pairwise (version 0.1.5)

ppML.dicho: ML Person Parameter dichotomous 1PL

Description

Calculation of the person parameters for dichotomous items, given the item parameters and and the datamatrix (argument daten) containing the person respose vectors (rows), according the 1PL Rasch Model using an ML approach.

Usage

ppML.dicho(daten, SIGMA, incidenz = NULL, na_treat = 0,
    limit = 1e-05, iter = 20, tecout = FALSE)

Arguments

daten

a matrix (or data.frame) containing (numeric) responses of the persons (rows) coded in the (0,1) manner.

SIGMA

numeric vector with item (difficulty) parameters.

incidenz

This is only relevant when items are assigned to different booklets. For such a booklet-design a matrix should be assigned to this argument with the same dimensions like daten containig 0 and 1 codes giving the information (for every

na_treat

numeric defining the type of treatment to missing responses in the argument daten. If set to na_treat=0 (default) missing responses are treated as wrong answers. An option is to set na_treat to any value betw

limit

numeric giving the limit at which accuracy the ML-algorithm stops.

iter

numeric giving the maximum numer of iteration to perform.

tecout

logical default set toFALSE. If set to TRUE the result will be a (very) long list with estimation details for every case in daten. In case of a booklet-design the list entrys will be divided by "booklet".

Value

• An object of class "ppMLdicho" or "list" (depending on techout) containing the person parameters.

Details

no detail in the moment.

Examples

############
data(cog);data(cogBOOKLET) # loading reponse and allocation data
d<-(cog[cog$BOOKID!=14,]) # skip persons which got booklet No.14.
inc<-make.incidenz(tab=cogBOOKLET, bookid=d$BOOKID) # make just the incidenz matrix
result<-ppML.dicho(daten=d[,4:34], SIGMA=itempar.dicho(d[,4:34]),incidenz = inc )
summary(result)
summary(result,FALSE) # return just the ML person estimates and their standard errors.