Monday, July 16, 2007

What do Polls Really Tell Us

When using polls as a guideline for action, it is imperative to be aware of some important issues in poll taking.

First – how is the question phrased? As example, consider the two questions below, both of which are attempting to evaluate the popularity of chocolate ice cream.

1. Do you like chocolate ice cream?
2. Do you prefer chocolate or vanilla ice cream?

Lets look at the results of the polls and let us assume they were carried out in an accurate and reproducible statistical manner.

Result of Question 1: 95% yes, 5% no.
Result of Question 2: Chocolate 50%, Vanilla 50%

Now lets change Question 2:

Do you prefer chocolate, vanilla, or some other flavor?

Result: Chocolate 50%, vanilla 20%, other flavor 30%

In this simple example it is easy to see how one may interpret the data differently depending on how a question is posed.

Now two more questions on a topic of national interest in our leadership:

1. Do you think President Bush is doing a good job? Yes or no?
2. Had Al Gore of John Kerry been elected instead of President Bush, do you believe that Gore or Kerry would have performed better than President Bush? Yes or No?

Results to Question 1. Yes 30%, No 70%
Results to Question 2. Yes 50%, No 50%

Interpretation of results to Question 1 appear quite clear-cut, but when the question is asked in another way, the results are open to a different interpretation. Take care before reaching conclusions. Consider in your own mind what may or may not have been actually polled!

Second issue – who is aware of all the facts needed for decision-making?

If a poll were taken among high school students regarding whether it is worthwhile to remain in school until graduation, it may show that there is a significant number of students who would rather leave school than stay. Have the students have answered with an awareness of all the facts regarding the pluses and minuses of such a decision?

A poll asking whether we should raise taxes, or leave Iraq, or invade Darfur, etc, asks for decisions to be made by people who most likely have very limited information. The vast majority of the polled public probably has not fully researched all the relative facts required to make an intelligent decision. Many of the answers are given based on “gut” feelings or because of media presentations. For example:

Presentation by Media 1: “2000 individuals lost their lives today.”
Presentation by Media 2: “2000 individuals were slaughtered today.”

In both instances 2,000 individuals are reported to have died, but how one regards the method of dying is quite different.

Individuals who are empowered to make decisions - parents, teachers, lawmakers, jurors, presidents, should make judgments not just on polls, but on the facts and all the facts, needed to make the proper decision – facts of which frequently only the decision-makers are fully aware.

No comments: