The Kos/R2K Scandal
I won’t belabor the details. Here’s a list of links instead:
- Three individuals with strong credentials for doing so publish a damning statistical analysis of R2K’s polling after failing to get much of a response by R2K.
- Markos Moulitas of DailyKos was shocked and appalled by the apparent fraud.
- A new blog by one of the writers of the report sheds more light on the analysis.
- Details emerged regarding the lawsuit DailyKos is planning to file against Research 2000.
- 538 and TPM break the news to their readership.
- Nate Silver at 538, whose rating of R2K as subpar initiated a lot of this, reveals details of his previous qualms about R2K, and then gets accused of ruining R2K’s reputation in a shoddily done “cease & desist” demand by their lawyer.
- Mark Blumenthal, the man behind Pollster.com and a frequent critic of Nate’s, reports on the scandal and has some sound advice for pollsters for the future of their profession.
In short, another pollster bites the dust. It’s a shame, really. Because it means fewer people will trust polls – and few trust them now anyway because these statistical tools for measuring public opinion are misrepresented and misinterpreted all the time. (Never mind that some polls ask about things that questionable value.)
The truth for pollsters is that Blumenthal is right – better disclosure, a more honest relationship with their clients, and adhering to more professional standards will help their industry. This goes for both public and private polling, for political and non-political work. Polling is useful, but we need a better job at educating non-pollster/non-statistically-inclined folks at the limitations inherent in survey research and the steps taken to minimize them.