A fundamental issue in research on individual differences is the type of evidence sufficient to justify an inference of selective or distinct deficits in relevant theoretical processes. It is proposed that an important consideration is the extent to which the individual differences in 1 variable are independent of those in another variable. Specifically, the suggestion presented here is that a strong conclusion of selective impairment requires evidence that there is significant group-related variance in 1 variable after the variance in the other relevant variable is controlled. Furthermore, an inference that the groups are equivalent on a particular theoretical process requires evidence that the variable presumed to reflect that process has sufficient unique variance to justify the claim that a distinct process is being assessed. The proposed methods are illustrated with two studies comparing adults of different ages in mental arithmetic tasks.

In six experiments, subjects read narratives describing varying spatial scenes with more than one point of view. They were probed with questions about objects located in six directions from each character's point of view. Subjects' response times were consistent with a one place-one perspective rule. They seemed to form separate mental models for separate places and to take a character's perspective when there was only one relevant character in a scene, but they seemed to take a neutral perspective when there was more than one probed point of view, rather than switch perspectives.From discourse describing spatial situations, readers construct mental models of scenes in addition to mental representations of the language of the text (see, e.g., Bransford, Barclay, & Franks, 1972; Bryant, Tversky, & Franklin, 1992;Denis & Cocude, 1989;Franklin & Tversky, 1990; Glenberg, Meyer, & Lindem, 1987;Mani & Johnson-Laird, 1982;Morrow, Bower, & Greenspan, 1989;Morrow, Greenspan, & Bower, 1987;Perrig & Kintsch, 1985;Taylor & Tversky, 1992). Mental models contain information about the characters and objects in a scene, their orientations and locations. The spatial information preserved in them includes categorical spatial relations, such as those expressed in the words above, in front of, north oj, across from, and so on, and sometimes it includes more analog information about distances. These mental models appear to be rapidly updated and transformed as narratives supply new information about objects, locations, and orientations. Experiments on the comprehension of spatial descriptions reveal processes underlying spatial thinking as well as those underlying language comprehension.In many cases, readers' performance indicates that they have taken a specific perspective on a described scene

The basic question motivating the present research was whether there were age-related differences in the ability to perform two concurrent tasks above and beyond the ability to perform the tasks in isolation. One reason why this question is of interest stems from Baddeley's (e.g., 1986Baddeley's (e.g., , 1992 Baddeley, Logie, Bressi, Della Sala, & Spinnler, 1986) hypothesis that the ability to coordinate two concurrent tasks is a function ofa central executive in a working memory system, and that the efficiency or effectiveness of the central executive might decline with increased age. The plausibility of this proposal would obviously decrease iffew or no age differences were found in dual-task performance after taking single-task performance into account. Accordingly, a major purpose of the present study was to attempt to assess the effectiveness of the hypothesized executive independently of performance in the constituent tasks.Another reason why the influence of age on the relationship between single-and dual-task performance is of interest is that it is related to a larger issue concerning the extent to which age differences in complex cognitive activities are predictable from age differences found in the performance of elementary cognitive tasks. That is, an important consideration in research on aging and cognition is the level at which age-related influences are manifested. One possibility is that there are direct and independent age-related effects on both simple and complex cognitive activities, suggesting that higher order cognitive processes are more affected by age-related influences than are lower ordero or elementary, processes. An alternative possibility is that all of the age-related influ- Previous research has yielded conflicting results regarding the relationship between adult age and the ability to divide attention between two concurrent tasks. At least some of the inconsistency is probably attributable to methodological variations, such as the manner in which divided-attention ability has been assessed, how single-task performance has been considered, and the degree of control over relative emphasis placed on each task. TWo experiments employing procedures sensitive to these concerns were conducted in which a speeded decision task was performed during the retention interval of a letter-memory task. The results of both experiments indicated that there were relatively few age-related influences on dual-task performance vis-i-vis those on single-task performance.

Two studies were conducted to investigate the aspect(s) of processing involved in the hypothesized speed mediation of adult age differences in memory. Both studies involved a serial memory task in which information was to be recalled either in the original order of presentation, or in a reordered sequence. Results from both studies indicated that task-specific processing durations were slower among older adults than among young adults, but that the attenuation of the age-related variance in memory was nearly as great after statistical control of a task-independent speed measure as after control of task-specific speed measures. These findings are consistent with the hypothesis that a substantial proportion of the adult age-related differences in memory is associated with a decrease with increased age in the speed of executing many cognitive processes, and not simply the speed of one or two specific processes.

Participants in two studies conducted by Salthouse (in press) were called 2 to 182 days after participation and asked to describe the activities that they had performed in the previous study. Hierarchical regression analyses were used to examine the prediction of activity recall from age, speed, and retention interval. Overall, age was associated with 20% of the variance in activity recall, and retention interval was associated with 19%, but there was no significant interaction of age and retention interval. When perceptual speed was entered into the regression equation before age, the age-related variance was reduced by 70%. A small, but statistically significant, amount of age-related variance in activity memory remained after controlling for speed and retention interval.