The true distribution was a mixture of two normal distributions, 60:40, with means 7 and 14, and standard deviation both 1.5. (In effect the true distribution was really a histogram of this in which all the values had been rounded down to the nearest whole number, because all values drawn from the distribution were truncated.) The sample of 150 points had histogram: its mean: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ---------------------------------------------------------------- 0 0 0 0 4 18 14 17 22 11 5 6 9 12 16 7 6 2 1 0 0: 9.6067 and the three bootstrap samples drawn from that sample (and their means) were: 0 0 0 0 4 16 18 19 20 14 6 6 9 13 12 6 7 0 0 0 0: 9.3067 0 0 0 0 4 17 18 13 17 5 7 4 16 13 21 8 4 1 2 0 0: 9.9133 0 0 0 0 6 9 4 10 32 17 8 7 10 15 16 10 5 1 0 0 0: 10.1000