Calculate National Trends in International Large-Scale Assessments

Share:

Description

This function processes LSA data and returns trend estimates in latent variable means and proficiency level frequencies.

Usage

1
2
3
4
5
6
eatTrend(itParsIntT1, PVsT1, countriesT1, itParsNatT1=NULL, 
jkzoneT1=NULL, jkrepT1=NULL, weightsT1=NULL, groupsT1=NULL, itParsIntT2, PVsT2, 
countriesT2, itParsNatT2=NULL, weightsT2=NULL, jkzoneT2=NULL, jkrepT2=NULL, 
groupsT2=NULL, GES=TRUE, testletNam = NULL, transfTo500=TRUE, mtT=500, sdtT=100, 
mRefPop=NULL, sdRefPop=NULL, cutScores=NULL, type =c("FCIP", "MM"), writeCsv=FALSE, 
path=NULL, plots=FALSE, backwards=FALSE, groupNam=NULL, landNam=TRUE, FRZ=FALSE)

Arguments

itParsIntT1

A data frame containing the international item parameter estimates at time 1. The first column has to be an ID variable, the second column contains the estimates.

PVsT1

A data frame containing plausible values for all persons at time 1. The first column has to be an ID variable, the following columns contain plausible values.

countriesT1

A vector containing the group membership for every person at time 1. The order has to correspond to the ID column in PVsT1.

itParsNatT1

A list of data frames. The lists' names should correspond to the groups specified in countriesT1. The first column of every data frame has to be an ID variable, the second column contains the national item parameter estimates.

jkzoneT1

A vector containing the primary sampling unit at time 1. The order has to correspond to the ID column in PVsT1. If empty a nonsene default will be set.

jkrepT1

A vector containing the jackknife replicate ID at time 1. The order has to correspond to the ID column in PVsT1. If empty a nonsene default will be set.

weightsT1

A vector containing the case weights at time 1. The order has to correspond to the ID column in PVsT1.

groupsT1

A vector or a data.frame indicating subgroup-membership ("1") or subgroup- nonmembership ("0") at time 1.

itParsIntT2

A data frame containing the international item parameter estimates at time 2. The first column has to be an ID variable, the second column contains the estimates. Please assure that the item parameters stem from an analysis in which the person parameters were centered around zero or that an equivalent adjustment has taken place.

PVsT2

A data frame containing plausible values for all persons at time 2. The first column has to be an ID variable, the following columns contain plausible values. Please note that the columns should be centered around zero.

countriesT2

A vector containing the group membership for every person at time 2. The order has to correspond to the ID column in PVsT2.

itParsNatT2

A list of data frames. The lists' names should correspond to the groups specified in countriesT2. The first column of every data frame has to be an ID variable, the second column contains the national item parameter estimates.

jkzoneT2

A vector containing the primary sampling unit at time 2. The order has to correspond to the ID column in PVsT2. If empty a nonsene default will be set.

jkrepT2

A vector containing the jackknife replicate ID at time 2. The order has to correspond to the ID column in PVsT2. If empty a nonsene default will be set.

weightsT2

A vector containing the case weights at time 2. The order has to correspond to the ID column in PVsT2.

groupsT2

A vector or a data.frame indicating subgroup-membership ("1") or subgroup- nonmembership ("0") at time 2.

GES

Logical. If TRUE, trends for the whole sample will be computed.

testletNam

Character vector containing the names of the units/testlets in which items are clustered.

mtT

Numeric. Mean of arbitrary metric to which parameters are transformed.

sdtT

Numeric. Standard deviation of arbitrary metric to which parameters are transformed.

mRefPop

Numeric. The mean of the reference population.

sdRefPop

Numeric. The standard deviation in the reference population.

transfTo500

Logical. If TRUE, the function transformTo500 will be called.

cutScores

Named vector indicating the proficiency level cut scores in descending(!) order. If is not NULL, the function addCuts will be called.

type

Character. If MM, all linking is done via mean-mean-equating (recommended for diagnostic purposes). If FCIP, international and national item parameters will be treated as the same at every time point.

writeCsv

Logical. If TRUE, .csv files will be written.

path

A character string containing the path for the output files.

plots

Logical. If TRUE, country DIF will be plotted.

backwards

Logical. IQB internal option to transform cutScores internally.

groupNam

A character vector containing group names.

landNam

Logical. Bundesland-Klarnamen.

FRZ

Logical. Standardisiert PV-weise wenn FALSE.

Value

A list with two data frames: Countrywise trends in means and proficiency levels.

Author(s)

Karoline Sachse