* program to generate 100 samples, n=10, with normal distribution (mu=50, sigma=6)
*    Bert Kritzer, June 27, 2002
*      Set Seed added October 28, 2002
*      updated to Version 9.0, November 1, 2005

version 9.0

* clear everything in memory
drop _all

*   set number of repetitions (100); each row will be one sample)
*   to change number of replications, reset the value of 100 below
*   to desired number

set obs 100

* set random number seed
* students should reset this to a random number chosen from
* table B
set seed 53

* get observations
*   to get different mean and standard deviation, reset 50 and 6 in
*   line below

for new case1-case10: gen X = 50+6*invnorm(uniform())


* get mean for each replications (the rmean function computes the
* mean across the variables given as arguments)
*
*   to use less than 10 observations for each replication, change
*      case1-case10 TO case1-case__ 
*   where __ is the number of observations you want to use

egen mean=rmean(case1-case10)

histogram mean, bin(12) xlabel(42(2)60) normal