itemstats {mirt} | R Documentation |
Function to compute generic item summary statistics that do not require prior fitting of IRT models. Contains information about coefficient alpha (and alpha if an item is deleted), mean/SD and frequency of total scores, reduced item-total correlations, average/sd of the correlation between items, response frequencies, and conditional mean/sd information given the unweighted sum scores.
itemstats( data, group = NULL, use_ts = TRUE, proportions = TRUE, ts.tables = FALSE )
data |
An object of class |
group |
optional grouping variable to condition on when computing summary information |
use_ts |
logical; include information that is conditional on a meaningful total score? |
proportions |
logical; include response proportion information for each item? |
ts.tables |
logical; include mean/sd summary information pertaining to the unweighted total score? |
Returns a list containing the summary statistics
Phil Chalmers rphilip.chalmers@gmail.com
Chalmers, R., P. (2012). mirt: A Multidimensional Item Response Theory Package for the R Environment. Journal of Statistical Software, 48(6), 1-29. doi: 10.18637/jss.v048.i06
# dichotomous data example LSAT7full <- expand.table(LSAT7) head(LSAT7full) itemstats(LSAT7full) # behaviour with missing data LSAT7full[1:5,1] <- NA itemstats(LSAT7full) # data with no meaningful total score head(SAT12) itemstats(SAT12, use_ts=FALSE) # extra total scores tables dat <- key2binary(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)) itemstats(dat, ts.tables=TRUE) # grouping information group <- gl(2, 300, labels=c('G1', 'G2')) itemstats(dat, group=group) ##### # polytomous data example itemstats(Science) # polytomous data with missing newScience <- Science newScience[1:5,1] <- NA itemstats(newScience) # unequal categories newScience[,1] <- ifelse(Science[,1] == 1, NA, Science[,1]) itemstats(newScience) merged <- data.frame(LSAT7full[1:392,], Science) itemstats(merged)