Quantitative Research: Closed-end alternatives for a market research campaign

This is the concluding eTip! of the three -part series on quantitative market research. October's issue described quantitative research in general terms, specifically its differences from qualitative methods and its value in predicting the effect of marketing decisions. November's issue illustrated the value created when closed-and open end questions are carefully meshed within a given study.

Our most recent eTip! focused on the use of closed- and open-end questions in quantitative research and the benefits and limitations associated with each question type. We now direct attention to some of the most commonly used closed-end questions, specifically, and their uses in data collection. Generally, six question types are utilized when conducting quantitative research: rating scale, dichotomous, paired preference, forced preference, list, and numeric.

Rating scale questions present a fixed scale that participants choose from when evaluating an attribute or feature of a product, company, etc. These questions are constructed to measure agreement, likelihood or other perceptual and behavioral attitudes. Scale ranges generally use 5, 7 or 9 points on a continuum, with lesser values indicating a "negative" rating and higher values indicating a "positive" rating. Odd number scales are preferable, as they allow for a distinct neutral rating point on the continuum. When quantitative data is aggregated, rating scale questions are often calculated to yield a mean rating which is evaluated in context to the scale.

Dichotomous questions allow only one response and are most frequently seen as "yes/no" questions. These questions "force" participants to commit to one response and are useful when the question topic does not require the measurement of multiple options. Similar to dichotomous questions, paired preference questions offer an either/or choice; however, they present two products, attributes or features with only one selection allowed so that preference may be determined.

Forced preference questions "force" a choice among a list of potential features, products, companies, etc. These questions differ from paired preference questions in that several choices may be offered; however, the participant may still select only one option. This question type identifies the one preeminent option when multiple options are available. List questions also present a fixed list of options to a participant but allow selection of one or more responses from that list. This is particularly useful when assessing advertising message recall or other subject areas where multiple options may be possible.

Finally, numeric questions are those that request a single numeric response, such as percent of time spent on a task, number of patients, etc. This question type is particularly suited for demographic questions where numeric value ranges may be extensive. Mean values and ranges for these questions may be used to segment quantitative data into categories for analyses as well.

This palette of closed-end questions offers the researcher a comprehensive array of ways to assess the beliefs and attitudes of a given sample. A qualified survey designer will use these tools to directly reflect the objectives of the research project. When reviewing your research instrument, your project manager should be able to readily explain the logic behind each question's design.