GfA, Dortmund (Hrsg.): Frühjahrskongress 2019, Dresden Beitrag B.7.3 Arbeit interdisziplinär analysieren – bewerten – gestalten 1 Studying Individual Satisficing-Optimizing Decisions in an Uncertain Explore/Exploit Information Environment Thomas WIEBRINGHAUS ifes – Institut für Empirie & Statistik, FOM University of Applied Sciences Herkulesstr. 32, D-45127 Essen Abstract: One of the most prevalent cognitive tasks in human-computer interaction (HCI) is seeking information using web searches. Although simple information retrieval is trivial, more complex questions evolve by new meaningful information. The optimal balance to exploit the current resource for gathering the available information or to explore new information is a main topic in Reinforcement Learning Theory (e.g. motion planning for robots), Information Foraging Theory (e.g. improving search engines) and Usability/ UX. Because people are rationally bounded by restricted capacity, information and time, the optimal search strategy is intractable for humans in search decisions under uncertainty. The aim of this short paper is to present and discuss first observations of a calibration test to study empirically the individual satisficing-optimizing relationship. Keywords: Satisficing, Uncertainty, Optimal Foraging, Web Search 1. Introduction and Aim 1.1 Information Foraging Theory and Satisficing vs. Optimizing Information Foraging Theory provides a framework where the exploration of the information space is similar to the process when animals search for new food patches, and the exploitation of the information is described as depleting the food patch (Stephens & Krebs 1986; Pirolli & Card 1999). The amount of a single information source is restricted, resulting in switching the food source to explore other patches as information gain diminishes with time. Charnov’s Marginal Value Theorem (Charnov 1976) states that a forager should leave the patch when benefits and costs are equal (Olsson & Brown 2006), that is when the gain of the current patch is lower than the average gain. However, the decision to explore/ exploit depends on the information environment and the available resources of the forager (Cohen et al. 2007; Hills et al. 2015). Human beings are rationally bounded because of limited processing and working memory capacity, restricted information of environmental structure and confined time horizons for finding an optimal solution for a decision task. People therefore do not maximize the reward or expected utility, but are rather satisfied with a sufficient good enough solution depending on their aspiration level, a process called satisficing (Simon 1955, Ward 1992). From a computational perspective, optimizing means choosing the best solution after testing all the available options (maximizing). However, searching for the optimum is costly for human beings (Gorod et al, 2017; Shervais & Shannon 2012).