Harris Poll Chairman Stands Up to The Economist
// January 19th, 2009 // Comments // Journal

- Image via Wikipedia
Cheers to Humphrey Taylor of Harris Interactive for defending online research in his letter to The Economist:
SIR – Your article about the accuracy of opinion polling in America’s election stated that “online surveys are notoriously biased” (“Poll, baby, poll!”, October 25th). Their track record says otherwise. In the almost 80 elections in the United States and Britain where we can carry out a comparison, the final forecasts of online opinion polls have on average been somewhat more accurate than telephone polls. What’s more, they include most of the “cellphone only” population.
You also wrote that “most experts” consider as “sloppy” the weighting of surveys to compensate for biases in the sample. As the raw samples in all opinion polls contain biases, whether the data are collected in person, by mail, on the phone or online, it would be very sloppy not to weight them. Who are these experts?
Humphrey Taylor
Chairman
The Harris Poll
New York

![Reblog this post [with Zemanta]](http://img.zemanta.com/reblog_c.png?x-id=6ad9718b-3558-4de1-9479-5fa3fe17f6d1)
![Reblog this post [with Zemanta]](http://img.zemanta.com/reblog_c.png?x-id=a6486740-e8c8-4789-bbec-2424622f9321)


