Measuring Constructs

Feelings & Sensations: Where Survey Designs Fail Badly

Survey research is pretty good at allowing people to describe “things” in such a way that the researcher winds up with a fairly accurate idea of the thing being described. The most straight-forward example is a survey question that asks, “Which of the following features came with Hotel experienceyour new Toyota Corolla?” followed by a list of possible features. However, survey research can also get at descriptions of more experiential phenomena with questions such as, “On a scale from ‘1’ to ‘5’, how does each of the following statements describe your experience in buying a new home?” In these cases, the use of survey methods to research a great number of people, and compile and report the data as efficiently as possible, make good use of closed-ended questions to gain an understanding of respondents’ accounts of the “things” of interest. This can also be said of beliefs. Pew’s recent survey pertaining to the Christmas story that asked, Read Full Text

Satisfaction Research & Other Conundrums

Greg Allenby, marketing chair at Ohio State’s business school, published an article in the May/June 2014 issue of Marketing Insights on heterogeneity or, more specifically, on the idea that 1) accounting for individual differences is essential to understanding Conundrumthe “why” and “how” that lurks within research data and 2) research designs often mask these differences by neglecting the relative nature of the constructs under investigation. For instance, research concerning preference or satisfaction is useful to the extent it helps explain why and how people think differently as it relates to their preferences or levels of satisfaction, yet these are inherently relative constructs that only hold meaning if the researcher understands the standard (the “point of reference”) by which the current question of preference or satisfaction is being weighed – i.e., my preference (or satisfaction) compared to…what? Since the survey researcher is rarely if ever clued-in on respondents’ points of reference, it would be inaccurate to make direct comparisons such as stating that someone’s product preference is two times greater compared to someone else’s.

The embedded “relativeness” associated with responding to constructs such as preference and satisfaction is just one of the pesky problems inherent in designing this type of research. A related but different problem revolves around the personal interpretation given Read Full Text