Parallel effects of memory set activation and search on timing and working memory capacity

Abstract

Accurately estimating a time interval is required in everyday activities such as driving or cooking. Estimating time is relatively easy, provided a person attends to it. But a brief shift of attention to another task usually interferes with timing. Most processes carried out concurrently with timing interfere with it. Curiously, some do not. Literature on a few processes suggests a general proposition, the Timing and Complex-Span Hypothesis: A process interferes with concurrent timing if and only if process performance is related to complex span. Complex-span is the number of items correctly recalled in order, when each item presented for study is followed by a brief activity. Literature on task switching, visual search, memory search, word generation and mental time travel supports the hypothesis. Previous work found that another process, activation of a memory set in long term memory, is not related to complex-span. If the Timing and Complex-Span Hypothesis is true, activation should not interfere with concurrent timing in dual-task conditions. We tested such activation in single-task memory search task conditions and in dual-task conditions where memory search was executed with concurrent timing. In Experiment 1, activating a memory set increased reaction time, with no significant effect on time production. In Experiment 2, set size and memory set activation were manipulated. Activation and set size had a puzzling interaction for time productions, perhaps due to difficult conditions, leading us to use a related but easier task in Experiment 3. In Experiment 3 increasing set size lengthened time production, but memory activation had no significant effect. Results here and in previous literature on the whole support the Timing and Complex-Span Hypotheses. Results also support a sequential organization of activation and search of memory. This organization predicts activation and set size have additive effects on reaction time and multiplicative effects on percent correct, which was found.

Keywords: activation; additive factor method; complex memory span; memory search; retrieval; selective influence; time production; working memory capacity.

Figure 1

In the Dual Task Condition, a tone and probe are presented simultaneously. Time production and the memory task begin simultaneously (left side of figure). In the process organization proposed by Wickens et al. (1981), the first memory-task process is activation of the memory set. When it finishes, memory search begins. When the target time interval has elapsed and the memory task is finished, the response is made (right side of figure). In the Dual-Task Condition, time production is concurrent with activation and search. The time production process is not present in the Single-Task Condition.

Figure 2

A trial in the search phase. Single-Task Condition: The interval between the probe onset and the left/right key press response is the Reaction Time. Dual-Task Condition: The trial is identical except the participant must attend to the tone's duration: the presence or absence response is to be given only when the tone has reached the subjective target duration. The interval between the probe onset and the left/right key press response is the Time Production.