Quantitative Marketing Research

During the next four weeks of class you will 1) learn about quantitative marketing research, 2) design an online survey, 3) administer your online survey, and 4) analyze your online survey data, and 5) present your results with cool charts and graphs.

I used SurveyMonkey while working on my MBA, but it looks like PollDaddy might be more appropriate for the purposes of this class–creating a web survey that users can conduct while they remain on your tumblr website.  I looked at some alternatives to PollDaddy.  If you are already using Vertical Response or MailChimp then you might want to do your survey with these services.  If you have a WordPress website the WP-Polls plugin seems popular.

Welcome to the world of quantitative marketing research (based on questioning).  Bosses like numbers, and quantitative marketing research techniques give the boss numbers and statistics.  A good example of the power of numbers is discussed in Words That Work by Frank Luntz.  The book The Emperor of Wine discusses the power of the 100 point scale in wine ratings.  But, quantitative marketing research isn’t perfect.  Marketers can come up with some pretty weird scales like the lovemarks scale discussed in The Lovemarks Effects (same guy from The Persuaders).

Some things to consider when designing your questionnaire:

• Questions to include:  What do you think is The Ultimate Question to include in a marketing survey?  It’s good practice to use qualitative/exploratory research results to make your questionnaire.  We’ll discuss content analysis of qualitative research results and do a class exercise on content analysis.  Let’s take a look at some questionnaire construction guidelines.
• Scaling your questions: Marketing academics like to use multiple-item scales (like self-esteem, service quality, the lovemarker, etc.), as opposed to single-item scales (like The Ultimate Question) because it’s easy to measure construct reliability and validity.  You’re probably going to use single-item scales for this class.  Let’s take a look at comparative and noncomparative scaling techniques.  Let’s take a look at some more scaling considerations.
• Levels of measurement: Analyzing ordinal data is not as easy as analyzing interval/ratio data, so you should try to use Likert or semantic differential scaling.
• Worrying about scaling and levels of measurement is important because the types and number of variables you include in your questionnaire will determine which statistical techniques you can use.  We’ll spend more time on statistics later.

If you would like to gather some freelist data in class to do a content analysis on let me know.

Next lab we’ll go over designing an online survey with PollDaddy.  Please come to class prepared with the items/questions you want to include in your questionnaire and know how you are going to scale your items/questions.  Try to include some relevant demographic background questions in your questionnaire (age, gender, etc.).