Bland-Altman plot.

The Bland-Altman plot (AKA Tukey mean-difference plot) is a graphical tool to measure agreement between two estimates of a single variables methods.

blandAltman(x1, x2, xlab = "Average", ylab = "Difference", alpha = 0.05)

Arguments

x1	the first estimate
x2	the second estimate
xlab	a label of x-axis on which the mean of the two estimates is plotted
ylab	a label of y-axis on which the difference between the two estimates is plotted
alpha	significance level of the confidence interval

Details

Code taken from http://rstats.tiddlyspot.com/#Bland-Altman and updated slightly.

Value

a list of 'mean' (the of the two estimates), 'diff' (the diff of the estimates, 'se' (the standard error of the differences), and the '1-alpha/2' confidence intervals for the 'mean +/- 2 * SD of diffs' given as 'lower' and 'upper'

References

Bland JM, Altman DG (1986) _Statistical methods for assessing agreement between two methods of clinical measurement.__ Lancet 327 (8476): 30710.

Examples

n<-20
x<-rnorm(n)
x1<-x+rnorm(n,0,.1)
x2<-x+rnorm(n,0,.1)
blandAltman(x1,x2)
#> $mean
#>  [1] -1.37313405  0.24150845 -2.50498903  0.01874336  0.73821782  1.28131023
#>  [7] -1.86628346 -0.21923736 -0.27962931 -0.45391146 -0.46382690  0.62564464
#> [13]  2.06690481 -1.54885308  0.51950689 -1.84221997 -0.73122263  0.06756409
#> [19]  0.50926298 -0.85004165
#> 
#> $diff
#>  [1]  0.039811957  0.100207458 -0.125457865  0.098925976  0.144370781
#>  [6] -0.285286268 -0.098237873 -0.059365938 -0.094498383  0.039932084
#> [11]  0.007327671  0.041972535  0.044977266  0.160437129  0.008247733
#> [16] -0.068382455  0.036402739 -0.296179443  0.004777528  0.085395213
#> 
#> $se
#> [1] 0.02769502
#> 
#> $lower
#> [1] -0.3395950 -0.1772907
#> 
#> $upper
#> [1] 0.1558285 0.3181328
#>