Park, D. C., Hertzog, C., Kidder, D. P.,
Morrell, R. W., & Mayhorn, C. B. (1997) Effect of
age on event-based and time-based prospective memory. Psychology and Aging,
12, 314-327.
The two experiments were designed to examine the effect
of age on both event- and time-based prospective memory tasks. Because the
authors used the model of a dual-task procedure, they included two additional
control conditions as well as using a prospective memory task which was not
related to the primary working memory task. The working memory task was a very
demanding verbal task while the prospective task was a non-verbal task using a
perceptual cue. It was predicted that having little relationship between the
two tasks would lead to an age difference on theevent-based
prospective task. The authors also varied the density of events to which the
subjects had to respond.
Experiment 1
The 48 younger and 48 older adult subjects were assigned
to the cells of a 2 x 3, between-subject design, resulting in 16 subjects in
each cell. The variables were age (younger, older), and density of prospective
memory events (none, 6, 12). There were also 12 older subjects who performed
only the 6 event prospective task and 12 who performed only the 12-event task.
Phase was a within-subject variable, and in the first phase subjects performed
only the working memory task. In the second phase all subjects performed both
the working memory and prospective memory tasks.
For the working memory task subjects were presented a
series of high-frequency words at the rate of 3-s each, on one of 6
backgrounds. and they were to retain the most recent 3
words in their memory. The subjects were asked to recall those words
periodically during the 12-minute task. In the second phase, subjects were also
required to monitor the background, and when the target background appeared, the were to press the zero key.
Results
Analysis of working memory task indicated no effect of
density (0, 6, 12) of the prospective-memory target.
There was an age effect, with younger recalling more words (.72) than older
(.52), and an effect of phase, indicating an improvement due to practice. The
age by phase interaction was significant, with the younger subjects showing
more improvement than the older.
Analysis of the prospective memory task (Table 2) yielded
only an effect of age; younger subjects got more correct than older subjects.
There were no significant correlations of the individual difference variables
with the prospective memory task, although there were some significant
correlations in the older age group between the individual difference variables and the working memory task.
There appeared to be no trade off between performance on
the working memory and the prospective memory tasks for either age group.
Experiment 2
This experiment used essentially the same materials and
procedure as in experiment 1, except that the
prospective task was changed to a time-based task by having subjects pull a
black lever every 1 or 2 minutes. They also could view a clock which showed how
much time had elapsed from the onset of the experimental task by pushing a red
button located near the lever.
Of the 56 younger and 56 older subjects in the
experiment, 64 were assigned to the conditions resulting from the 2 x 2 (age by
density), between-subject design. The remaining 48 were assigned to four
prospective-memory only control groups resulting from crossing age and density.
Data from the 32 subjects who performed only the working memory task were also
used in this experiment.
Results
To analyze the accuracy of the prospective memory
responses, the authors used various time windows, and reported on those of 3, 7
and 19 s. These windows ranged from stringent to lax in how close the
prospective memory response came to when it was required. Analysis of absolute
deviations from target gave similar results. Subjects did better on the
prospective task when they were not doing the working memory task; there was an
Age x Load (phase) interaction (Figure 2). There was no change in the percent
correct for younger subjects when the prospective task was added, but there was
a decrease for the older adults. There was also a main effect of density with
poorer performance (.54) on the 6 interval-task than on the 12-interval task
(.72).
Examination of the number of clock checks per quartile in
the one minute prior to the time the response was due yielded main effects of
age, working memory condition and quartile, as well as 4, 2-way, and a 3-way
interaction. The triple interaction is presented in Figure 3. Results indicated
that younger subjects showed an increase in clock-checking behavior in the
final quartile, while there was little increase for the older subjects.
Comparisons of the working memory task across experiments (event
vs. time) indicated that in Phase 1, when there was no prospective memory task,
performance was equivalent, .56 and .59 for the event and time based tasks
respectively. However, when the prospective task was added n phase 2, there was
a difference; .65 for the event-based task and .74 for the time-based task.