10.07.2015 Views

Using R for Introductory Statistics : John Verzani

Using R for Introductory Statistics : John Verzani

Using R for Introductory Statistics : John Verzani

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Bivariate data 77> stripchart(list(ephedra=ep,placebo=pl), # named list+ method = "stack", # stack multiples+ pch=16,offset = 1/2, cex=3) # big circles—notsquaresFigure 3.3 shows the graphic (slightly modified).Figure 3.3 Strip chart of placebo andephedra group3.2.4 Quantile-quantile plotsThe boxplot uses the quartiles (essentially) of a data set to graphically represent a data setsuccinctly. If we use more of the quantiles, a very clear picture of the data can be had atthe expense of a more complicated graph to read. A quantile-quantile plot (q-q plot)plots the quantiles of one distribution against the quantiles of another as points. If thedistributions have similar shapes, the points will fall roughly along a straight line. If theyare different, the points will not lie near a line, in a manner that can indicate why not.A normal quantile plot plots the quantiles of a data set against the quantiles of abenchmark distribution (the normal distribution introduced in Chapter 5). Again, thebasic idea is that if the data set is similar to the benchmark one, then the graph willessentially be a straight line. If not, then the line will be “curved” in a manner that can beinterpreted from the graph.Figure 3.4 shows the q-q plot <strong>for</strong> two theoretical distributions that are clearly not thesame shape. Each shaded region is 5% of the total area. The difference in the shapesproduces differences in the quantiles that curve the q-q plot.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!