The majority of people show persistent poor performance in reasoning about “stock-flow problems” in the laboratory. An important example is the failure to understand the relationship between the “stock” of CO2 in the atmosphere, the “inflow” via anthropogenic CO2 emissions, and the “outflow” via natural CO2 absorption. This study addresses potential causes of reasoning failures in the CO2 accumulation problem and reports two experiments involving a simple re-framing of the task as managing an analogous financial budget. In Experiment 1 a (...) financial version of the task that required participants to think in terms of controlling debt demonstrated significant improvements compared to a standard CO2 accumulation problem. Experiment 2, in which participants were invited to think about managing savings, suggested that this improvement was fortuitous and coincidental rather than due to a fundamental change in understanding the stock-flow relationships. The role of graphical information in aiding or abetting stock-flow reasoning was also explored in both experiments, with the results suggesting that graphs do not always assist understanding. The potential for leveraging the kind of reasoning exhibited in such tasks in an effort to change people's willingness to reduce CO2 emissions is briefly discussed. (shrink)

Stocks and flows are building blocks of dynamic systems: Stocks change through inflows and outflows, such as our bank balance changing with withdrawals and deposits, or atmospheric CO2 with absorptions and emissions. However, people make systematic errors when trying to infer the behavior of dynamic systems, termed SF failure, whose cognitive explanations are yet unknown. We argue that SF failure appears when people focus on specific system elements, rather than on the system structure and gestalt. Using a standard SF task, (...) SF failure decreased by a global as opposed to local task format; individual global as opposed to local processing styles; and global as opposed to local perceptual priming. These results converge toward local processing as an explanation for SF failure. We discuss theoretical and practical implications on the connections between the scope of attention and understanding of dynamic systems. (shrink)