Treatment of Non response in MR

Proper field methods can reduce the non-response rate – both refusals and not-at-home – but they cannot eliminate the problem completely. Since most researchers have agreed that the bias of non-response is often considerable, definite provisions to reduce it are almost always warranted.

In door-to-door interviews with probability samples, neighboring dwelling units are often substituted for not-at-homes or refusals. It is clear that this procedure does not solve the problem if the universe is thought of as consisting of two strata – one from which information will be obtained if its members are included in the sample and one from which no response will be obtained. Substituting new households just adds to the sample from which information is obtained; it does nothing to obtain information from the other stratum.

Call backs are the only useful method of obtaining information from the non response stratum. This applies to both telephone and door-to-door personal interview studies. For most surveys it is wise to select a smaller sample and make four or more callbacks than to spend the same budget on a larger sample with no call backs.

In door-to-door surveys the cost of call backs can be reduced by leaving the questionnaire at the sample home with a request that the resident complete it and return it in the envelope provided. This has some of the problems of mail surveys, but it does reduce the size of the non response group. Costs may also be reduced by using the telephone to schedule call backs. The telephone is particularly useful in locating, screening, and interviewing relatively unique populations, such as individuals who have exercycles in their homes.

The problem of non-response has been growing and gives every evidence of continuing to grow in the future. Not-at-homes can be reduced by call backs as indicated above, but refusals are another problem. Care in the construction of the questionnaire particularly in the opening part, is helpful in reducing refusals; attention and interest must be achieved very early in the interview. Experiment in the pre-testing phase can be helpful in improving the introduction.

Improved interviewing especially in telephone interviews can reduce refusals; training of interviewers should emphasize how to maintain respondent rapport.

As most refusals occur in the first few minutes of the interview, the introduction is a key factor, especially in telephone interviews. In general the introduction must be as brief as possible to permit the respondent to become involved in the subject of the interview as soon as possible; the longer the explanation the more likely the respondent is to hang up. Four items should be covered in the introduction:

1. Verification of the telephone number dialed (or of other respondent qualification if personal interview).
2. Identification of the interviewer and the interviewer’s organization.
3. Brief explanation of the purpose of the survey and the universe being studied.
4. Some positively worded phrase to encourage cooperation.

Continued growth in the non response problem is causing research organizations to concentrate more resources on efforts to overcome the problem. Whether or not new solutions will be found is not yet certain.

One research firm that has given a good deal of attention to the refusal problem has come up with the following list of 11 tips to interviewers to help them achieve better cooperation in surveys:

(1) Be honest about the amount if time needed for the interview. This may encourage people to be more willing respondents in the future.
(2) Educate the public by encouraging your clients to say publicly that survey information caused them to change or create products or services.
(3) Remember that this may be your 10th or 40th interview, but it’s the respondent’s first. Make every contact you make your first.
(4) Try to avoid calling during dinner time and don’t call after 9 pm.
(5) Avoid asking personal questions early in the interview.
(6) Wording should be carefully chosen to not insult the respondent’s intelligence.
(7) Involve people in the screening process by asking questions that require some thought, not just a yes / no response.
(8) When approaching people in a shopping center, use an energetic purposeful movement to establish eye contact. Look at that person and have a friendly but not stilted facial expression. Speak clearly in a conversational tone and talk to the people, don’t read to them.
(9) Avoid calling people during the upper Bowl, Mother’s Day or a world crisis i.e. Challenger astronaut tragedy, Libyan invasion.
(10) Mae sure your introduction does not sound like the beginning of a sales pitch.
(11) Say please and thank you. Make each respondent feel he or she is the most important person in your life during the interview situation because they are.