You are correct in stating that you do not need to poll a majority of people in order to make a statistically sound assertion, however you do need to poll a reasonable sample size and use questions that are neutrally biased to accurately make a statistically sound assertion. Can polling a sample of two people allow one to make a statistically sound assertion for a population of forty thousand? In my above post I left out two zeros between the decimal and the five just to prove a point that many people will accept or reject statistics and numbers without actually checking them out personally. That sample group actually represents less than 0.00004225% of the population that makes up the NRA membership. This sample size if far from reasonable and is another trick used by firms to generate biased poll results. Selectively timing the poll so that it follows a major attack where the media is claiming the shooter got his guns without a background check will bias the poll towards background checks. Choosing a sample population mostly from Los Angeles, NYC, and Chicago will further bias the poll. Asking "Do you support background checks on all firearms sales to prevent terrorists from acquiring machine guns" will bias the poll results more than "Do you support background checks on all firearms transfers including a father gifting a family heirloom to his son?"
The Sample size is too small relative to the population it is intended to represent in order to reflect an sound assertion. Without knowing where and when the members of the sample group where polled and the exact questions used we can not accurately assume anything about the poll.
There is a flaw in the above statement. The NRA keeps it's member lists secret therefore one can not reliably poll the majority of the NRA members other than the NRA. The polls used to support this claim are usually biased towards getting results that support gun control measures and interestingly enough recent political polls have proven that polling on a political issue can generate false results (the recent presidential election being one example). The poll cited by this article used a sample of one hundred sixty nine NRA members out of over four million members the NRA had in 2013. Consider the sample of 169 from 4,000,000 represents less than.005% of the NRA membership and the four million number is under-representing the NRA's membership claim.
Slashdot Mirror
You are correct in stating that you do not need to poll a majority of people in order to make a statistically sound assertion, however you do need to poll a reasonable sample size and use questions that are neutrally biased to accurately make a statistically sound assertion. Can polling a sample of two people allow one to make a statistically sound assertion for a population of forty thousand? In my above post I left out two zeros between the decimal and the five just to prove a point that many people will accept or reject statistics and numbers without actually checking them out personally. That sample group actually represents less than 0.00004225% of the population that makes up the NRA membership. This sample size if far from reasonable and is another trick used by firms to generate biased poll results. Selectively timing the poll so that it follows a major attack where the media is claiming the shooter got his guns without a background check will bias the poll towards background checks. Choosing a sample population mostly from Los Angeles, NYC, and Chicago will further bias the poll. Asking "Do you support background checks on all firearms sales to prevent terrorists from acquiring machine guns" will bias the poll results more than "Do you support background checks on all firearms transfers including a father gifting a family heirloom to his son?" The Sample size is too small relative to the population it is intended to represent in order to reflect an sound assertion. Without knowing where and when the members of the sample group where polled and the exact questions used we can not accurately assume anything about the poll.
There is a flaw in the above statement. The NRA keeps it's member lists secret therefore one can not reliably poll the majority of the NRA members other than the NRA. The polls used to support this claim are usually biased towards getting results that support gun control measures and interestingly enough recent political polls have proven that polling on a political issue can generate false results (the recent presidential election being one example). The poll cited by this article used a sample of one hundred sixty nine NRA members out of over four million members the NRA had in 2013. Consider the sample of 169 from 4,000,000 represents less than .005% of the NRA membership and the four million number is under-representing the NRA's membership claim.