klausur.data: A function to create data objects with given and correct...

In klausuR: Multiple Choice Test Evaluation

A function to create data objects with given and correct answers to a test.

Description

klausur.data automatically parses the variable names in answto decide which variables are actual test items, if they are named according to the given scheme Item###. To help in constructing a data.frame with correct column names one can call the klausur.gen utility to generate an empty data object of a given number of items and test subjects.

Usage

klausur.data(
  answ,
  corr,
  items = NULL,
  marks = NULL,
  wght = NULL,
  corr.key = NULL,
  rename = c(),
  dummies = c(),
  disc.misc = FALSE,
  na.rm = TRUE,
  item.prefix = c(),
  sort.by = "Name",
  maxp = NULL,
  wrong = NULL,
  keep.cases = NULL,
  recode.na = 0
)

Arguments

answ	A data.frame which has to include at least these variables: No, Name, FirstName, MatrNo, as well as Pseudonym (optional) and variables for the answered items (according to the scheme Item###, where ### is a number with leading zeros, if needed).
corr	A vector with the correct answers to all items in answ (named also according to Item###).
items	Indices of a subset of variables in answ to be taken as items.
marks	A vector assigning marks to points achieved (see details). Leave NULL if not available.
wght	A vector with weights for each item (named also according to Item###). Leave NULL if not available.
corr.key	If test has several test forms: A data.frame or matrix indicating the positions of all items (columns) in all forms (rows). Must have a column called Form (like answ), and the item columns must follow the explained name scheme Item###. NULL if not needed.
rename	A named vector defining if variables in answ need to be renamed into the klausuR name scheme. Accepts elements named No, Name, FirstName, MatrNo, Pseudonym and Form. The values of these elements represent the variable names of the input data.
dummies	A vector of dummy variables to be created, e.g. if you don't need/want actual data in the id slot. Can include "No", "Name", "FirstName", "MatrNo" and "Pseudonym". Columns will just be filled with increasing integers.
disc.misc	Logical. If TRUE, left over columns from answ will not be stored in slot misc but silently discarded.
na.rm	Logical, whether cases with NAs should be ignored in answ. Defaults to TRUE.
item.prefix	A named character vector with two optional elements, item and corr, defining the name prefix used for the items in the test data and the vector with correct answers, respectively. Defaults to item="Item" and corr="Item".
sort.by	A character string naming the variable to sort the answ data by. Set to c() to skip any re-ordering.
maxp	Optional numeric value, if set will be forced as the maximum number of points achievable. This should actually not be needed, if your test has no strange errors. But if for example it later turns out you need to adjust one item because it has two instead of one correct answers, this option can become handy in combination with "partial" scoring and item weights.
wrong	If you want full pick-n scoring: A vector similar to corr, but this time listing all alternatives that are wrong.
keep.cases	A vector of MatrNo values, if you want to prevent these cases from being dropped even if they contain missing data. If not NULL, missing values in all test items are replaced by the value given to recode.na, before na.rm is evaluated.
recode.na	A value to replace missing data with in all cases specified by keep.cases. Ignored if keep.cases=NULL.

Details

If you have items with multiple correct answers you can easily code these as one single item: All alternatives a subject has marked should be combined to a single value without spaces. The vector with correct answers will have to be coded accordingly, of course. An example: If someone marked the first, third and fourth answer, you would code this as "134". See klausur.gen.corr for a helpful function to create such an answer vector. Internally klausur checks for equality of given answers and correct values, that is, it will only give that person a point if the correct answer was coded as "134" as well.

Data for (Number Right) Elimination Testing

If your test is to be evaluated according to elimination testing (ET), number right elimination testing (NRET) or number right (NR, which is actually multiple choice) scoring, the data has to be in a different format: In contrast to the usual MC procedure, ET items are answered by eliminating all alternatives a subject considers wrong; in an NRET test subjects are asked to eliminate all wrong alternatives and mark the one they consider the correct answer. That is, for both scoring functions, you need to know for each answer alternative whether a subject saw it as right, wrong or was not sure and left it open.

In this implementation, these answers are to be coded as a plus sign "+" (right answer), a minus sign "-" (wrong answer) or a zero "0" (missing). If you need to code errors (like both "right" and "wrong" have been marked), use the asterisk "*" for these cases. All answers to one item belong into one column. E.g., if you have four answer alternatives, a subject thought the second one to be the correct answer and eliminated the rest, you'd have to code this item as "-+--". The same is true for the vector of correct answers, of course.

Marks

The assigned marks are expected to be in a certain format as well, as long as you don't want klausur to suggest them itself. Just create an empty vector to start with (say your.marks <- c()) and fill it according to the scheme your.marks[<points from>:<points to>] <- <mark>. For example: Should one get a 1.7 if in sum 27 to 30 points were achieved, you'd assign these points as indices to the vector with your.marks[27:30] <- "1.7" (see example section below). It is crucial to assign marks to the whole range of points that can be achieved in the test. On the other hand, it's irrelevant wheter you assign decimal marks as in the example, only integer values, a 15 marks scheme or whatever. The convenience function klausur.gen.marks can assist you in creating such a valid vector.

Value

An object of class klausuR.answ-class.

Examples

data(antworten)

# vector with correct answers:
richtig <- c(Item01=3, Item02=2, Item03=2, Item04=2, Item05=4,
 Item06=3, Item07=4, Item08=1, Item09=2, Item10=2, Item11=4,
 Item12=4, Item13=2, Item14=3, Item15=2, Item16=3, Item17=4,
 Item18=4, Item19=3, Item20=5, Item21=3, Item22=3, Item23=1,
 Item24=3, Item25=1, Item26=3, Item27=5, Item28=3, Item29=4,
 Item30=4, Item31=13, Item32=234)

# vector with assignement of marks:
notenschluessel <- c()
# scheme of assignments: marks[points_from:to] <- mark
notenschluessel[0:12]  <- 5.0
notenschluessel[13:15] <- 4.0
notenschluessel[16:18] <- 3.7
notenschluessel[19:20] <- 3.3
notenschluessel[21]    <- 3.0
notenschluessel[22]    <- 2.7
notenschluessel[23]    <- 2.3
notenschluessel[24]    <- 2.0
notenschluessel[25:26] <- 1.7
notenschluessel[27:29] <- 1.3
notenschluessel[30:32] <- 1.0

# now combine all test data into one object of class klausur.answ
data.obj <- klausur.data(answ=antworten, corr=richtig, marks=notenschluessel)

# if that went well, get the test results
klsr.obj <- klausur(data.obj)

klausuR documentation built on April 5, 2022, 1:15 a.m.