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

Single Variable Research

During the holidays I received a new textbook from a good friend, Dr. Richard Fox, University of Georgia "Masters of Marketing" program. It is easy reading, and it is loaded with case studies that bring back many memories of past successes, failures, fun times and frustrations.

One particular topic I'd like to discuss today was brought to mind on page 80 where Rich was discussing "interaction and how to measure it." My frustration for many years at P&G was the restriction imposed on multiple variable testing. While Procter has changed, I find companies that still restrict their testing to single variable studies.

It has been my belief that single variable testing is expensive, time consuming and can be misleading. For example, consider a project I worked on involving a product today called Bounce. In the early development, we were looking at evaluating the following aesthetic variables: two shapes, four colors, two designs, and four odors. Following the single variable concept, we would have selected the preferred shape, then with the preferred shape, determined the preferred color, etc. After four, single variable studies, we would have the aesthetics selected for the new product.

Luckily, this was not what we did. We designed a single, factorial study aimed at measuring all combinations. In this way we assessed the main variables and their interactions. (For some of my readers it may be more helpful to look at interactions as synergisms.) The combination of variables selected in this approach may have been the same ones that would have been selected in the single variable approach if there were no pronounced interactions. But with the factorial approach we reduced the time and expense involved and more importantly we had assessments for all 64 combinations of the variables. This was especially important since the combination selected (round, yellow, sunburst, and a perfume called outdoor fresh) was later found to be impractical. At this point we were able to go back to the aesthetic study and select the most promising combination that did not have the negative in-use properties.

Single variable studies are much easier to analyze than a factorial, but if there is a strong synergistic effect, you can miss the better opportunity.

For those interested, the book is titled, Marketing Research Principles & Applications, by Crask, Fox & Stout, a Prentice Hall publication.


[Back][Index][Forward]