Reliability Of Political/Opinion Polls

Before I place faith in a political or opinion poll, here are some of the things I should verify:

1. The administrator has used objective standards.
2. A small sample group can be used to make inferences for everyone.
3. The sample is representative of the population it is attempting to estimate.
4. The sample reflects all significant factors that affect responses.
5. Subjective answers from a small group can be extended to the large group.

At a minimum, these five conditions must be met before relying on poll results. I must also ask additional questions.

How accurate would it be to sample half the students in a school and then say this is what everyone wants to eat? Not even with sampling 50% of the population would I be right. Percentage-wise, political and opinion polls take an extremely small sample and then claim to be accurate. Samples might work when estimating fish in a pond, but not when estimating personal choices and preferences which depend on a different set of factors for each person. It is more scientific when the population is likely to behave the same as the sample. This same behavior will only occur when the sample and the population reliably use the same set of criteria to guide behavior.

Undoubtedly, nature is guided by instincts; therefore, all members of a population will exhibit a similar set of behaviors. This makes sampling reliable. Humans use individual choices based on many variables. This makes sampling/polls unreliable when asking humans subjective questions.

Why haven't I been polled on anything in 50 years? Does this mean there are others who have been able to accurately speak for me?

If I am polled, will my answer be an indicator of how hundreds of thousands of people will respond? Even two people who call themselves soulmates respond differently to the same question. Trust a poll on matters of personal preference? Come on now! The game show, Family Feud, is a good example of how often an individual response doesn't match a sample group.

If I had been polled, I'd probably respond off the top of my head instead of first making sure I arrived at my answer independently. I wouldn't be thinking how important my answer might be to others. I would respond quickly which rarely involves careful consideration. I would say what I could quickly recall. I would likely recall what I had heard and read frequently, accurate or not.

If I had been polled, I would probably change my response at least half the time after further consideration.

Political and Opinion polls are also problematic in that they are never an indication of which direction people will head when the time comes or where they will finally take root. To be swayed by a poll is to navigate by dandelion seed.

Truth is, many polls are not taken to help the public but to help spinmeisters gauge the effect of their spinning. They are checking “whether signs,” to see if the hype has us sufficiently off balance. Going along with polls is thinking by substitute thought. It's like using an unreliable replacement head.

Related: Financial Information Sources, Gaming The Economy & Our Education