This is a guest post by Markus Haverland, Professor at Erasmus University Rotterdam and author of a recent book on research methods.
Causal knowledge about the world proceeds by testing hypotheses. The context of discovery precedes the context of justification. We all know that journalists and pundits often do it the other way around: providing for an explanation after the fact.
A particularly hilarious example can be found in today’s issue of “Spits”, a Dutch daily newspaper. Anticipating that the result of the elections for the president of the US would arrive after the newspapers went to press, the newspaper prepared for both situations. It has turned the backpage into a second frontpage. Depending on the results the reader is advised to either read the frontpage or the backpage. On both pages the well-known Dutch journalists, a former correspondent in Washington, Charles Groenhuijsen analyses the results. On the “Obama wins” page he explains that it was evident that Obama would win, because he is a better campaigner and Romney’s economic program is inconsistent. On the “Romney wins” page he explains this outcome, by stating that, ultimately, the US is a conservative country, that voters were afraid of a turn to the left, laws against gun possession, and tolerance towards gay marriage, and that voters thought he was not effectively dealing with the economic crisis.
Well, Spits could have saved themselves the double work and the embarrassment if they just listened to Nate Silver and the rest of the quants. Instead, in a typically journalistic fashion they chose to follow the pundits who can of course ‘explain’ convincingly every possible result post factum.