A recent story in State Impact focused on the results of one poorly phrased and misleading question out of several in a poll that ended up showing broad public support for natural gas development. That one question suggested citizens of both Michigan and Pennsylvania desired moratoriums on hydraulic fracturing but other survey results contrasted sharply with this conclusion.

When polling results don’t fit the template, some media sources try bending the polling curve to suit their storyline. When the poll itself is also designed to deliver a particular result, we get headlines like this one from State Impact :

“Poll Shows Support for a Drilling Moratorium in Pennsylvania”

The poll in question was one conducted by the National Surveys on Energy and Environment (NSEE), and was entitled Public Opinion on Fracking: Perspectives from Michigan and Pennsylvania.

State Impact’s headline, unfortunately, reflects the answer to a single (poorly phrased) question among a total of more than two dozen – and an equally poor understanding of how phrasing can alter the results. (More on that later.)

Nonetheless, given the actual results throughout the poll, a far more accurate headline – and one that actually describes the overarching results – would have been:

“Poll Shows Residents Strongly Support Shale Development”

http://eidmarcellus.org/marcellus-shale/bending-the-natural-gas-pol...