Topic 16: Learning About a Proportion

Minitab Local Macro p_disc

Computes the posterior distribution for a proportion using a discrete set of models

TO RUN:

Name two Minitab columns 'p' and 'prior'. Place the values of p in 'p' and the prior probabilities in 'prior'. Here we take the proportion to be nine equally-spaced values between .1 and .9 and give each proportion a probability of 1/9 = .111.

MTB > prin p prior

Row p prior

1 0.1 0.111111

2 0.2 0.111111

3 0.3 0.111111

4 0.4 0.111111

5 0.5 0.111111

6 0.6 0.111111

7 0.7 0.111111

8 0.8 0.111111

9 0.9 0.111111

Suppose that 5 successes and 15 failures are observed. To compute the posterior probabilities, type the following in the Minitab Session Window. The 'plot' command gives a display of the prior, likelihood, and posterior probabilities.

MTB > %p_disc 'p' 'prior' 5 15;

SUBC> plot.

Other subcommands:

The full macro command is

%p_disc P1 PRI s f;

store POS;

plot.

The 'store' command will store the posterior probabilties in the column 'pos'.