To ensure that the results are accurate? Especially if you are measuring pH, volume of gas given off, etc.
What was the context of the question?
Oh, right. Erm, presumably it's something to do with being sure of the full extent of the temperature change? It completely depends on so many other problems though, is it for an experiment or just a demo? If it's for an experiment you must always wait for every reaction to complete, otherwise the results will not be comparable as a variable not being compared isn't being properly controlled.
To make sure that the results completely are finished.
This is because sometimes results do vary within time and in this case it can be within seconds. So if your test was 11 seconds the 9tth can have a different result to what it would have in 11
Not a stupid question , but it was stupid to say that it was stupid lol -)