Twenty-four young (23 years) and 24 old (71 years) adults performed arithmetic tasks with working memory loads ranging from 1 to 4. Age groups were equivalent in mean accuracy and speed of arithmetic operations under minimal working memory load, but old adults were slower than young with memory demands .1. Access to a new object in working memory as the basis of computation required additional time. This object-switching cost increased with increases in memory demand, but was unaffected by age, indicating that old adults have no deficit in selective access to working memory.

In four of the five experiments presented by Bargh and colleagues, achievement-related words to prime the goal of doing well subsequently led to better achievement. However, as the experiments used verbs for achievement words and nouns as neutral words, the results might be due to confounding conditions. To avoid this problem, we conducted an online experiment with a 2 (achievement-priming: achievement words, neutral words) x 2 (word class: verbs, nouns) factorial design. 89 men and 123 women whose mean age was 25.8 yr. (SD = 7.1), who were recruited via a student e-mail list of the university and former participants in other experiments, took part. Barge, et al.'s findings that the achievement content is responsible for the priming effect were confirmed. Moderator effects of task difficulty and achievement motivation are discussed.