This research was conducted to determine if students receiving complex systems instruction in the form of software simulations recognize patterns and underlying elements of complex systems more effectively than students receiving traditional instruction. Complex systems were investigated with an analytic (reductive) approach in a control group and with a synthesis approach in the treatment group. Exploration of this top-down approach to learning complex systems counters traditional bottom-up methodologies, investigating systems and subsystems at the component level. The hypothesis was that students experiencing complex systems scenarios in a computer-based learning environment would outperform their counterparts by constructing a greater number of explanations with emergent-like responses. A mixed method experimental, pretest posttest, control group triangulation design research study was designed for high school students enrolled in an Introduction to Technology and Engineering course. A pretest consisting of one open-ended near transfer problem and one far transfer problem was administered, investigating the generation of reductive (clockwork) and complex (emergent-like) mental models. A stratified sampling procedure was used to assign students to control or treatment groups. Following treatment, an analysis of covariance failed to reveal statistically significant evidence supporting the hypothesis. However, qualitative data in the form of student transcriptions, daily lab reports, and data entry worksheets revealed evidence of emergent-like response and behaviors.