<p>Theories predicting performance of human control of complex dynamic systems must assess how decision makers capture and utilise knowledge for achieving and maintaining control. Traditional problem solving theories and corresponding measures such as Ravens matrices have been applied to predict performance in complex dynamic systems. While they assume stable properties of decision makers to predict control performance in decision-making tasks these tests have shown to provide only a limited degree of prediction in human control of complex dynamic systems. This paper reviews theoretical developments from recent empirical studies and tests the theoretical predictions of a model of dynamic decision-making using a complex dynamic microworld – Moro. The requirements for control of the microworld is analysed in study one. Theoretical predictions from the reviewed theory and results from study one are tested in study two. In study three additional hypotheses are derived by including meta cognitive dynamics to explain anomalies found in study two. A total of 21 Hypotheses are tested. Results indicate that for predicting human control of complex dynamic opaque systems a number of meta cognitive processes play an important role in determining outcome. Specifically, results show that we cannot expect a lower risk of failure in complex dynamic opaque systems from people with high problem solving capabilities when these also express higher goals. Further research should seek to explore the relative contribution of task characteristics to determine conditions under which these meta cognitive processes of decision makers take a dominant role over problem-solving capabilities – enabling improved decision-maker selection and support.</p> / Rapportkod: LiU-Tek-Lic-2002:4. I den tryckta versionen är bilagorna bifogade som CD-skiva. Den elektroniska versionen innehåller den kompletta avhandlingen.

Theories predicting performance of human control of complex dynamic systems must assess how decision makers capture and utilise knowledge for achieving and maintaining control. Traditional problem solving theories and corresponding measures such as Ravens matrices have been applied to predict performance in complex dynamic systems. While they assume stable properties of decision makers to predict control performance in decision-making tasks these tests have shown to provide only a limited degree of prediction in human control of complex dynamic systems. This paper reviews theoretical developments from recent empirical studies and tests the theoretical predictions of a model of dynamic decision-making using a complex dynamic microworld – Moro. The requirements for control of the microworld is analysed in study one. Theoretical predictions from the reviewed theory and results from study one are tested in study two. In study three additional hypotheses are derived by including meta cognitive dynamics to explain anomalies found in study two. A total of 21 Hypotheses are tested. Results indicate that for predicting human control of complex dynamic opaque systems a number of meta cognitive processes play an important role in determining outcome. Specifically, results show that we cannot expect a lower risk of failure in complex dynamic opaque systems from people with high problem solving capabilities when these also express higher goals. Further research should seek to explore the relative contribution of task characteristics to determine conditions under which these meta cognitive processes of decision makers take a dominant role over problem-solving capabilities – enabling improved decision-maker selection and support. / <p>Rapportkod: LiU-Tek-Lic-2002:4. I den tryckta versionen är bilagorna bifogade som CD-skiva. Den elektroniska versionen innehåller den kompletta avhandlingen.</p>