This paper investigated the process by which decision-makers use corporate accounting and financial information to make decisions, and the performance outcomes of those decisions. By employing both the traditional input-output method and the process tracing method, which allows understanding of the decision-making process, this study simultaneously examined the decision-making process and performance to enhance understanding of graph usage and to provide guidelines for the increasingly prevalent use of computer-generated graphs. The results showed that decision-makers using graphs perceived the task as easier than those using tables, expended less cognitive effort, and exhibited a stronger tendency to use a holistic processing approach in information acquisition. Furthermore, the graph-using group employed compensatory information evaluation methods more frequently—which require more cognitive effort but can yield better outcomes—when task complexity was high. Regarding decision-making performance, the group using graphs demonstrated superior performance regardless of task complexity.