Shared Attention in Pigeons: Retrieval Failure Does Not Account for the Element Superiority Effect

Divided attention performance and the matching law

“…The finding that the bias parameter was greater than zero reflects proportionally higher accuracies on color trials than on line trials across the range of reinforcement ratios. Similarly higher accuracies on color than on line trials have been obtained in previous research using related procedures (Langley & Riley, 1993;Santi et al, 1982;Zentall et al, 1997).…”

“…Lamb (1991) suggested that the divided attention effects obtained with pigeons and compound samples may similarly be the result of retrieval failures, rather than capacity limitations at the time of initial processing. Zentall et al (1997), however, provided empirical data inconsistent with the notion that divided attention effects with pigeons reflect a retrieval failure at the choice point. Nonetheless, it is possible that the differential reinforcement in the present experiment could have its effects by affecting performance at the choice point.…”

“…Similarly, Maki and Leith (1973) found that when sample duration was fixed at the same duration for compound and element sample trials, accuracy was consistently higher following element samples (cf. Gottselig, Wasserman, & Young, 2001;Langley & Riley, 1993;Leith & Maki, 1975;Santi, Grossi, & Gibson, 1982;Zentall, Sherburne, & Zhang, 1997). Although other hypotheses have been proposed (e.g., generalization decrement, coding decrement, comparison uncertainty, retrieval deficit), the bulk of the data support the suggestion that this "element superiority effect" results from the detrimental effects of a division of attention between the elements with compound samples (see Zentall, 2005, andZentall &Riley, 2000, for reviews).…”

Divided attention and the matching law: Sample duration affects sensitivity to reinforcement allocation

“…In addition, overall accuracy on the divided-attention task was lower with a shorter sample duration than with a longer sample duration. This finding is consistent with a number of previous experiments in which sample duration has been manipulated within similar tasks (e.g., Maki & Leith, 1973;Santi et al, 1982;Zentall et al, 1997). Most important, sensitivity of accuracy to changes in relative reinforcement was greater with a longer sample than with a shorter sample.…”

“…Thus, the time available for processing the sample has been a major variable of concern in previous experiments in which the divided-attention task has been used (see Zentall & Riley, 2000, for a review). A reliable finding in this previous research is that decreases in sample duration decrease accuracy on the task (e.g., Maki & Leith, 1973;Santi, Grossi, & Gibson, 1982;Zentall, Sherburne, & Zhang, 1997). Here, we ask whether the impact of variations in relative reinforcement on divided-attention performance (i.e., a in Equation 1) depends on sample duration.…”