The Effects of Strategy Training and Text Organization on Mental Models of Natural Causal Complex Systems

Abstract

The Effects of Strategy Training and Text Organization on Mental Models of Natural Causal Complex Systems Cathy Jalali In the absence of direct experience with physical stimuli, people generate mental models to facilitate cognitive processes involved in problem solving. Mental models are viewed as internal representations that help people understand, reason about, predict and investigate causal and functional relationships contained within physical systems. In addition, more recent findings distinguish mental models from other mental representations by suggesting that people actively construct mental models, switch between different mental models of a given problem, and alternate between mental models and rules to solve problems. Piecemeal animation of a system’s action has been identified as the cognitive strategy people use to reason about dynamic physical systems. Emphasizing function over form in multimedia presentations about physical systems has been shown to improve the leaners’ mental models. Novice leaners’ mental models are reported to emphasize structure, while the experts’ mental models include more functional and behavioral elements. Mental models of causal complex systems that operate in perceptually novel environments remain largely unstudied. Two experiments were conducted to study peoples’ mental models of single and functionally coupled causal complex systems and investigate the influences of text organization and strategy training on novice learners’ mental model development. A modified version of the Structure Behavior Function (SBF) framework was developed to facilitate the investigation of mental models of causal complex systems. Results suggest that in the absence of explicit instruction, people utilize a combination of cognitive strategies ranging from rote memorization to piecemeal simulation of the systems’ functions to form incomplete, but coherent mental models of causal complex systems from text and static diagrams. A brief training session was effective in teaching novice learners to use envisioning strategy and help promote piecemeal animation of the system. Envisioning strategy training significantly improved the subjects’ mental model complexity as well as knowledge of structural, functional, and mechanistic components involved in achieving the systems’ overall functions in both single and functionally coupled causal complex systems (p<0.05). For the single causal complex system (CCS), function-salient text improved subjects’ structure and function subscores significantly (p<0.05), but did not affect the mechanism subscore. The interaction between text and strategy training was not significant. For the functionally coupled causal complex system (FC-CCS), text was not a significant predictor of mental model complexity or its subcomponents. The interaction between text and strategy training was also insignificant. Though subjects’ mental models included more structural than functional elements for both CCS and FC-CCS, majority demonstrated understanding of the overall systems’ functions (formation of end products). The subjects appeared to pay selective attention to the change producing steps and points of functional coupling (product of the first system, which serves as the template for the second system).