Professional Development and Mba

Conducting Survey-Interviews
Assignment #1
Interviews can be incredibly valuable research tools, but there are definitely problems associated with them. For one, in an interview, there is opportunity clarify the question, so if the respondent needs clarification, it can be obtained. This is unlike most other forms of research. The interviewer can also encourage participation, and there is the opportunity for deeper probing of a subject. However, the interview could lead or bias the interviewee, skewing the results, and with an interviewer the respondent could be either more open or less open. That ends up being dependent on the interviewer and the interviewee, depending on subject matter and overall context.
Surveys have some merits, too, however. They are easier to administer, and less time-consuming, so the cost per response is usually much lower than for interviews. However, there is a fairly low response rate, and many times a survey risks falling below the threshold for statistical validity. Further, there is less opportunity to clarify questions, so careful attention must be paid to the wording. So there is a trade-off between interviews and surveys. Both can, however, deliver in-depth answers, if used correctly.
For expedience, I would use a survey. First, a survey allows me to craft the questions precisely, whereas in an interview there is the risk that I would get off track. Further, I feel that gathering a large number of responses is important. With quantitative results, you really have to emphasize the need to get a statistically valid sample size.
There is absolutely no need to use just one type of question in a survey. That does not even make sense. With a survey, there exists the ability to vary the type of response to fit the question, and that is what I will do. If the question calls for a Likert scale, that will be used. If it calls for an open-ended question, that is what I will do. This is the beauty of writing out a survey – you have a lot of flexibility with the design and can tailor each individual question to the type of information that you are seeking.
Data collection is going to be manual. Basically, in a relatively small survey the cost of automated data collection is not worth the value. It might be a bit tedious to perform the analysis manually, but ultimately that is the best way to make sure you get all the data. Plus, if there are any open-ended questions then those should be evaluated manually.
That does not mean that any statistics will be done by hand; far from it. Excel will be used to handle simple statistics. If SPSS is needed, it can be used, but for anything up to a single regression, Excel is fine as a means of generating the numbers to analyze the different responses.