Entropy_plot: Plot item entropy curves for selected items or questions.

In TestGardener: Information Analysis for Test and Rating Scale Data

Plot item entropy curves for selected items or questions.

Description

Item the value of the entropy curve at a point theta is the expected value of the surprisal curve values. Entropy is a measure of the randomness of the surprisal value, which is maximized when all the surprisal curves have the same value and has a minimum of zero if all but a single curve has probability zero. This is unattainable in the calculation, but can be arbitrarily close to this state.

Usage

  Entropy_plot(scrfine, SfdList, Qvec, dataList, plotindex=1:n, 
               plotrange=c(min(scrfine),max(scrfine)), height=1.0, value=0,  
               ttlsz=NULL, axisttl=NULL, axistxt=NULL)

Arguments

scrfine	A vector of length nfine (usually 101) containing equally spaced points spanning the plotrange. Used for plotting.
SfdList	A numbered list object produced by a TestGardener analysis of a test. Its length is equal to the number of items in the test or questions in the scale. Each member of SfdList is a named list containing information computed during the analysis.
Qvec	The five marker percentile values.
dataList	A list vector containing objects essential to an analysis.
plotindex	A set of integers specifying the numbers of the items or questions to be displayed.
plotrange	A vector of length 2 containing the plot boundaries within or over the score index interval c(0,100).
height	A positive real number defining the upper limit on the ordinate for the plots.
value	Number required by ggplot2. Defaults to 0.
ttlsz	Title font size.
axisttl	Axis title font size.
axistxt	Axis text(tick label) font size.

Details

An entropy curve for each question indexed in the index argument. A request for a keystroke is made for each question. The answer to question strongly defines the optimal position of an estimated score index value where the curve is high value. Values of entropy curves typically range over [0,1].

Value

The plots of the entropy curves specified in plotindex are produced as a side effect. If saveplot is TRUE, the plots of item entropy curves specified in plotindex are bundled into a single postscript or .pdf file and the file name is defined by paste(dataList$titlestr,i,'-entropy.pdf',sep=""). The file is then output as a returned value.

Author(s)

Juan Li and James Ramsay

References

Ramsay, J. O., Li J. and Wiberg, M. (2020) Full information optimal scoring. Journal of Educational and Behavioral Statistics, 45, 297-315.

Ramsay, J. O., Li J. and Wiberg, M. (2020) Better rating scale scores with information-based psychometrics. Psych, 2, 347-360.

See Also

Sensitivity_plot, Power_plot, Ffuns_plot, ICC_plot

Examples

  #  Example 1.  Display the item entropy curves for the 
  #  short SweSAT multiple choice test with 24 items and 1000 examinees
  #  plot the entropy curve for the first item
  dataList <- Quant_13B_problem_dataList
  SfdList  <- Quant_13B_problem_parmList$SfdList
  Qvec     <- Quant_13B_problem_parmList$Qvec
  scrfine  <- seq(0,100,len=101)
  oldpar   <- par(no.readonly=TRUE)
  Entropy_plot(scrfine, SfdList, Qvec, dataList, plotindex=1)
  par(oldpar)

TestGardener documentation built on Nov. 24, 2023, 5:08 p.m.