dirf {lazy.irt}

R Documentation

Calculation of Derivative of Item Response Function

Description

Calculation of Derivative of Item Response Function

Usage

dirf(
  param,
  theta = NULL,
  weight = NULL,
  zero = 1,
  smallP = 1e-09,
  thmin = -4,
  thmax = 4,
  npoints = 21,
  DinP = 1,
  numderiv = 0,
  eps = 1e-06,
  log = 0,
  print = 1,
  debug = 0,
  plot = 0
)

Arguments

param	Item Parameter Data Frame
theta	Discrete theta values
weight	Weight data frame
zero	= 0 to exclude the xzero-th category from output
smallP	Minimum value of probability
thmin	Minimum value of discrete thata value
thmax	Maximum value of discrete thata value
npoints	# of discrete points for theta
DinP	= 1 to include D=1.7 in logistic function
numderiv	= 1 to use numerical derivatives
eps	eps for JacobianMat
log	= 1 to calculate log derivatives
print	= 1 to print result
debug	= 1 to print intemediate result
plot	= 1 to plot result

Value

A list of
dICRF, dIRF, dTRF, fromP, toP=toP, vecv, minscore_i, maxscore_i, minscore_t, maxscore_tt, log
where
dICRF npoints x sum(ncat)
dIRF npoints x nitems weighted by item category weight
dTRF npoints x 1 weighted by item category weight
and item weight
fromP, toP location of each item category in dICRF
vectorize category weights
minscore_i mimimum score of each item
maxscore_i maximum score of each item
minscore_t mimimum score of test
maxscore_t maximum score of test
Note that when log=1, dICRF etc are the log derivatives, namely, the derivative of log ICRF w.r.t. theta, etc.

Examples

dirf( paramS1, plot=1 )

# compare analytic and numeric derivative
res1=dirf( paramS1, print=0, plot=0 )$dICRF
res2=dirf( paramS1, print=0, numderiv=1, plot=0 )$dICRF
Print(max(abs(res1-res2)))

---