Academic journal article Memory & Cognition

Traveling Economically through Memory Space: Characterizing Output Order in Memory for Serial Order

Academic journal article Memory & Cognition

Traveling Economically through Memory Space: Characterizing Output Order in Memory for Serial Order

Article excerpt

How do people report the contents of short-term memory when information about order must be retained but items can be retrieved in any order? We report an experiment using an unconstrained reconstruction task in which people can report list items in any order but must place them in their correct serial positions. We found (1) a tendency to report recent items first in immediate but not in delayed reconstruction, (2) a tendency to recall temporally isolated items first, (3) a preference for forward recall order, and (4) a preference for output orders that minimize the length of the path that must be traversed through memory space during retrieval. The results constrain most current models of short-term memory in which retrieval is ballistic and is assumed to run to completion autonomously once initiated.

There has been much research emphasis on the order in which people retrieve items from long-term memory during free recall (see, e.g., Howard & Kahana, 2002; Kahana, 1996). This research has uncovered a number of empirical regularities: First, people tend to commence free recall with the report of one of the last list items (e.g., Howard & Kahana, 2002); second, once an item has been recalled, people's next report tends to be in a forward direction-that is, involving a later list item, preferably from nearby positions (i.e., short lags; see, e.g., Kahana, 1996). This information about output order has been crucial in shaping theories of free recall (e.g., Davelaar, Goshen-Gottstein, Ashkenazi, Haarmann, & Usher, 2005; Howard & Kahana, 2002), and its examination continues to provide new constraints on theories (Farrell & Lewandowsky, 2008).1

In short-term memory, by contrast, virtually nothing is known about the variables that determine people's report order, and no existing quantitative theories consider output order worthy of explanation. At first glance, this may not be entirely surprising, because most tests of short-term memory involve a mandatory report order. For example, in the classic immediate serial recall task, people must reproduce the list in the order of presentation. Nonetheless, several researchers have manipulated report order within a serial recall task-for example, by asking participants to commence recall with the second half of a list before reporting the first half (see, e.g., Beaman, 2002; Cowan, Saults, Elliott, & Moreno, 2002). This simple manipulation has turned out to be theoretically quite diagnostic. For example, the fact that recency increases at the expense of primacy when the last list items are reported first has been taken to suggest that primacy may, in part, result from output interference (Cowan et al., 2002). Similarly, taking a further step to disentangle presentation order from report order, Oberauer (2003) selectively probed for the recall of specific list positions in random order. The absence of recency when performance was plotted with respect to output position (thus canceling out the effect of input serial position) was interpreted as evidence against response suppression (see Brown, Preece, & Hulme, 2000; Lewandowsky, 1999). These precedents underscore the theoretical diagnosticity of manipulations or examinations of report order; however, existing precedents are limited in one important respect, because participants were unable to choose on their own the order in which to report items. This limitation, which turns out to have considerable theoretical import, can be overcome either by use of unconstrained recall (see, e.g., Tan & Ward, 2007)-a procedure in which people are free to report items in any order but must assign them to their list positions (unlike free recall)-or by use of an unconstrained reconstruction-oforder task (e.g., Lewandowsky, Nimmo, & Brown, 2008; Nairne, 1992). In the present article, we will focus on the latter.

In an unconstrained reconstruction task, participants are re-presented at test with the list items in random order and, rather than recalling them, must rearrange the items back into their order of presentation using an array of response options corresponding to the possible list positions. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.