These approaches are helpful but, in our opinion, are not as effective as having students participate in sampling experiments. A simple experiment is to sample a rectangular distribution, either from a table of random numbers, by drawing chips from a bowl, or by computer. If a computer is used, it will also be easy to sample other kinds of populations. Sampling a moderately skew population may help convince students of the Central Limit Theorem in the absence of symmetry. Indeed, the use of several populations (e.g., rectangular, exponential) can demonstrate to the student that the rapidity with which the sampling distribution of (x − μ ) ÷ (σ n ) approaches a normal distribution as n increases depends on the population from which the samples are selected.