And the survey says...

There are such a myriad of purveyors of surveys and opinion polls (Gallop, Harris, Nielsen, Roper, Pew Research) that it’s difficult to know if one provides a better pulse of consumer, constituent or citizen attitudes than others. Another source of gathering people’s viewpoints are conducted by colleges and universities, of which the University of Michigan Institute for Social Research has a 60-year history of being “one of the largest and oldest academic survey and social research organizations in the world”.

Then there are polls that are less than scientific on social websites, including MySpace, Facebook and Sodahead. Some of these surveys are well intentioned and thought provoking while others are nothing other than childish inquiries on such topics as the Jonas Brothers, Britney Spears and other teenybopper interests.

Others website networks give people and businesses an opportunity to design and post online surveys to suit their specific needs, some of which are free, such as polldaddy, micropoll, twiigs, proprofs, and surveypopups.

There are industry-specific questionnaires – medical, education, employee satisfaction, customer satisfaction, student opinions, plus surveys that measure how the electorate views the performance of international, national, state and county governments.

The wording and placement of survey questions can produce different results. Survey results can be slighted with the intent to direct questions that may result in respondents to give answers that serve their cause. To avoid this tactic, it’s necessary to consider two important factors in validating the worthiness of a survey – demographics and methodology.

Demographics include household income, gender, education, race, age and marital status, all of which are basic considerations. Understandably, political opinion polls require party affiliation. The more information given by a targeted segment of the population, the results hold a greater level of accuracy than if the cross-section of respondents has its limitations.

Methodology identifies how the poll was administered. For instance, the accuracy of a random sampling of 1,000 respondents is typically +/- 3% while a larger sampling of 10,000 reduces the margin to a 1% error rate, which also takes into consideration that some participants may not be entirely honest in that they try to second guess mainstream opinions.

To make the sampling as random as possible, surveys conducted by large pollsters may take a computer-generated list of, say 50,000, active home telephone nationwide numbers. From there, a computer may then be programmed to randomly select perhaps 1,000 residential numbers. A random selection of exchanges and digits takes the process one step further.

Another measure to be taken into consideration: generally speaking there’s a 95% accuracy rate of the +/- accuracy rate! Figure that one out!

‘No opinion/Prefer not to answer’ on a multiple-choice question restricts information for analysis of information, yet they’re appropriate for answers concerning household income, race and religious affiliation. It’s like saying, “It’s none of your business.” But without providing age and sex, results of many surveys invalidate the intended focus of the subject at hand.

Strongly agree, somewhat agree, agree, slightly disagree, disagree, strongly disagree. Take your pick. This format provides a more specific opinion than all other question-answer polls. It may also be more affective in giving respondents an easier task of making a choice if they should have a borderline opinion on a given subject. It helps to avoid the guesswork out of the actual mood of the respondent. Depending on the question, I appreciate the option to ‘strongly’ disagree rather than be restricted to ‘disagree’.

Some opinion polls are sought on a regular basis, say weekly and called a “tracking poll”, when there is deemed a frequent change of attitude, such as health care legislation, the economy, foreign policy, abortion or, coming soon to everyone’s contention, the debate on immigration reform, which is sure to be another round of partisan alignment.

A bandwagon effect occurs when the poll prompts voters to back the candidate shown to be winning in the poll, which may have been the case during the 2008 Presidential primaries, which not to meant to put in doubt the popularity of Obama.

In the 2008 General Election, Barack Obama garnered 52.9% of the vote, which happens to be extremely close to the 52% overall national approval ratings in July and August, so it should be of little surprise that, after the 100-day honeymoon, it should level off from the 60% in February when he was given the benefit of the doubt for his anticipated efforts to tackle the leftover debris from the Bush years, and the 61% approval rate in May, which may have been largely influenced by taxpayer rebates.

My favorite surveys are those sought from businesses that are requested for marketing purposes. Wal-Mart’s an excellent example where I could actually win $1,000 of in-store purchases. The odds aren’t likely to be any better than those for the Lottery.

For relaxation I watch Family Feud. Whether be the original host Richard Dawson or followers Ray Combs, Richard Carr or today’s John O’Hurley some are repetitive whereby I tend to coach them on the answers. It’s obvious some have watched previous broadcasts as they sweep the answers toward victory.

Although unscientific, the results on questions can be very telling of people’s attitudes. There a few ‘Fast money round’ responses asked of “one hundred people” that were very surprising.

The #1 answer as to what age people start to lose their memory: 50 years old! I beg to question the validity on that one. The #1 answer to “On a scale of 1 to 10, how satisfied are you with your job?” was “5” – not surprising by any means.

The one that got shocked me a bit was “On a scale of 1-10, how prepared is the USA for a crisis?” I would have given it a “7” but the #1 answer was “5”! from a hundred people more astute to reality than myself.