marktforschung.de
Interview with Laura Kern, GIM "Experiments have largely replaced introspective procedures"
Interview with Laura Kern, GIM
"Experiments have largely replaced introspective procedures"
The choice of method increasingly falls on experiments - why? For Laura Kern, GIM, the greatest advantage of the method lies in the possibility of ascertaining causalities. The typical question-and-answer game has come to an end, and questions to which the respondent cannot know the answer find answers.
What are experimental approaches? And what distinguishes them from other quantitative approaches/methods?
Laura Kern: In experimental approaches we break away from the typical "question-answer" principle and generate answers without asking respondents directly. Example: We want to make conclusions about the influence of a single design element of an ad on the overall perception of the ad: we do not get reliable answers from respondents. With experimental approaches, on the other hand, we can vary and combine individual elements of the ad, such as the logo, color scheme, illustration, font, additional elements such as the organic seal, etc., so that we can ultimately make statements about each individual element and their interactions (e.g. font and illustration). The main advantage is thus that causal statements are possible. Causality does not result from introspection, but from the fact that possible influencing factors are systematically varied in the design of the experiments. In a nutshell: For experimental approaches we dissolve existing structures, vary individual elements in a new way in order to be able to measure the individual effects of these elements.
How can responses within experimental study approaches be measured?
Laura Kern: Responses can be measured by a standardized scale, by a change in behavior, or by reaction time. For example, we can use a scale to ask how typical an ad shown is for a brand or whether respondents would be more likely to buy or not buy the product shown. The reaction time until a brand was attributed an image can also be recorded using keystrokes. The exciting thing about experiments, however, is usually the influencing factors that we consciously take apart and recombine, since the measure or reaction is downstream.
In which settings do you conduct experiments? Have experiments been possible at all in recent years?
Laura Kern: Experiments can be conducted in any setting. It can be face-to-face or online. In other words, experiments are possible wherever responses and behavior can be measured or observed - in fact, anywhere.
For what kind of questions does it make most sense to avoid direct questioning?
Laura Kern: The typical question here is: "What influence does ... have on ...?" because respondents can seldom name the exact magnitude of influence of individual elements on the overall perception. Incidentally, a conjoint is also an experimental approach, because here existing products are "broken down" into their individual elements and recombined. This variation can be used to measure the influence of product features such as price or brand on purchase decisions.
Which topics are particularly suited to the use of experiments?
Laura Kern: A typical field of application is the topic "design and effect of design (elements)". Experiments are particularly useful when specific design elements are to be evaluated in terms of their effect. One question, for example, would be: How big is the effect of an organic label on the perception of the product as "sustainable"? And again, more fundamentally, experiments are always appropriate when the question is not accessible by introspection, i.e., when a person cannot know the answer by inwardly directed observation.
What is most important in a good questionnaire design?
Laura Kern: That you only ask questions that respondents can answer! This sounds trivial at first, but in everyday research it is anything but self-evident. The relatively simple and typical question "How has your attitude changed?" requires that respondents can remember what their attitude was like in the past (at whatever point in time) and what it is like today. This might still work for basic beliefs, such as attitudes toward abortion, but probably much less well for chocolate brands. Also difficult: statements about causality, or the question "For what reasons did you ...?" With such questions, one can elicit implicit explanatory models of the respondents, but not causalities, since these are usually not consciously accessible.
How do experimental methods fit with quantitative studies?
Laura Kern: Standardization and high numbers of cases are key characteristics of experimental methods. However, they can of course be combined with other quantitative methods based on introspection.
Why should your web seminar not be missed?
Laura Kern: Experiments have become the method of choice in many disciplines and have largely replaced introspective procedures. In the web seminar, we explain the use of experimental approaches as well as their implementation and evaluation possibilities using several examples. We thus show how research questions can be answered by deliberately varying elements without having to ask "classical questions" by means of introspection.
ABOUT
Laura Kern is a Research Manager at GIM's Heidelberg office and has been responsible for conducting quantitative ad hoc projects - predominantly in the FMCG sector - since 2019. Previously, she studied "BW / Market Research and Consumer Psychology" (B.A.) and "Creative Communication and Brand Management" (M.A.) at Pforzheim University.
Interview published in German on marktforschung.de
