In part (iv) of the exam type question we are asked for the confidence interval for the "mean" score corresponding to an "individual" response of 53. In the answer they calculate the variance using the formula [1/n+...] as opposed to [1+1/n+...]. Obviously a trick question (well tricked me anyway). How do I know when to use the mean vs the individual variance formula? Thanks.
I think we've been a bit cruel - and it isn't representative of exam questions. Frankly your default will be the mean response (as this is what we use in practice) and only use the individual response if it explicitly says "individual" (which has been twice counting from 1993).
Fair enough. I mean the question did say CI for the "mean" score - so a bit more careful reading of the question probably wouldn't go astray either. Thanks.