the example goes something like this:
Assuming a set of data is collected to show warping of a long metal pipe.
0.756 0.856 0.943 0.978 1.12 1.23 1.45 1.67 1.87
Plot a probability plot to ascertain if the data is of normal distribution.
I know how to use minitab to plot the probability plot and then check if it is
more or less a straight line or is there any out-lier.
But I would like to learn how to manually plot a probability plot.
Normally a graph plot will require 2 set of datas, 1 set for the x-axis and the other set for the Y-axis. But now that I have only 1 set of data, how to I use this particular set of data to generate the other set of data for the other axis?
Is there any standard parameters for the x-axis and y-axis?
The textbook says something about using percentile to calculate but I don't really understand what its trying to say..... :(
I can see you are on the right track. What you are looking for is called qq plot. The way to tell if the data is normally distributed just by plotting their quantiles.
you have observed value as x values. you need rank based z score for the y axis.