You’ve collected your data, coded it, got your license for SPSS, inputted (is inputted a word?) it, and finally run correlations for all variables against all other variables and you’ve got an interesting selection of significant correlations. If they mean anything, you could have an or even a whole thesis! But how do you decide if they mean anything? That is, how do you move from correlation to causation?
It will be helpful to those who are venturing into this area in their research for the first time, and will also be useful if you’ve got undergraduate or taught Masters students who’re unsure about stats and how to interpret data.
To illustrate the sort of issues involved, consider these two correlations taken from a great site described by one of my colleagues as ‘possibly the best site I’ve seen all year..’
The first shows the very high correlation between the consumption of Mozzarella cheese and the award of Civil Engineering PhDs in the US:
The second, a similarly high correlation between the award of Computer Science PhDs in the US and amusement arcade revenues.
Food for thought!