Abstract
This study investigated the performance on a computer game called Mastermind, playing of which consumes a large amount of working memory resource. First, 143 college students took a Japanese version of reading span test (RST). They then played Mastermind under two conditions: with or without display of past game records. There were 105 participants who relied under both conditions on a tactical strategy that caused high memory load, and their performance data were analyzed. Results indicated that low-span readers (87 students) were more constrained than high- and medium-span readers (18 students) under the condition without display, and performance of the low was affected by the content of display. It was argued that working memory resource that RST measured was involved in the process of cognitive problem solving.
