 3.7 | 35 |
| 116242 | 5.0 | 31 |
| 117454 | 3.0 | 44 |
| 117853 | 1.3 | 66 |
| 120798 | 5.0 | 28 |
| 121289 | invalid | 0 |
| 132477 | invalid | 0 |
| 134927 | 5.0 | 24 |
| 136639 | 1.0 | 73 |
| 137457 | 5.0 | 27 |
| 139899 | 2.7 | 47 |
| 140726 | 2.7 | 48 |
| 141328 | 5.0 | 25 |
| 150027 | 2.3 | 51 |
| 154577 | 2.0 | 57 |
| 156298 | 1.3 | 63 |
# without no-shows / invalidated:
public_all <- results |> dplyr::filter(grade != "invalid") |>
publish_grades(scheme = scheme, max_points = MAX_POINTS) ## HTML table copied to your clipboard.
public_all |> knitr::kable()| Matrikelnummer | Grade | Points |
|---|---|---|
| 105621 | 3.7 | 35 |
| 106914 | 1.3 | 64 |
| 107662 | 4.0 | 34 |
| 109735 | 5.0 | 27 |
| 110417 | 3.3 | 38 |
| 112259 | 3.7 | 35 |
| 116242 | 5.0 | 31 |
| 117454 | 3.0 | 44 |
| 117853 | 1.3 | 66 |
| 120798 | 5.0 | 28 |
| 134927 | 5.0 | 24 |
| 136639 | 1.0 | 73 |
| 137457 | 5.0 | 27 |
| 139899 | 2.7 | 47 |
| 140726 | 2.7 | 48 |
| 141328 | 5.0 | 25 |
| 150027 | 2.3 | 51 |
| 154577 | 2.0 | 57 |
| 156298 | 1.3 | 63 |
results_new <- dplyr::mutate(results,
grade_num = as.numeric(as.character(grade)) * 100,
geschl = "",
abschl = "",
Stg = "",
pversuch = "",
pvermerk = "",
email = "") |>
dplyr::rename(mtknr = "matriculation",
nachname = "nn",
vorname = "vn",
poversion = "po",
bewertung = "grade_num") |>
dplyr::select(mtknr, nachname, vorname, geschl, abschl, Stg,
poversion, pversuch, pvermerk, bewertung, email)## Warning in mask$eval_all_mutate(quo): NAs introduced by
## coercion
results_new |>
print.data.frame(row.names = FALSE, max = nrow(results_new)) |>
clipr::write_clip() ## mtknr nachname vorname geschl abschl Stg poversion
## 120798 Xukuraqus Goy Ethnobotanik
## pversuch pvermerk bewertung email
## 500
## [ reached 'max' / getOption("max.print") -- omitted 20 rows ]
# the last command puts the table on the clipboard,
# so you can <ctrl-v> it directly into the template .xls-fileActually need to generate separate tables for each subject, like so:
for (this_po in unique(results_new$poversion)) {
message("####\n grades for ", this_po, ":\n")
results_new |> dplyr::filter(poversion == this_po) |>
print.data.frame(row.names = FALSE, max = nrow(results_new)) |>
clipr::write_clip()
readline("ready for next group? ")
}not automated, sorry.
CS majors can get their grades from uni2work, maths majors (theoretically) as well (... I think?), the files to upload there expect this format:
results_cs_maths <- dplyr::filter(results,
po != "Statistik als HF")
export_uni2work(results_cs_maths, outfile = "demo-uni2work.csv")## writing to file demo-uni2work.csv.
Now log in at uni2work, create an "external exam", navigate to its "Participants" tab & upload this CSV file. All done.
gpg -e -r notenliste@stat.uni-muenchen.de <YOUR FILE>