A number of factors may affect the choice of a survey method in a given situation. The researcher should choose the survey method that will provide data of the desired types, quality, and quantity at the lowest cost. The major considerations in the selection of a survey method are summarized in Exhibit 6.6.
Sampling Precision
The required level of sampling precision is an important factor in determining which survey method is appropriate in a given situation. Some projects, by their very nature, require a high level of sampling accuracy, whereas this may not be a critical consideration in other projects. If sampling accuracy were the only criterion, the appropriate data-collection
Sampling Precision
The required level of sampling precision is an important factor in determining which survey method is appropriate in a given situation. Some projects, by their very nature, require a high level of sampling accuracy, whereas this may not be a critical consideration in other projects. If sampling accuracy were the only criterion, the appropriate data-collection
technique would probably be call center telephone interviewing, an online survey of a sample drawn from a huge Internet panel, or some other form of polling of a sample drawn from customer lists. The appropriate survey method for a project not requiring a high level of sampling accuracy might be the mail approach or some type of mall survey. The trade-off between the call center telephone survey, Internet panel, and the mail survey methods in regard to sampling precision is one of accuracy versus cost. A call center telephone survey employing a random-digit dialing sampling procedure that includes cell phone and smart phones might produce a better sample than the mail survey method. However, the mail survey will most likely cost less. Often, Internet samples will provide both lower cost and greater accuracy.
Budget
The commercial marketing researcher frequently encounters situations in which the budget available for a study has a strong influence on the survey method used. For example, assume that for a particular study the budgetary constraint for interviewing is $10,000 and the sample size required for the necessary accuracy is 1,000. If the cost of administering the questionnaire using the mall-intercept method is $34.50 per interview and the cost of administering it via Internet survey is $1.50 per interview, the choice is fairly clear assuming that nothing about the survey absolutely requires face-to-face contact.
Requirements for Respondent Reactions
In some studies, the marketing researcher needs to get respondent reactions to various marketing stimuli perhaps product prototype usage (a new style of smartphone keyboard) or a taste test. In these cases, the need to get respondent reactions to stimuli normally requires personal contact between interviewer and respondent. Taste tests typically require food preparation. This preparation must be done under controlled conditions so that the researcher can be certain that each person interviewed is responding to the same stimulus. The only viable survey alternative for tests of this type is the mall-intercept approach or some variant. One variant, for example, is recruiting people to come to properly equipped central locations, such as community centers, to sample products and be interviewed. Some surveys require face-to-face interaction because of the need to use special measurement techniques or eye tracking devices, or to obtain specialized forms of information. The tasks are so complex that the interviewer must be available to explain the tasks and ascertain whether the respondents understand what is required of them.
Quality of Data
The quality of data required is an important determinant of the survey method. Data quality is measured in terms of validity and reliability. Validity refers to the degree to which a measure reflects the characteristic of interest. In other words, a valid measure provides an accurate reading of whatever the researcher is trying to measure. Reliability refers to the consistency with which a measure produces the same results with the same or comparable populations.
Many factors beyond the interviewing method affect data quality. Sampling methods, questionnaire design, specific scaling methods, and interviewer training are a few of them. However, each of the various interviewing methods has certain inherent strengths and weaknesses in terms of data quality. These strengths and weaknesses are summarized in Exhibit 6.7. The important point here is that the issue of data quality may override other considerations such as cost. For example, although the least expensive way to get responses to a long questionnaire with many open-ended questions might be via a mall-intercept interview, thedata obtained by this method might be so biased because of respondent fatigue, distraction, and carelessnes that the results would be worthless at best and misleading at worst
Budget
The commercial marketing researcher frequently encounters situations in which the budget available for a study has a strong influence on the survey method used. For example, assume that for a particular study the budgetary constraint for interviewing is $10,000 and the sample size required for the necessary accuracy is 1,000. If the cost of administering the questionnaire using the mall-intercept method is $34.50 per interview and the cost of administering it via Internet survey is $1.50 per interview, the choice is fairly clear assuming that nothing about the survey absolutely requires face-to-face contact.
Requirements for Respondent Reactions
In some studies, the marketing researcher needs to get respondent reactions to various marketing stimuli perhaps product prototype usage (a new style of smartphone keyboard) or a taste test. In these cases, the need to get respondent reactions to stimuli normally requires personal contact between interviewer and respondent. Taste tests typically require food preparation. This preparation must be done under controlled conditions so that the researcher can be certain that each person interviewed is responding to the same stimulus. The only viable survey alternative for tests of this type is the mall-intercept approach or some variant. One variant, for example, is recruiting people to come to properly equipped central locations, such as community centers, to sample products and be interviewed. Some surveys require face-to-face interaction because of the need to use special measurement techniques or eye tracking devices, or to obtain specialized forms of information. The tasks are so complex that the interviewer must be available to explain the tasks and ascertain whether the respondents understand what is required of them.
Quality of Data
The quality of data required is an important determinant of the survey method. Data quality is measured in terms of validity and reliability. Validity refers to the degree to which a measure reflects the characteristic of interest. In other words, a valid measure provides an accurate reading of whatever the researcher is trying to measure. Reliability refers to the consistency with which a measure produces the same results with the same or comparable populations.
Many factors beyond the interviewing method affect data quality. Sampling methods, questionnaire design, specific scaling methods, and interviewer training are a few of them. However, each of the various interviewing methods has certain inherent strengths and weaknesses in terms of data quality. These strengths and weaknesses are summarized in Exhibit 6.7. The important point here is that the issue of data quality may override other considerations such as cost. For example, although the least expensive way to get responses to a long questionnaire with many open-ended questions might be via a mall-intercept interview, thedata obtained by this method might be so biased because of respondent fatigue, distraction, and carelessnes that the results would be worthless at best and misleading at worst
Length of the Questionnaire
The length of the questionnaire the amount of time it takes the average respondent to complete the survey is an important determinant of the appropriate survey method to use. If the questionnaire for a particular study takes an hour to complete, the choices of survey method are extremely limited. Telephone, mall-intercept, and most other types of surveys, with the exception of personal interviews, will not work. People shopping at a mall ordinarily do not have an hour to spend being interviewed. Terminations increase and tempers flare when interviewers must try to keep respondents on the phone for an hour. Response rates plummet when people receive through the mail questionnaires that take an hour or more to complete. The trick is to match the survey technique to the length of the questionnaire. While there is no hard-and-fast rule, the following maximum lengths have been recommended:
▪ Face-to-face (in-home, office) 25 minutes
▪ Telephone 20 minutes
▪ Mall intercept 15 minutes
▪ Online 20 minutes
▪ Individual depth interview 40 minutes
Incidence Rate
Recall that the incidence rate refers to the percentage of people, households, or businesses in the general population that would qualify as interviewees in a particular study. Search costs, which correlate with the time spent trying to locate qualified respondents, sometimes exceed the costs of interviewing. In situations where the researcher expects incidence rates to be low and search costs high, it is important to select the method or combination of methods that will provide the desired survey results at a reasonable cost. Doing a low-incidence rate study in a mall would be very expensive. This approach should be taken only if there is some compelling reason for doing so a long in-depth interview, for example. The lowest-cost survey alternative for the low-incidence study is probably the Internet panel, assuming that this approach meets the other data-collection requirements of the study. One advantage of the Internet panel is that it can be prescreened; people can be asked a number of questions, usually including some on product usage. For example, if panel members were asked during prescreening whether anyone in their household participated in downhill or alpine skiing, the Internet panel operator could at very low cost pull out only those households with one or more skiers for a survey of Alpine
skiers.
Structure of the Questionnaire
In addition to the length of the questionnaire, the degree of structure required in the questionnaire may be a factor in determining which survey method is most appropriate for a given study. Structure refers to the extent to which the questionnaire follows a set sequence or order, has a predetermined wording of questions, and relies on closed-ended (multiplechoice) questions. A questionnaire that does all these things would be structured; one that deviates from these set patterns would be considered unstructured. A questionnaire with little structure, such as an individual depth interview, requires a face-to-face interview. Very brief, highly structured questionnaires do not require face-to-face contact between interviewer and respondent. Mail, telephone, self-administered, and online surveys are viable options for studies of this type.
Time Available to Complete the Survey
If the client needs to have survey results quickly, the Internet is the best choice. Generally, call center telephone and mall-intercept interviews can also be completed in a timely manner.
The length of the questionnaire the amount of time it takes the average respondent to complete the survey is an important determinant of the appropriate survey method to use. If the questionnaire for a particular study takes an hour to complete, the choices of survey method are extremely limited. Telephone, mall-intercept, and most other types of surveys, with the exception of personal interviews, will not work. People shopping at a mall ordinarily do not have an hour to spend being interviewed. Terminations increase and tempers flare when interviewers must try to keep respondents on the phone for an hour. Response rates plummet when people receive through the mail questionnaires that take an hour or more to complete. The trick is to match the survey technique to the length of the questionnaire. While there is no hard-and-fast rule, the following maximum lengths have been recommended:
▪ Face-to-face (in-home, office) 25 minutes
▪ Telephone 20 minutes
▪ Mall intercept 15 minutes
▪ Online 20 minutes
▪ Individual depth interview 40 minutes
Incidence Rate
Recall that the incidence rate refers to the percentage of people, households, or businesses in the general population that would qualify as interviewees in a particular study. Search costs, which correlate with the time spent trying to locate qualified respondents, sometimes exceed the costs of interviewing. In situations where the researcher expects incidence rates to be low and search costs high, it is important to select the method or combination of methods that will provide the desired survey results at a reasonable cost. Doing a low-incidence rate study in a mall would be very expensive. This approach should be taken only if there is some compelling reason for doing so a long in-depth interview, for example. The lowest-cost survey alternative for the low-incidence study is probably the Internet panel, assuming that this approach meets the other data-collection requirements of the study. One advantage of the Internet panel is that it can be prescreened; people can be asked a number of questions, usually including some on product usage. For example, if panel members were asked during prescreening whether anyone in their household participated in downhill or alpine skiing, the Internet panel operator could at very low cost pull out only those households with one or more skiers for a survey of Alpine
skiers.
Structure of the Questionnaire
In addition to the length of the questionnaire, the degree of structure required in the questionnaire may be a factor in determining which survey method is most appropriate for a given study. Structure refers to the extent to which the questionnaire follows a set sequence or order, has a predetermined wording of questions, and relies on closed-ended (multiplechoice) questions. A questionnaire that does all these things would be structured; one that deviates from these set patterns would be considered unstructured. A questionnaire with little structure, such as an individual depth interview, requires a face-to-face interview. Very brief, highly structured questionnaires do not require face-to-face contact between interviewer and respondent. Mail, telephone, self-administered, and online surveys are viable options for studies of this type.
Time Available to Complete the Survey
If the client needs to have survey results quickly, the Internet is the best choice. Generally, call center telephone and mall-intercept interviews can also be completed in a timely manner.
No comments:
Post a Comment