error.crosses {psych} | R Documentation |
Given two vectors of data, plot the means and show standard errors in both X and Y directions.
error.crosses(x, y, labels = NULL, pos = NULL, arrow.len = 0.2, ...)
x |
A vector of summary statistics (from Describe) |
y |
A second vector of summary statistics (also from Describe) |
labels |
name the pair |
pos |
Labels are located where with respect to the mean? |
arrow.len |
Arrow length |
... |
Other parameters for plot |
For an example of two way error bars describing the effects of mood manipulations upon positive and negative affect, see http://personality-project.org/revelle/publications/happy-sad-appendix/FIG.A-6.pdf
The second example shows how error crosses can be done for multiple variables where the grouping variable is found dynamically.
William Revelle
revelle@northwestern.edu
To draw error bars for single variables error.bars
, or by groups error.bars.by
, or to find descriptive statistics describe
or descriptive statistics by a grouping variable describe.by
desc <- describe(attitude) x <- desc[1,] y <- desc[2,] plot(x$mean,y$mean,xlab=rownames(x),ylab=rownames(y)) #in graphics window error.crosses(x,y) #in graphics window #now for a bit more complicated plotting desc <- describe.by(attitude,(attitude[,7]>41)) #select a high and low group g1 <- desc$'FALSE' g2 <- desc$'TRUE' plot(g1$mean,g2$mean,xlab = "Low Advance",ylab="High Advance",xlim=c(30,80),ylim=c(50,80)) error.crosses(g1,g2,labels=rownames(g1),pos=rep(1,7)) title("Attitudes grouped by high and low scores on Advance")