As mall interviewing has grown in importance as a method of data collection, questions have been raised regarding the way respondents are recruited. In particular there is concern that certain types of people haunt the mall concourses, some of whom are just waiting to be "corralled" by a research interviewer.

The scenario goes something like this: Mary goes to the mall at least once or twice a week. Like many others, she isn't there because she has something to buy, she just likes to browse in the bookstore, look in the pet shop, see what's new in her favorite department store and just "shop." When she sees the mall interviewers she always stops to say hello and usually asks what they are interviewing on today.

Mary likes to be interviewed and she is very cooperative. Sometimes she tastes products, sees some commercials before they're on TV, or maybe gets a product to take home. Occasionally she gets paid for being interviewed.

Mary is the answer to an interviewer's prayer. She knows to always say she hasn't been interviewed in the past six months (even though it was only last week). With a little leading, she can guess what the screening interviewers want her to say. And Mary can even remember what brand of dog food she said she used in a survey last month, and go through a standard validation like a stealth bomber.

There are people like Mary at every mall who truly want to be helpful. But when repeat respondents make up even a minor proportion of your quota, their "help" can seriously damage the results of a study. The difficulty is that these days the marketing research industry is increasingly troubled by declining cooperation rates. So why should it be concerned about those who enjoy participating?

Aside from the classic problems of "professional" respondents (who are far from the "naive" respondents on whom survey research depends) there is a much more insidious aspect of their use. Cutting a corner on the past participation question can soon lead to more serious violations of the qualifications.

Cooperative, repeat respondents can develop a conspiratorial familiarity with an interviewing staff. Having such cooperative respondents "fudging" the frequency of product use, the brand used (or even use of the product category altogether) and other key respondent qualifications dilutes and can even destroy the validity of the results.

There are no figures to indicate if the practice of using repeat respondents occurs often in mall research. But certainly the temptation to do so exists. In this business, as in many others, any practice that can result in faster turnaround and/or lower costs-and which might not be detected under normal conditions-should not be assumed to be rare. Even though the industry continues to emphasize quality in data collection, competitive pressures are at an all time high. Mall location managers, often miles away from their management offices, face the decision alone. They can either actively encourage the use of repeat respondents, they can ignore the practice or they can actively discourage it.

One company, Ruth Nelson Research Services, has put into action a program to safeguard their client's projects against date that results from using overly cooperative respondents. They have a custom-designed program that matches the telephone numbers of all mall respondents in each city over a three month program. Each day they input the names, telephone numbers, date of interview, interviewer, project number and other details for each person interviewed at the malls. By running the matching program regularly, repeat respondents, and their interviewers are noted; when a repeat occurs an investigation is begun.

Some matches are found to be innocent. A husband and wife interviewed on separate days or a second project with no past participation requirement are spotted. But whenever a violation is noted the company notifies the client, the interviewer, and the supervisor in charge of the project.

The main benefit of this system hasn't necessarily been in catching the occasional errors that might be made under the best of circumstances. It has been in preventing widespread use of repeat respondents. The match program insures that everyone-clients, interviewers, and supervisors-knows that this research field service not only discourages the practice but that they can also detect it immediately.