Reliability of verbal–numerical working memory tasks

A large body of research has shown substantial correlations between working memory capacity and measures of intelligence. In contrast, studies focusing on test–retest reliability and other psychometric properties of working memory tasks have been less common and provided mixed results. The purpose of the present study therefore was to further explore the test–retest reliability and scalability of four verbal–numerical working memory tasks. The tasks were administered twice to N = 70 students with 4.5–6 months between testing attempts. Test–retest reliabilities of single tasks ranged from .69 to .77. Exploratory factor analyses revealed a single factor at the first testing attempt, but two method-specific factors at the second testing attempt, possibly indicating a task-related practice effect. Reliability analyses based on item response theory revealed that task reliability depended both on subjects’ working memory capacity and practice effects. Results are discussed in light of other research investigating psychometric properties of working memory tasks.