You would avoid that orange, but you wouldn't say to someone "don't eat oranges, I had one once and it was bad", which is what you are doing.

You can't always do a large study, but when you're presented with one it doesn't help do give anecdotes in response.