How statistically meaningful is telephone polling?

The Watchdog received a call from a national firm conducting a poll of Southeast Pennsylvania.

The Watchdog responded that if the poll would not take more than two minutes, he would cooperate.  The party responded that the poll would take ten to thirteen minutes.

 Our question:  How representative of the general population are people willing to stay on the phone with a stranger for thirteen minutes answering questions?   

What happens is that many clients are served by a single phone call.  To assure accuracy, it is better for a firm or organization to spend more for a short poll, perhaps one specific to their purpose.  They are much more like to achieve the random sample essential for accuracy.

Share