ra
{
these 'special' items to a some small degree. Size would be
a composite variable, presumably representive of cost,
durability, and social prestige as well. The broad post hoc
measure of size had a small but significant relationship to
both the recall listings, as well as to whe relationship of
the two listings to each other. Subjects tended to list
large items, more so for citings not owned. The
relationships of item size to recall listings remaind even
with items in territorial categories were removed. About
11% of the relationship between the two recall listings was
due to the effect of size, and this decreased to 6.5% with
the territorial items deleted. Finally, the size measure
was predictive of recall order, with larger items being
listed earlier in the recall listings of both the owned
things and the things not owned. With territorial items
deleted, the relationship of size to recall position
decreased for both recall listings. Thus, there was a
slight tendency to list territorial items earlier in both
recall tasks, and subjects listed larger items earlier in
both tasks. About 9% of the shared variance in the two
-call tasks was due to the tendency to list larger items.
Thus, a preference for listing Larger items accounted for
about half of the common qualities in the two recall tasks.
Recall Strategies
The common qualities of the two recall lists can also .