Downloads & Free Reading Options - Results

We are interested to research how memory abilities contribute to numerical estimation tasks. Recall is an ability that is essential to many of our daily tasks, of which most require memory of some kind. A common finding when researching free recall of sequentially presented items are effects of recency and primacy, i.e. easier recall for the first and last items presented in a sequence (e.g. Murdock, 1962). Primacy and recency effects have also been shown to influence human decision making. For instance, candidates who perform on the first or last position are typically evaluated more favourably. It is, however, unclear how much primacy and recency effects in decision making can be explained by memory recall. In two previous experiments, we investigated how participants estimated the average of sequences of only numbers and found that the serial position weight of an item in this averaging task correlated higher with the recall probability of an item on this position in the free recall task when participants did not know which task lie ahead (randomly interpersed; r=0.6, Exp. 1) than when they did know which task lies ahead (blocked such that the first half of sequences were averaging tasks, and the second half were free recall tasks; r=0.38, Exp. 2). These correlations were mainly due to recency effects. Differences arising from such knowing or not knowing were investigated in recall tasks by Oberauer (2003) who called these procedures "pre-cueing" and "post-cueing", respectively. For a third experiment, we aim to generalize our findings to a cued recall task in which participants are presented not solely by a number, but a number associated to an object. As objects, we use shopping carts filled with six random products, each from a different product category. We aim to replicate our findings of numerical averaging from our first two experiments. Further, we try to get a more general understanding on contribution of memory to numerical estimation and thus include a further task in our experiment: summation. Interestingly, Scheibehenne (2019) found that participants especially rely on the middle items of a sequence when summing numbers up. This would suggest an opposite contribution of memory to this task. More likely, however, participants use different strategies when averaging and summing (possibly even adjusting for a bias in recency and primacy).