I try to stay in touch with the graduate students I meet throughout the years at the various colleges that I visited.  The most prevalent response I get when I ask how things are going in their jobs, centers around how market research is abused.  It seems that research is conducted without regard for good fundamental research principles.  Actually this comes as no surprise.  I have frequently seen the same abuses.  I have found that it does little good to hit the problem head on.  As a result, I have been recommending that every market research department have a good supply of textbooks.  My personal favorite is, which I usually recommend, Market Research -- Principles & Applications by Melvin Crask, Richard Fox and Roy Stout.  The authors are not only writers and educators but I know personally that two were heavily involved in corporate market research business, namely with Coca-Cola and Procter & Gamble.  Their first publication, 1995, is my standby.  This book does what many textbooks do not do and that is point out the weaknesses of various designs and also how the designs are frequently abused.

It is my experience that one of the most frequently abused test designs is the Protomonadic Test design.  [Note:  The Protomonadic design includes both single product and paired comparison evaluation.  Participants use one product for a period of time, after which a single-product interview is administered.  They then use a second product for an equal period of time, after which a paired comparison interview is administered.]  The authors point out that "No questions pertaining to specific product attributes can be included in the single-product interview to avoid biasing the participant's evaluations of the second product."  In other words the single product interview, conducted after using the first product, is limited to likes and dislikes along with an overall assessment.  Ignoring this concern is one of the most common abuses I have seen in this particular design.  Unfortunately there are those that believe they can obtain additional data at no extra cost.  They ask all kinds of questions about the first product assuming that bias does not enter into the assessment of the second product.

The Sequential Monadic test design is a version of the Protomonadic Design, where a product is placed, a single product interview is conducted (again only likes & dislikes along with an overall assessment), and then a second product is placed and a second single product interview is conducted.  The authors point out that the second product is not truly evaluated in a monadic fashion, but rather is evaluated relative to the first product.  This can make the second single product interview data meaningless.

Note that the above examples deal with the psychological effects in interviewing and not physical product effects.  There are even greater hazards when physical performance characteristics are involved.

Every market research organization should have a good supply of market research textbooks.  Also for anyone interested, the University of Georgia, through its Continuing Education Center, administers an excellent study at home market research program for the MRA/ARF.  This is an excellent way for those in the work force to receive formal marketing research training.