isop.calc2set {ISOP}R Documentation

ISOP model for two sets of items

Description

The function calculates nonparametric indices for testing the validity of an ISOP-Model, suggested by Scheiblechner, for two given sets of items. Having one item in criterion and rest of the items in predictor set can be used for item selection. Having many predictors and many criterions corresponds to canonical correlation.

Usage

isop.calc2set(raw, set1, set2, verbose = TRUE)

Arguments

raw

The data, ISOP should be applied on

set1

Indices of items in criterion set

set2

Indices of items in predictor set

verbose

Print information about progress

Value

ns, nd, nt, nss, ndd, nsd

Frequencies of concordant, discordant and tied vectors

ps, pd, pt, pss, pdd, psd

Probabilities of concordant, discordant and tied vectors

predictability

Multivariate index of ordinal association. Varies between -1 and +1. Does not include ties.

stddev

Standard devation of predictability

isotonicity

Nonparametric measure of positive diagonalization of an ordered contingency table. Varies between 0 and +1. Includes ties.

compatibility

Number of pairs concordant or tied

n

Number of response vectors

z

z-value of predictability

stderr

Standard error of predictability

ubas

Upper bound of asymptotic

Author(s)

Stefan Nachtnebel

References

Hartmann Scheiblechner (2003) Nonparametric IRT: Testing the bi-isotonicity of isotonic probabilistic models (ISOP) Psychometrika, 68, 1, 79–96

See Also

isop.calc1set


[Package ISOP version 1.0 Index]