entropy.lca {CDM} R Documentation

## Test-specific and Item-specific Entropy for Latent Class Models

### Description

Computes test-specific and item-specific entropy as test-diagnostic criteria of cognitive diagnostic models (Asparouhov & Muthen, 2014).

### Usage

entropy.lca(object)

## S3 method for class 'entropy.lca'
summary(object, digits=2,  ...)


### Arguments

 object Object of class din, gdina or mcdina. For the summary method, it is the result of entropy.lca. digits Number of digits to round ... Further arguments to be passed

### Value

A list with the data frame entropy as an entry.

### References

Asparouhov, T. & Muthen, B. (2014). Variable-specific entropy contribution. Technical Appendix. http://www.statmodel.com/7_3_papers.shtml

See cdi.kli for test diagnostic indices based on the Kullback-Leibler information and cdm.est.class.accuracy for calculating the classification accuracy.

### Examples

#############################################################################
# EXAMPLE 1: Entropy for DINA model
#############################################################################

data(sim.dina, package="CDM")
data(sim.qmatrix, package="CDM")

# fit DINA Model
mod1 <- CDM::din( sim.dina, q.matrix=sim.qmatrix, rule="DINA")
summary(mod1)
# compute entropy for test and items
emod1 <- CDM::entropy.lca( mod1 )
summary(emod1)

## Not run:
#############################################################################
# EXAMPLE 2: Entropy for polytomous GDINA model
#############################################################################

data(data.pgdina, package="CDM")

dat <- data.pgdina$dat q.matrix <- data.pgdina$q.matrix

# pGDINA model with "DINA rule"
mod1 <- CDM::gdina( dat, q.matrix=q.matrix, rule="DINA")
summary(mod1)

# compute entropy
emod1 <- CDM::entropy.lca( mod1 )
summary(emod1)

#############################################################################
# EXAMPLE 3: Entropy for MCDINA model
#############################################################################

data(data.cdm02, package="CDM")

dat <- data.cdm02$data q.matrix <- data.cdm02$q.matrix

# estimate model with polytomous atribute
mod1 <- CDM::mcdina( dat, q.matrix=q.matrix )
summary(mod1)
# computre entropy
emod1 <- CDM::entropy.lca( mod1 )
summary(emod1)

## End(Not run)


[Package CDM version 8.2-6 Index]