Cape Gazette
http://capegazette.villagesoup.com/p/977801

Look at real gun control statistics

By Theresa Garcia | Mar 23, 2013

On March 19, the Delaware Coalition Against Gun Violence proudly released a poll, completed by a Washington polling company, that stated, "66 percent in New Cstle County, 56 percent in Kent County and 54 percent in Sussex County supported stricter laws for gun control." That sounds pretty overwhelming until you consider the following.

There are 917,092 people in Delaware. This company "randomly" selected 600 registered voters in Delaware to conduct their poll. That means .065 percent of Delawareans were contacted. That is less than 1 percent. I wonder what the other 99.94 percent think?

At Gov. Markell's gun control legislation debate March 19, I took a poll of around 200 Delawareans, mainly from New Castle Coumty. It revealed an overwhelming 99 percent of the attendees were against all five of Markell's bills! Keep in mind that this meeting was open to the public and heavily advertised by the Delaware Coalition Against Gun Violence, which supports all five of Markell's bills.

We feel that this is a more accurate poll as those who were in attendance are the people who are engaged, and who have taken the time to research the proposed laws.

Theresa Garcia
Magnolia

Comments (1)
Posted by: Thomas Adams | Mar 24, 2013 17:26

The number of people who live in Delaware is not a relevant consideration in this instance.  Polling firms looking at questions concerning public policy will sample from the pool of registered voters and will purposely eliminate non-voters.  The polling firm in question did just that.  Their sample of 600 out of 627,784 registered voters in Delaware represents a tad under 1% of their targeted population.   Ms Garcia seems to think that there can be no or little predictive value in such a small sample size.  There is no basis in the field of statistics for reaching such a conclusion.  Reputable national polling firms typically use a sample size of 1,000 to make predictions about what 146 million registered voters think!

 

Lastly, Ms Garcia writes as though the word “randomly” means “haphazardly.”  It does not.  In statistics, a random sample means that every individual has the same chance of being included in the sample.  Random sampling is desirable.

 

There may well be flaws in the methodology of the poll, but Ms Garcia fails to identify any of them.



If you wish to comment, please login.