klausur.gen.marks: Generate mark assignments

Description

Create a vector of marks to be used by klausur.

Usage

klausur.gen.marks(
  mark.labels = NULL,
  answ = NULL,
  wght = NULL,
  suggest = list(mean = NULL, sd = NULL),
  minp = 0
)

Arguments

mark.labels

Either a vector with labels (names) for all marks that should be assigned to certain test scores, or one of the arguments 6, 11, 16, "DIHK", "UK", "USA" or "A" (see Details). If NULL, you will be asked to type in labels.

answ

Either an object with item names in klausuR scheme (see klausur), e.g. your observation data, or an integer representing the maximum score of the test. If NULL, you will be asked for the maximum score.

wght

A vector with weights for each item. If NULL, the number of items is used as maximum score, that is, each item gives a point.

suggest

A list with the elements mean and sd. If both are not NULL, this function will suggest marks for achieved points assuming normal distribution. That is, "mean" and "sd" should be set to the corresponding values of the test's results.

minp

An integer value, in case there is a minimum score no-one can fall below (which can happen, e.g., with ET/NRET scoring and different numbers of answer alternatives). Should be left as is in most cases.

Value

A character vector.

Details

If mark.labels is set to one of the arguments 6, 11, 16, "DIHK", "UK", "USA" or "A", often used schemes for marks will be used as a preset. In case of mark.labels=6, marks will go from 1 (best) to 6 (failed), as widely used in German schools. In case of mark.labels=11, marks will range from 1.0 (best) to 5.0 (failed), with the marks 1 through 3 being split into three decimal steps .0, .3 and .7, as is often used in academic institutions. In case of mark.labels=16, marks will be a range of points from 15 (best) to 0 (failed), as often used in German gymnasiums. If mark.labels="A", marks A to F are given.

For the other cases some more probably useful assumptions are being made, which percentage of achieved points leads to which mark. If mark.labels="DIHK", marks will be 1 through 6, and calculated according to usual standards of the Deutsche Industrie- und Handelskammer (1 > 92%, 2 > 81%, 3 > 67%, 4 > 50%, 5 > 30%, 6 below that). If mark.labels="UK", marks are A > 90%, B > 65%, C > 35%, D > 10% and E below that, and for mark.labels="USA" it's A > 90%, B > 80%, C > 70%, D > 60% and F below that. Please note that the percentages indicate individual test results and not "the best X percent of the sample". If you'd rather use your own system, either declare it as a vector, or leave as NULL, and you'll be asked (be sure to begin with the worst mark!).

The parameter answ is quite versatile as well. You can just feed it your observation data, if it complies with the naming scheme for items (Item###, see also klausur.gen), and klausur.gen.marks will calculate the maximum score automatically. Or you assign the maximum directly as an integer value.

Another feature can be toggled with the parameter suggest. If you feed it with the mean and standard deviation values of your test's results, marks are automatically assigned to the achieved score under the assumption of normal distribution. Please understand that the naming "suggest" is not an accident! This is only a suggestion, please review it, tweak it, revise it, until it fits your needs. However, this feature can directly be called by klausur.

Examples

Run this code

# NOT RUN {
notenschluessel <- klausur.gen.marks(mark.labels=11,answ=antworten)
# }