If you have the Statistics toolbox
doc normrnd
will give you normally distributed numbers from a population with a given mean, but a sample of 10 from this will give you a mean that is often a notable distance away from 0.3. The only way you could guarantee exactly 0.3 is to fix the numbers instead of taking random numbers. You could do a loop to get as close as possible to 0.3 if that is acceptable. Though as an aside I think I am correct in saying 0.3 cannot be exactly represented by a double anyway.
