Page 83 - Design for Six Sigma for Service (Six SIGMA Operational Methods)

64 Chapter Four

For example, the population of fast-food eaters may include all people except
homeless and sick people. But if a telephone interview is to be conducted, only
the people with known telephone number can be reached, so “people with a
telephone number” will be the sampling frame for fast-food eaters. The concepts
of population and sampling frame will be discussed further in Sec. 4.4.

Stage 5: Determination of Sample Size and Sample Selection Procedure
The survey researcher will have to select a sample that adequately represents
the population under study. In general, larger samples will yield greater
accuracy than small samples in terms of analysis results. The sample size is
usually determined by balancing between analysis accuracy and the increased
cost and time due to larger sample size. Once the sample size is determined,
the method of sampling will be determined. The commonly used sampling
methods include random sampling, stratified random sampling, and cluster
sampling. The sample size determination and sampling methods will be
discussed in Sec. 4.4.

Stage 6: Design of the Survey Instrument

The development of a survey instrument or questionnaire is a key component
of the customer survey. At this stage, a survey researcher must design a series
of unbiased, well-structured questions that can systematically obtain the
information based on the survey goals and objectives developed in stage 1.
The input from the focus group is a major source for survey questions. The
development of a questionnaire can be an extremely detailed and time-
consuming process. The best-designed questionnaire should be short and
concise with well-worded questions. Long and wordy questionnaires will
often result in a lower response rate and a higher survey cost. The design of
the survey instrument is discussed in Sec. 4.2.

Stage 7: Pretest of the Survey Instrument

After a draft questionnaire is designed, it is important to pretest the ques-
tionnaire with a small group of respondents. During this pretest, poorly
worded questions will be identified and refined. The refined questionnaire
will have a better quality.

Stage 8: Selection and Training of Survey Interviewers

For telephone and in-person interviews, trained interviewers are required.
The source of interviewers can be college students, part-time workers, and
so on. Interviewers should be familiar with the questionnaire and know how
to handle uncooperative respondents.