PRV: This function calculates the proportion reduction in variance...

In mustfa5/test.adaptation: Adaptation - level of adaptation of a computerized adaptive test

Description

This function calculates the proportion reduction in variance (PRV) index as described in Reckase, Ju, and Kim (2019)

Usage

PRV(diff, N, n.items, item.diff)

Arguments

diff	a data frame in which rows represent test takers while columns represent the difficulty level of items that were administered to that test taker
N	total number of test takers who participated in a CAT that uses a specific CAT pool
n.items	test length
item.diff	item difficulties administered to a test taker during a computerized test administration

Value

This function returns numeric value

References

Reckase, M. D., Ju, U., & Kim, S. (2019). How adaptive is an adaptive test: Are all adaptive tests adaptive?. Journal of Computerized Adaptive Testing, 7(1), 1-14.

Examples

library(catR)
N=1000 #number of students
bank=250 #number of items
items=45
theta=rnorm(N,0,1) #level of trait
model="2PL" #IRT model to use
start <- list(theta = -1:1, randomesque = 1)
stop <- list(rule = c( "length"), thr = items)
final <- list(method = "ML")

test=list(method = "ML", itemSelect = "MFI")
bank=genDichoMatrix(items =bank, cbControl = NULL,
                    model = model)

res <- simulateRespondents(thetas = theta, bank,
                           start = start, test = test, stop = stop,
                           final = final, model = NULL)
t.hat=res$final.values.df$estimated.theta

items.administered=res$responses.df[,grepl("items.administrated",
                                           names( res$responses.df ) ) ]
colnames(items.administered)=NULL
item.diff=matrix(ncol = ncol(items.administered),nrow = nrow(items.administered))
for (k in 1:nrow(items.administered)) {
  xx= as.numeric(items.administered[k,])
  item.diff[k,]=bank[xx,2]
}
PRV(diff = item.diff, N=N, n.items = items, item.diff = bank$b )

mustfa5/test.adaptation documentation built on Dec. 21, 2021, 11:03 p.m.