It would be helpful to know exactly what statistical definitions you need to include, as that will, in part, determine the kind of project you need to do, but here are a few ideas:
1) If you have terms related to probability, you could use a bag of M&M's to illustrate things like: the probability of getting a (insert color of your choice here) M&M, or of getting 1 blue and 1 red M&M when drawing only 2 M&M's (or if you draw 10 M&M's), etc.
2) if you have terms are related to study design (i.e. bias, power, etc.), then you could run a small study and point out how different aspects effect other pieces of the puzzle.
You get the idea?