Memory in children

Kindergarten, third, and sixth graders received one of two 22
item lists for cued-recall, with one-half of the items in each
list being typical examples of familiar categories, and onehalf
heing atypical category exemplars. For subjects in the
Age-Appropriate condition, the typicality of the items was
based on children's definitions of "item goodness," whereas
for subjects in the Adult-Norm condition, item typicality was
based on adult judgements. At all grade levels, typical items
were recalled to a greater extent than atypical items, and
recall in the Age-Appropriate condition was significantly
greater than in the Adult-Norm condition. In the Age-Appropriate
condition, processing differences between typical and
atypical category exemplars were interpreted as being due to
qualitative differences in how representative items were of
their categories, whereas the "typicality" effects in the Adult
Norm condition were hypothesized as being due to a quantitative
lack of category knowledge.

Several studies have found recall and clustering performance of
young children to be greater with non-taxonomic (NT) than with
taxonomic (T) materials, while other studies have found the
reverse. The present experiment has tried to resolve this
discrepancy by introducing the variable of criterion vs single
sorting prior to recall. A comparison of Immediate and Delayed
recall between child-generated T and child-generated NT categories
under criterion (two consecutive identical sorts) and single
sorting conditions was used to assess the differences in these
T and NT grouping patterns as a basis for organizing recall.
Although there were no significant interactions with delay, when
subjects sorted only once, recall performance was greater with
T related materials. However, when subjects sorted to a stable
criterion of two consecutive identical sorts, recall performance
with NT related materials was greater than performance with T
related materials. These results suggest that under single
sorting conditions, the use of T categories may have resulted in
a better fit with the child's semantic memory structure than NT
groupings. However, with stable sorting, both T and NT grouping
patterns were equally consolidated into the memory structure,
making them both equally retrievable.

In the present experiment possible developmental differences in the
timing of organization in children's free recall were examined. It was
hypothesized that children who organize information at input would show
a smaller decrement in recall as a result of delayed testing with
related than with unrelated materials. If the categorical similarity
among items is discovered at input, the resulting organizational scheme(s)
would make the individual items more resistant to forgetting over time.
However, since no organizational scheme could be imposed upon unrelated
items, many items would be lost as a result of delayed testing. This
pattern of data was predicted only for junior high students and not
for second and third graders. However, contrary to expectation, both
the younger and older groups of subjects showed this pattern, indicating
that children of both age groups organize categorically related
information at input. The possibility of a semantic facilitative effect
for the younger subjects was discussed.

A probe recognition task tested memory for syntactic (active/passive arid
word order) changes and for semantic (meaningful and anomalous) changes.
On the basis of McNeill's theory of semantic development the following
predictions were made: (a) with a minimal retention interval (almost
immediate) 8-year-olds would recognize semantic changes better than
syntactic changes, while 6-year-olds would not perform differently on
the two types of changes, (b) with a longer retention interval, 8-year-olds
would recognize semantic changes better than syntactic changes. Results
supported (b) but in (a) the 8-year-olds recognized syntactic changes
better. This finding was discussed in terms of task differences and a new
experimental approach was proposed.

A stimulus class was generated by applying a fixed set of
transformational rules to a prototype, and selected members
of this class were presented during acquisition. It was found
that children discriminated members of this class from nonmembers.
For adults, further distinctions were obtained among
the stimuli within the class, as follows: Although the prototype
was not presented during acquisition, it was recognized
with greater confidence than stimuli from the acquisition set. Furthermore, class members that were not seen during acquisition
were recognized on the basis of this prototype. For those
class members that had been seen during acquisition, however,
there was no evidence that their recognition was based on the
prototype. Evidence that these previously seen class members
were recognized with greater confidence than new class members
indicated that recognition of stimuli from the acquisition
set was based on stored copies, or specific memory.

This series of experiments was designed to address two major questions concerning children's false-memory creation: (1) Are events that involve physical harm content more difficult to implant than events that do not involve physical harm content? and (2) Are individuals with particular cognitive and personality attributes more likely to create false-memories than others? In Experiment 1, 27 first-graders were presented with descriptions of two fictitious events (1 physical harm, 1 no physical harm) and two true events (1 physical harm, 1 no physical harm). Furthermore, individual differences in image descriptions, inhibition, verbal intelligence, working memory capacity and teacher ratings of cognitive and personality attributes were assessed. Results indicated that children generated more detailed accounts for fictitious events that did not involve physical harm in comparison to those that did. Furthermore, individual differences in image descriptions, verbal intelligence and teacher ratings of creativity, extroversion and risk taking were found. For Experiment 2, one-half of participants were presented with descriptions of two fictitious and two true events that involved physical harm content. The other one-half of participants were presented with descriptions of two fictitious and 2 true events that did not involve physical harm content. Furthermore, all children were told that two of the events (1 fictitious, 1 true) happened when they were very young and that the other two events (1 fictitious, 1 true) happened during the previous year. Individual differences in picture memory, inhibition and creativity were also assessed. The analyses failed to yield significant differences for event type or age of event occurrence. However, individual differences in creativity were related to children's false reports.