1 LaTina – An online system for teaching and learning stock-and- flow thinking skills Martin Schaffernicht Universidad de Talca Talca – Chile martin@utalca.cl Abstract This paper continues a line of work that took up previously published stock-and-flow thinking studies and proposed to apply the model of implicit learning to the case. According to this model, novices have to elaborate personal experience by following rules. After previous trials, a set of such rules is proposed together with a group of challenges that allow to apply and to learn them. A conceptual model for representing the rules, the challenges, the learners and their learning itinerary are proposed. Then, the design for an on-line system for publishing and working with challenges and to monitor progresses is introduced. This software system is currently under construction. Keywords: system thinking skills, tacit knowledge, learning, expertise Introduction Stock-and-flow thinking has been found to be a tricky business for most human beings (Booth-Sweeny and Sterman, 2000; Ossimitz, 2002; Kainz and Ossimitz, 2002): in all the cases studied, subjects failed in several vary fundamental tasks involving one stock and one or two flows. Schaffernicht (2005a) proposed to use Dreyfus and Dreyfus’ (1986) approach to skill acquisition, based on a review of Polanyi’s model of implicit integration (Polanyi, 1966; Neuweg, 1999): subjects transform themselves from novices into experts over several stages, by experience and reflection. The novice is given general and context-free rules to classify situations and determine courses of action. The advanced novice has already a stock of personal references and begins to “compile” his or her own directives (less specific than rules in this terminology). The competent has advanced in this same direction and starts recognizing situations implicitly. Finally the expert not only recognizes what the case is, he also knows what to do. Schaffernicht (2005a suggested a set of rules and an adaptation of Ossimitz’s tests in order to try this out. As reported in Schaffernicht (2005b), following the indicated rules appears to improve the results in the tests; however, the main finding was that instead of having subjects work through the battery of tests and then giving them feedback, assessment and feedback should