Judging event covariation: effects of age and memory demand

In this study, we investigated age differences in the accuracy of covariation judgement. Young and older adults were asked to solve covariation problems under low or high memory demand conditions. For each problem, subjects saw a sequential presentation of the event-state combination in a 2 X 2 contingency table. Subjects either kept a running tally of the frequencies of occurrence of each combination and used these tallies to make their covariation judgement for the events (Low Memory Demand), or they recalled the frequencies from memory, and then made their judgement (High Memory Demand). Solution patterns across the problems indicated which of four judgement strategies (i.e., Cell A, A vs B, Sum of Diagonals, or Conditional Probability) the subject preferred. The results showed that older adults were generally less accurate than young adults in judging event covariation. Additional findings suggested that this difference might be due to an age-related decline in memory for the frequency of event combinations and to older adults' use of simpler, less accurate judgement strategies.