Skip navigation



To be sure that we can trust an answer, we need a minimum number of observations. If we pick one ball out of a box, and it is blue, what does that tell us about all the other balls in the box. Are they all blue, or are some of them red? If the next ball is red, does that mean we have equal numbers of blue and red balls? Clearly we need a bigger sample than this.

Random chance, quality, and size are all related topics in clinical trials and meta-analysis. To look at them together needs a fair amount of space, so Bandolier has a downloadable PDF to examine aspects of all of them. First it looks at the effects of random chance that we can get just by rolling dice. Then it examines how likely our results are to change, depending on the amount of information we have available, and finally it reverts to the answers to the socks drawer, or ibuprofen question. The link is below:

Bandolier Extra on size .