reliability indices

Description Usage Arguments Details Value References Examples

Modified Bland-Altman plot for multiple raters

BAplotMultipleR(rater, subject, variable, data, LoA = c("Default", "loess"),
  ArgzLoess = list(span = 1, degree = 1), symbols.graph = NULL,
  colSymbols = NULL, Legend = T, ArgzLegend = list(x = "topright"),
  xlab = "Average", ylab = "Difference", ...)

`rater`	Variable indicating which rater made the rating.
`subject`	Variable for the subject.
`variable`	Variable containing the ratings of the subjects.
`data`	The dataframe
`LoA`	The limits of agreement that have to be used. See Details.
`ArgzLoess`	Arguments for `loess.smooth` if `LoA = "loess"`.
`symbols.graph`	Optional, the symbols to be used in the graph for the raters.
`colSymbols`	Optional, the color of the symbols
`Legend`	Logical, indicates whether a legend has to be printed. Default is `TRUE`.
`ArgzLegend`	The arguments for the legend, see `legend`.
`xlab`	The label for the x-axis.
`ylab`	The label for the y-axis.
`...`	Additional arguments to be passed to `plot`.

The limits of agreement (LoA) have to be interpreted differently than in Bland-Altman plots with 2 raters. In the modified Bland-Altman plots, the LoA indicate how different an individual rater can be when compared with the mean of all the raters (Jones et al., 2011). However, as with the Bland-Altman plots for 2 raters, it may be that the variability of the differences increase as the magnitude of the measurement increases (Bland and Altman, 1999). Alternative LoA can then be plotted using a method based on the method of Royston and Wright (1998). Specifying LoA="loess", we get approximate 2.5 and 97.5 percentile curves. In this method, the standard deviation of the difference scores per subject is calculated and the LoA per subject are calculated as -/+ 1.96 times the standard deviation. A loess fit is then used to connect these LoA per subject. Note that we assume that the difference scores are normally distributed (with mean 0 as we are working with centered values).

Returns a list with the following objects

`AnovaSummary`	The summary of the two-way ANOVA
`AvgPerSubject`	The average score per subject.
`LoA`	The limits of agreement calculated according to Jones et al. (2011).
`ICC`	The intraclass correlation coefficients. See `ICC`

Bland, J.M., Altman, D.G.(1999). Measuring agreement in method comparison studies. Statistical Methods In Medical Research, Vol.8(2), pp. 135-160

Jones, M., Dobson, A., O'Brian, S. (2011). A graphical method for assessing agreement with the mean between multiple observers using continuous measures. Int J Epidemiol. Vol.40: pp. 1308-1313.

Royston, P., Wright, E.M. (1988). How to construct 'normal ranges' for fetal variables. Ultrasound Obstet Gynecol, Vol.11: pp. 30-38.