Touch-Screen Surveys

A recent article in Marketing News, September 23, 1996 edition, titled "Debunking the Myths of Touch-Screen Surveys," by Joe Duket, president of Q&A Inc. provided interesting reading.

Mr. Duket made some very interesting points. I agreed with some but felt he did not address the single most important aspect of self selection in surveys of this type. The three myths he addressed were self selection, demographic and socioeconomic representivity, and confidentiality.

Self Selection: His point is that the appearance of bias is a result of the early use of the touch-screen as a vehicle to disseminate information, where playing was a frequent activity on the equipment. He also points out that all methodologies have a non-response bias and his "smart" software can identify children and separate their responses.

I agree that true, random samples exist only in theory. However, it is our responsibility to minimize these biases and take them into consideration in the analysis of the results. What he does not address, and I consider it the most overlooked bias in research is illiteracy. In my research on self-administered, computer-assisted research, I found a great deal of reluctance to participate. At the time, in the early 1980's, I attributed this to the fear of using computers. It was not until a later awareness of the problem of illiteracy, did I realize that what I thought was computer fear was actually a combination of two effects, computer fear and illiteracy. In recent years, computer fear has been reduced but illiteracy has not. Note, illiteracy has been reported to range between 22% and 32% of the adult population. I have been using the most frequently cited rate of 27%. No matter which number you use in the range, it is a significant portion of the population.

Demographic and Socioeconomic Representivity: His results show the representivity of all segments to be present. I believe this is the old dilemma of "Reliability vs. Validity" which will be a topic of its own in a future Views.

A point he makes in this area is "For reaching certain demographic groups, it is far more practical and cost effective to survey where they congregate, than to try hit-and-miss telephone research." Here he is comparing site research, In-Store vs. telephone. I am in complete agreement with this position.

Separately in the article, Mr. Duket claims that touch-screen surveys have a response rate of 85%. While I have no data to refute this claim, I have serious doubts. My data of protocol effectiveness shows the highest cooperative rates to be among door-to-door research at 65%.

Confidentiality: Here his position appears to be based on research comparing comment card data with touch-screen data and maybe a touch of "Construct validity." I believe in this context, his position has merit.

Bottom Line: From my point of view, I would not discourage the use of touch-screen research for some uses but I would encourage experimentation and caution. If anyone has any experience in the use of touch-screen surveys, I'd like to hear about it.