Rant coming up! But before we get started, take a moment to reflect on what happiness means to you.
ok. so now: This paper purports to analyse the relationship between happiness and mortality. But even before reading it, we know it can't possibly do that. It must be a study of the relationship between mortality and response to some question like 'Are you happy today?'. Which is a different matter altogether. (In technical language I'm querying the validity of their measure of a construct which, to my mind, is pretty complicated.)
Here is the key question:
“How often do you feel happy?” Possible responses were “most of the time”, “usually”, “sometimes”, or “rarely/never”.
This was analysed by grouping 'most of the time' and 'usually' as 'happy', and the others as 'unhappy'.
The study may have had a Million Women, but that doesn't mean they came up with the right answer.