Gaissmaier, W. (2007). The mnemonic decision maker: How search in memory shapes decision making. Doctoral dissertation, Freie Universität Berlin, Germany.
The starting point of the dissertation was the question how people search for information in memory when they make decisions. Following the perspective of ecological rationality (e.g., Gigerenzer et al., 1999), successful decision strategies are anchored both in the human mind and in the environment. Adaptive decision making thus requires that people adapt their strategies both to the structure of the environment and to the limitations of the cognitive system. In this regard, I am sympathetic to the view put forward for example by Schooler and Hertwig (2005) that these limitations may be functional. Among other functions, they shape how people search for information in memory by facilitating certain ways of searching for information, but hindering others.
In Chapter 1, I have explored the counterintuitive finding that people with a lower short-term memory capacity outperform people with a higher short-term memory capacity in a correlation detection task (Kareev et al., 1997), which is highly similar to classic binary choice probability learning tasks. It turned out that the success of people with lower capacities lies in the simplicity of their strategy. In contrast, people with higher capacities explored too much and looked for patterns in the sequence of events and ended up with behavior that looked like probability matching. Since there were no patterns, this more exploratory behavior was counterproductive in this task. It helped, however, to detect changes in the environment. It could very well be that the process underlying probability matching, searching for patterns, is usually smart, because often the cost of missing a non-random sequence could well be higher than the price of detecting patterns where there are none (Lopes, 1982). But it looks stupid in stationary binary choice tasks with conditions that rarely hold outside of psychological laboratories and casinos (Ayton & Fischer, 2004). Probability matching, or its underlying process, could thus smarter than it appears at first glance.