mirt (version 1.42)

key2binary: Score a test by converting response patterns to binary data

Description

The key2binary function will convert response pattern data to a dichotomous format, given a response key.

Usage

key2binary(fulldata, key, score_missing = FALSE)

Value

Returns a numeric matrix with all the response patterns in dichotomous format

Arguments

fulldata

an object of class data.frame, matrix, or table with the response patterns

key

a vector or matrix consisting of the 'correct' response to the items. Each value/row corresponds to each column in fulldata. If the input is a matrix, multiple scoring keys can be supplied for each item. NA values are used to indicate no scoring key (or in the case of a matrix input, no additional scoring keys)

score_missing

logical; should missing data elements be returned as incorrect (i.e., 0)? If FALSE, all missing data terms will be kept as missing

Author

Phil Chalmers

References

Chalmers, R., P. (2012). mirt: A Multidimensional Item Response Theory Package for the R Environment. Journal of Statistical Software, 48(6), 1-29. tools:::Rd_expr_doi("10.18637/jss.v048.i06")

Examples

data(SAT12)
head(SAT12)
key <- c(1,4,5,2,3,1,2,1,3,1,2,4,2,1,5,3,4,4,1,4,3,3,4,1,3,5,1,3,1,5,4,5)

dicho.SAT12 <- key2binary(SAT12, key)
head(dicho.SAT12)

# multiple scoring keys
key2 <- cbind(c(1,4,5,2,3,1,2,1,3,1,2,4,2,1,5,3,4,4,1,4,3,3,4,1,3,5,1,3,1,5,4,5),
              c(2,3,NA,1,rep(NA, 28)))
dicho.SAT12 <- key2binary(SAT12, key2)

# keys from raw character responses
resp <- as.data.frame(matrix(c(
  "B","B","D","D","E",
  "B","A","D","D","E",
  "B","A","D","C","E",
  "D","D","D","C","E",
  "B","C","A","D","A"), ncol=5, byrow=TRUE))

key <- c("B", "D", "D", "C", "E")

d01 <- key2binary(resp, key)
head(d01)

# score/don't score missing values
resp[1,1] <- NA
d01NA <- key2binary(resp, key) # without scoring
d01NA

d01 <- key2binary(resp, key, score_missing = TRUE) # with scoring
d01