MTB > # predictive inference using a beta prior
MTB > #
MTB > # let p=prob(coin lands head)
MTB > # suppose that your prior opinion about p is described
MTB > # by a beta(10,10) distribution
MTB > #
MTB > # want distribution for number of heads in 10 tosses
MTB > 
MTB > exec 'p_beta_p'
Executing from file: p_beta_p.MTB
 
INPUT VALUES OF BETA PARAMETERS A AND B:
DATA> 10 10
 
INPUT NUMBER OF TRIALS:
DATA> 10
 
INPUT RANGE (LOW AND HIGH VALUES) FOR NUMBER OF SUCCESSES:
DATA> 0 10
 
PREDICTIVE DISTRIBUTION OF NUMBER OF SUCCESSES:

 Row   SUCC       PRED

   1      0   0.004612
   2      1   0.024274
   3      2   0.066752
   4      3   0.125651
   5      4   0.178660
   6      5   0.200100
   7      6   0.178660
   8      7   0.125651
   9      8   0.066752
  10      9   0.024274
  11     10   0.004612