lwrc {irtQ}

Lord-Wingersky Recursion Formula

Description

This function computes the conditional distributions of number-correct (or observed) scores given probabilities of category responses to items or given a set of theta values using Lord and Wingersky recursion formula (1984).

Usage

lwrc(x = NULL, theta, prob = NULL, cats, D = 1)

Arguments

x

A data frame containing the item metadata (e.g., item parameters, number of categories, models ...). See irtfit, info or simdat for more details about the item metadata. This data frame can be easily obtained using the function shape_df. If prob = NULL, this data frame is used in the recursion formula. See below for details.

theta

A vector of theta values where the conditional distribution of observed scores are computed. The theta values are only required when a data frame is specified in the argument x.

prob

A matrix containing the probability of answering each category of an item. Each row indicates an item and each column represents each category of the item. When the number of categories differs between items, the empty cells should be filled with zeros or NA values. If x = NULL, this probability matrix is used in the recursion Formula.

cats

A numeric vector specifying the number of categories for each item. For example, a dichotomous item has two categories. This information is only required when a probability matrix is specified in the argument prob.

D

A scaling factor in IRT models to make the logistic function as close as possible to the normal ogive function (if set to 1.7). Default is 1.

Details

The Lord and Wingersky recursive algorithm is an efficient way of calculating the compound probabilities of any number-correct scores on a test based on IRT models. This algorithm is particularly useful when computing the IRT model-based observed score distribution for a test.

To compute the conditional distributions of observed scores, either the item metadata set specified in x or the probability matrix specified in prob can be used.

Value

When the prob argument is provided, this function returns a vector of the probabilities of obtaining every observed score on a test. When the x argument is specified, the function returns a matrix of conditional probabilities across all possible observed scores and theta values.

Author(s)

Hwanggyu Lim [e-mail]

References

Kolen, M. J. & Brennan, R. L. (2004) Test Equating, Scaling, and Linking (2nd ed.). New York: Springer.

Lord, F. & Wingersky, M. (1984). Comparison of IRT true score and equipercentile observed score equatings. Applied Psychological Measurement, 8(4), 453-461.

Examples

## example 1: when a matrix of probabilities is used as a data set
## this is an example from Kolen and Brennan (2004, p. 183)
# create a matrix of probabilities of getting correct and incorrect answers for three items
probs <- matrix(c(.74, .73, .82, .26, .27, .18), nrow=3, ncol=2, byrow = FALSE)

# create a vector of score categories for the three items
cats <- c(2,2,2)

# compute the conditional distributions of observed scores
lwrc(prob=probs, cats=cats)

## example 2: when a matrix of probabilities is used as a data set
## with a mixed-format test
# category probabilities for a dichotomous item
p1 <- c(0.2, 0.8, 0, 0, 0)
# category probabilities for a dichotomous item
p2 <- c(0.4, 0.6, NA, NA, NA)
# category probabilities for a polytomous item with five categories
p3 <- c(0.1, 0.2, 0.2, 0.4, 0.1)
# category probabilities for a polytomous item with three categories
p4 <- c(0.5, 0.3, 0.2, NA, NA)

# rbind the probability vectors
p <- rbind(p1, p2, p3, p4)

# create a vector of score categories for the four items
cats <- c(2, 2, 5, 3)

# compute the conditional distributions of observed scores
lwrc(prob=p, cats=cats)

## example 3: when a data frame for the item metadata of
## a mixed-format test is used.
# import the "-prm.txt" output file from flexMIRT
flex_prm <- system.file("extdata", "flexmirt_sample-prm.txt", package = "irtQ")

# read item parameters and transform them to item metadata
x <- bring.flexmirt(file=flex_prm, "par")$Group1$full_df

# compute the conditional distributions of observed scores
lwrc(x=x, theta=seq(-4, 4, 0.2), D=1)