Effect of a packer on speed of checkout operators

This diagram shows that random allocation of treatments to experimental units results in an unbiased estimate of the treatment effect — there is no tendency to over- or under-estimate.

Again click Allocate treatments then Conduct experiment to see typical experimental results. Observe that there is now no tendency for younger or older checkout operators to get packers.

The estimated effect of getting packers is quite variable but there is no tendency to over- or under-estimate it.