empICC: Empirical ICCs
In mihock/iana: GUI for Item Analysis

empICC

R Documentation

Empirical ICCs

Description

Plot of empirical ICCs of the items in a data frame.

Usage

empICC(
  x,
  score = c("factor.thomson", "factor.bartlett", "mean", "sum"),
  method = "loess",
  span = 0.75,
  alpha = 0,
  jitter = 0.4
)

Arguments

`x`	data frame in which rows are subjects and colums are test items
`score`	total score to use
`method`	function to use for curve fitting, e.g. "loess" or "lm"
`span`	degree of smoothing used for loess (see `loess`)
`alpha`	opaqueness of the points in the scatterplot
`jitter`	amount of jitter

Details

To examine item characteristics, item scores are plotted against total scores or factor scores. To avoid overplotting, a small amount of jitter is added to overlapping points. In the upper left corner, the correlation of the total/factor score and the item score is given. The lines are locally weighted regression lines, the shaded regions represent 95% confidence intervals around the expected item scores. If many data points are present it may be useful to decrease the opaqueness of the points and/or to deactivate jitter.