Views from the Hills by R. E. Stevens, GENESIS II (The Second Beginning) E-Mail views@aol.com

A Reminder about Dealing with Test Biases -- From a News Magazine

Since retiring from P&G almost five years ago, I've been encouraging my friends to be concerned about the effects of the test environment on their test results. I recently came across a new article on research directed at the problems of WHO, WHEN, WHERE and HOW questions are being asked. While the article focuses on polling, the points are relevant to market research.

In the December 4, 1995 issue of U.S. News & World Report, titled "Consulting the Oracle," (page 52), they provide data from reputable research firms showing extreme differences as a result of various conditions.

Data in the article show the order effect in a question resulting in a win of 65/35 or a loss of 23/77, depending upon the order.

The article cites an experiment to determine how many people will give an opinion when they are unfamiliar with the specific topic. A third of those interviewed gave attitudinal responses about a non-existent government act.

How many of us have assessed the environmental effects when designing the research? Your research partners should bring this perspective to you. Their abilities and resources provide value beyond the execution of the research.

The effects of the environment on decision making was recalled yesterday when listening to a sermon. The minister was talking about risk, confidence and faith. He asked the congregation to consider how we would respond to the willingness to walk a plank that was three inches thick, 18 inches wide and 20 feet long. First, if it was lying on the floor and secondly, if it was suspended ten floors above the ground between two buildings. Few, if any, would have difficulty walking the plank on the floor. But not many would be willing to walk it when it was suspended 10 floors above the ground. It's the same plank, only the location was changed, which had nothing to do with the physical ability to walk it, only the psychological.

I believe the above types of experiments should be routinely conducted by market research companies. Recently I have been involved with some "Researching Research" experiments with Elrick & Lavidge, utilizing their SCOPE panels (a real cost-effective way to do research).

From our family to your family -- Happy Holidays and a Successful New Year!


[Back][Index][Forward]