Relevance & Research Question
In brand awareness research, even universally recognized brands that should reach maximum awareness (for example, Google among internet users) often fall short in actual surveys. This suggests that true awareness may be systematically underestimated. Participants in online research panels frequently take part in surveys and are highly familiar with questionnaire formats. Therefore, respondents may adopt behaviors that reduce cognitive effort or optimize earnings per hour. Our research examines whether this underreporting reflects conscious strategic choices or whether it arises more implicitly through mechanisms such as fatigue.
Methods & Data

We embedded a survey experiment in a client study among 996 driving license holders in Switzerland from the nationally representative YouGov panel. Respondents were randomly assigned to one of four experimental groups. We used a 2x2-design with list length (either 8 or 24 brands) and a behavioral nudge (vs. no nudge), stating that the questionnaire length and payout were independent of their awareness of the specific brands) as the experimental conditions.
Audi, BMW and Mercedes-Benz can be assumed to have complete awareness among the target population. If the nudge increased reported awareness, this would indicate deliberate effort reduction. If the shorter list increased awareness, this would point to implicit mechanisms such as fatigue. We estimated the effects of the experimental conditions using a linear regression with the number of these brands being reported as recognized as the dependent variable.

Results
On average, respondents recognized 2.85 of the three brands. The linear regression indicates that the shorter list leads to significantly higher reported awareness among the three brands in this condition (p < 0.05). The behavioral reminder does not show a positive effect on reported brand awareness.
Added Value
The effect of list length indicates that respondents do not intentionally underreport awareness. Instead, the results suggest that underreporting is driven primarily by implicit mechanisms such as cognitive fatigue and reduced attentional effort. This implies that survey designers can improve data quality by using shorter and/or splitting brand lists for subgroups wherever possible. By minimizing mental load, researchers can obtain awareness estimates that more closely reflect true recognition levels.

Relevance & Research Question: Behavioural campaigns – communication initiatives aimed at changing behaviours – play a vital role in both social and market research. Their evaluation requires a holistic approach integrating behavioural, cognitive, and emotional indicators. This is particularly challenging when topics are sensitive to social desirability or strategic responding, such as traffic safety or pricing research. Traditional explicit self-reports (System 2) provide valuable insights into reflective attitudes/intentions but are prone to response bias and often overlook intuitive or automatic aspects of behaviour. Implicit methods complement them by capturing unconscious processes (System 1). This presentatoin examines how explicit and implicit methods can be combined to obtain a more comprehensive picture of behavioural campaign effects. We explore this in the context of a Swiss road-safety campaign addressing cyclists’ “illusion of control” – the tendency to overestimate control in traffic situations and underestimate associated risks.

Methods & Data: The evaluation follows a quasi-experimental repeated cross-sectional design combining explicit and implicit measures. Four online panel surveys are conducted between 2025 and 2027 in Switzerland (intervention) and Germany (control), each with 1’000 regular bicycle or e-bike users. Three outcome categories are assessed: knowledge, attitudes/intention, and behaviour. Explicit measures capture self-reported knowledge, perceived risk, and behavioural intentions. Implicit measures employ a reaction-time-based Single Association Test (SAT) with scenario vignettes covering four risk-prone situations: cycling without a helmet, mobile phone usage while cycling, running a red light, and riding under the influence of alcohol. These vignettes assess implicit risk perception and control illusions. The design enables pre/post comparisons between Switzerland and Germany to infer potential causal effects of the campaign over time.

Results: Preliminary findings show that explicit and implicit measures can be integrated within one online survey. Scenario-based SATs complement self-reports by revealing intuitive risk perceptions and control beliefs. Data from two waves (pre- and first post-campaign) provide first insights into early campaign effects.

Added Value: The study demonstrates a scalable framework for embedding implicit measures in large-scale online surveys. Combining scenario vignettes with reaction-time methods offers a practical blueprint for evaluating sensitive behavioural topics with higher ecological validity and depth.

Relevance & Research Question: Aesthetics is quickly processed, has a multitude of consequences for attitudes, recommendations, and purchase behavior–and thus is paramount for the success of products. However, differentiated and validated evaluation tools are lacking, particularly for interactive products. In order to close this gap, we aimed to develop and validate the Product Aesthetics Inventory (PAI) and a short version (PAI-S).

Methods & Data: Items were developed in a pre-study (N=6 design experts, N=4 product users). The resulting item pool with 54 questions was tested in an online survey on different types of household appliances (alongside with several validation measures; N=6,002). In Study 1, data from n=3,000 of these participants were used to determine the number of factors using exploratory graph analysis. In Study 2, the found structure was validated in a confirmatory factor analysis with the remaining n=3,002 participants. A third web-based study (N=1,028) was conducted to further determine construct validity and generalizability among different products such as IT products, home entertainment devices, and power tools.

Results: The final PAI consists of 34 questions covering eight dimensions: visual aesthetics, operating elements, brand logo, feedback sounds, operating noises, haptics, interaction aesthetics, and impression. A higher-order factor of product aesthetics can be determined both with the full PAI and with the 8-item short version PAI-S. Both versions demonstrate excellent reliability, as estimated by Cronbach’s alpha. The decrease in reliability from the full scale (.97) to the short scale (.89) is acceptable. Further, all eight subscales of the PAI achieved good to excellent reliability (range: .85-.92). Validity was confirmed by corresponding correlations with other established scales, intention measures, and overall judgments. Further, Study 3 confirms and extends these results with respect to reliability and validity. Finally, in all three studies, strict measurement invariance was achieved across different product classes for the majority of subscales.

Added Value: The PAI meaningfully supplements the measurement spectrum of product evaluation beyond classic usability and brand perception. Questionnaire templates, evaluation guides and interpretation aids are freely available online (https://doi.org/10.5281/zenodo.6478042). Further, we will complement this by discussing additional options of response scaling and benchmarks for the PAI.