[R] [R-pkgs] MiscPsycho Package 1.0

Doran, Harold HDoran at air.org
Tue May 8 17:59:22 CEST 2007

I have just submitted MiscPsycho to CRAN. MiscPsycho contains functions
for miscellaneous psychometrics that may be useful for applied
psychometricians. MML estimation already exists in the ltm package.
Hence, a jml option is provided for users who prefer this method. The
jml function gives back rasch difficulties and the same Infit and Outfit
statistics as Winsteps. 

Also, jml is known to return biased parameter estimates. So, a
correction for jml bias is provided for as an optional argument in the
function. Currently, the package includes the following:

1) jml() a function for joint maximum likelihood estimation of the Rasch
2) alpha() , coefficient alpha
3) alpha.Summary(), gives what alpha would be if item i were excluded
3) scoreCon(), create score conversion tables, 
4) class.acc(), classification accuracy statistics 
5) simRasch(), simulate Rasch item responses 
6) classical(), point-biserial and p-values 

Future versions will include Master's partial credit model, DIF
statistics, procedures for linking scales (e.g., Stocking-Lord), lattice
graphics for item characteristic curves, and bayesian scoring methods

Below is a sample session:

set.seed(1) # for replication only
xx <- simRasch(200, 10) # Simulate data for 200 persons and 10 items
classical(xx$data) # get p-values and point-biserials
alpha(xx$data) # coefficient alpha
aa <- jml(xx$data) # estimate item parameters

The student in row 1 has a number correct score of 3 out of 10. From the
score conversion table, his theta estimate is -1.36. So, we can ask,
what is the probability that his true score is above, say, .5 on the
theta scale.

class.acc(as.numeric(xx$d[1,]), aa$params, .5, aboveQ=T)

Any recommendations, bugs, hints, comments, are welcome.

Harold Doran

	[[alternative HTML version deleted]]

R-packages mailing list
R-packages at stat.math.ethz.ch

More information about the R-help mailing list