Dr. Thierry Worch
When consumer scientists encounter data science, they are often led to the encounter by questions such as:
Which analysis should I perform? What form should my data take? How should I run this analysis? While these questions are
all valuable, they disconnect various parts of data science - data collection, data formatting, data analysis, and so on - by
treating them as unrelated to other parts - objectives of the test, data collection methodology and so on. But what if these
parts were, in fact, all interrelated? What if quality data science was, in fact, an integral part of quality consumer science?
Through the examples of several real-life consumer studies, we show in this talk how data science influences consumer science from the very beginning, as the way data are collected influences the questions the data can answer. Moreover, the objectives of a study determine the general choice of the data collection tool and the specifics of that tool - experimental design, sample size, number of products, scale used, and so on - as well as the analyses to perform and the interpretation of results.