See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/11985900

Working memory and updating processes in reading comprehension

Article in Memory & Cognition · April 2001

DOI: 10.3758/BF03194929 · Source: PubMed

CITATIONS READS

282 1,360

4 authors:

Paola Palladino Cesare Cornoldi

Università degli studi di Foggia 341 PUBLICATIONS 12,766 CITATIONS
87 PUBLICATIONS 2,434 CITATIONS
SEE PROFILE
SEE PROFILE

Rossana De Beni Francesca Pazzaglia

University of Padova University of Padova
163 PUBLICATIONS 7,545 CITATIONS 128 PUBLICATIONS 4,453 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Paola Palladino on 08 January 2015.

The user has requested enhancement of the downloaded file.

Memory & Cognition
2001, 29 (2), 344-354

Working memory and updating

processes in reading comprehension
PAOLA PALLADINO, CESARE CORNOLDI,
ROSSANA DE BENI, and FRANCESCA PAZZAGLIA
University of Padua, Padua, Italy

In this study, we examine the relation between reading comprehension ability and success in work-
ing memory updating tasks. Groups of poor and good comprehenders, matched for logical reasoning
ability, but different in reading comprehension ability, were administered various updating tasks in a
series of experiments. In the first experiment, the participants were presented with lists of words, the
length of which (4–10 words) was unknown beforehand, and were required to remember the last 4
words in each series. In this task, we found a decrease in performance that was related to longer series
and poor reading ability. In the second experiment, we presented lists of nouns referring to items of
different sizes, in a task that simulated the selection and updating of relevant information that occurs
in the on-line comprehension process. The participants were required to remember a limited, prede-
fined number of the smallest items presented. We found that poor comprehenders not only had a poorer
memory, but also made a greater number of intrusion errors. In the third and fourth experiments, mem-
ory load (number of items to be selected) and suppression request (number of potentially relevant
items) were manipulated within subjects. Increases in both memory load and suppression requests im-
paired performance. Furthermore, we found that poor comprehenders produced a greater number of
intrusion errors, particularly when the suppression request was increased. Finally, in a fifth experi-
ment, a request to specify the size of presented items was introduced. Poor comprehenders were able
to select the appropriate items, although their recall was poorer. Altogether, the data show that work-
ing memory abilities, based on selecting and updating relevant information and avoiding intrusion er-
rors, are related to reading comprehension.

Several studies have considered the role of working can become too broad or even irrelevant. A good compre-
memory—that is, the temporary system that maintains hender must, therefore, update the available information
and processes information (Baddeley, 1986)—in reading contained in working memory.
comprehension. Indeed, it has been shown how the reader The aim of the present work was to study the relation
must keep in mind information that is useful for inter- between the ability to update information contained in
preting successive parts of the text. For example, in order working memory and reading comprehension ability. In
to solve an anaphora, a reader needs to recover the an- the literature, this aspect has not been considered, even
tecedent of a pronoun (Clark & Sengul, 1979; Daneman though particular attention has been paid to the updating
& Carpenter, 1980; Ehrlich & Rayner, 1983). Given the processes of mental representations maintained during
limits to working memory capacity, good use of it does comprehension. We can distinguish between generally
not consist in simply maintaining as much information as automatic updating in relation to what has been read be-
possible, but in continuously selecting and updating this forehand (automatically encoded information), which
information. A good comprehender is capable of choosing guarantees local coherence, and a somewhat more con-
the most relevant information in the context which he or trolled updating that involves conscious, strategic pro-
she is analyzing (Brown, Armbruster, & Baker, 1986; Gar- cesses. These processes occur, for example, when it be-
ner, 1987). He or she is also able to utilize that particular comes necessary to interpret a portion of the text when the
information at the right moment in order to interpret a cer- previous representation is inadequate (see, e.g., McKoon
tain part of the text. On the other hand, the amount of in- & Ratcliff, 1992; van Dijk & Kintsch, 1983). Various re-
formation kept available in the working memory system searchers (e.g., de Vega, 1995; Glenberg & Langston,
1992; Morrow, Bower, & Greenspan, 1989) have observed
that the participant updates the representation of the text
The present research was partially supported by CNR grants and in order to understand the new information supplied by
MURST–40% grants to the second and third authors. We thank Chiara the text. In particular, Robertson and Gernsbacher (1996)
Braga, Barbara Carretti, Martina Di Pieri, Anna Maffezzoni, and Maria found that the reader keeps track of the character being
Carmela Spoto for their help in data collection. Correspondence con-
cerning this article should be addressed to C. Cornoldi, Dipartimento di
discussed in a text and updates the reference when a new
Psicologia Generale, Università di Padova, via Venezia 8, 35131 character is introduced. The capacity to recall the previous
Padova, Italy (e-mail: [email protected]). character’s name is therefore reduced. The authors have

Copyright 2001 Psychonomic Society, Inc. 344

 WORKING MEMORY AND UPDATING IN READING COMPREHENSION 345

interpreted these data, using the structure building frame- ing memory updating test, when an individual is invited
work (Gernsbacher, 1990), as proof of an updating pro- to remember, he or she must give the information that is
cess (a shift to building a new substructure) whereby new relevant at that time. In order to do this, he or she must
information leads to the suppression of previous infor- have updated the relevance and, therefore, the degree of
mation. accessibility of information processed, thereby reducing
In this study, we examine the relation between working the accessibility of information that is (certainly or prob-
memory updating and the reading comprehension process ably) no longer useful. Success in the task is linked to re-
from the point of view of individual differences. membering the pertinent information and also to control-
The present research used the methodology of compar- ling for nontarget information with respect to items that
ing matched groups to see whether they differed in their were potentially relevant but did not satisfy the selection
capacity to update working memory. With respect to the criterion. This information, if not updated, can disturb the
well-known memory tests (e.g., Daneman & Carpenter, memory of pertinent information but can also be erro-
1980), a working memory updating test has the advantage neously remembered, thus producing an intrusion error. It
of measuring both aspects of working memory implicated has recently been observed (De Beni, Palladino, Pazza-
in text comprehension—not only the maintenance of se- glia, & Cornoldi, 1998) that intrusion errors in working
lected information, but also its updating. memory tasks are highly related not only to success in the
At first, we considered an updating ability test, devel- working memory tasks themselves, but also to reading
oped by Morris and Jones (1990), that follows a proce- comprehension ability.
dure presented by Pollack, Johnson, and Knaft (1959).
The test requires the participant to listen to a sequence of EXPERIMENTS 1–5
letters, varying in length (4–10), and afterward to remem-
ber only the last 4 elements. Since the length of the series Method for the Selection of Participants
is unknown, it is assumed that each element is initially For all five experiments, we selected a group of good compre-
held in memory and can only be updated as the presenta- henders and a group of poor comprehenders according to the fol-
tion continues. Morris and Jones consider this a measure lowing criteria. In a reading comprehension test, good comprehen-
of the central executive component of the working mem- ders were required to have a performance at least 1 SD higher than
the average score expected for their age, and the group of poor com-
ory, which is also hypothesized to be related to reading prehenders a performance at least 1 SD lower. The reading com-
comprehension ability (Baddeley, 1990). prehension test was based on the MT battery (Cornoldi, Rizzo, &
In five experiments, the present research tested five dif- Pra Baldi, 1991), the only standardized test for assessing reading
ferent groups of people with low text comprehension abil- comprehension available in Italy. The test includes passages fol-
ity, as compared with participants matched for age, sex, lowed by multiple choice questions. Groups were matched for age
education, and logical ability but with good comprehen- and general logical intelligence, which was examined with a battery
of logical reasoning subtests based on items (numbers and spatial pat-
sion ability. In the first experiment, the participants were terns) present in standard group intelligence tests (Anastasi, 1988).
administered the previously mentioned updating task used
by Morris and Jones (1990), in this case replacing the let-
EXPERIMENT 1
ters with words. In the second experiment, the memory
updating task was made more similar to the updating task
in the comprehension process. It required participants to The main goal of Experiment 1 was to test the hypoth-
select information, with reference not to a simple ordinal esis that poor comprehenders’ performance is lower than
position criterion (such as the last four words), but to a that of good comprehenders in a working memory updat-
predetermined semantic criterion (the words that mainly ing task. The participants were required to update not just
satisfied the criterion; in our case, the words referred to letters, as in the original updating task (Morris & Jones,
the smallest items). In the third experiment, we wanted to 1990), but more meaningful and semantically relevant
generalize the results to a sample of younger participants items—that is, concrete, highly familiar nouns. Further-
and to evaluate the role of required memory load and more, they were tested with a word span test in order to ex-
quantity of potentially relevant information. In the fourth clude the possibility that failure in the updating test was
experiment, we analyzed the typology of errors made in due to poor short-term memory.
the updating task. In the fifth experiment, we controlled
for the possibility that poor comprehenders’ difficulty in Method
Participants. Sixteen poor comprehenders (4 males and 12 fe-
working memory updating tasks was not due simply to a males), between 19 and 21 years of age, all undergraduate students
failure to identify the items that had to be remembered. at the University of Padua, and 14 good comprehenders (4 males
In summary, the aim of the five experiments was to test and 10 females), with similar education, age, short-term memory,
the hypothesis that comprehension ability is related to and logical intelligence, agreed to participate in the experiment.
the ability to update relevant information maintained in Short-term memory was measured by a forward word span test.
working memory. This hypothesis is the consequence of Procedures were drawn from the Wechsler Intelligence Scale–
Revised (Wechsler, 1987), standardized for the Italian population
a model that assumes that, among the operations required (Orsini & Laicardi, 1997). The test, which was administered indi-
in the comprehension process, there are operations that vidually, assesses serial recall of strings of words that are presented
involve the selection and use of information that is rele- orally. Different from the original procedure (Wechsler, 1987), the
vant to interpreting the passage. Similarly, in the work- participants were presented with only one list for each length. The
346 PALLADINO, CORNOLDI, DE BENI, AND PAZZAGLIA

Table 1 they may present a deficit in working memory tasks in-

Means and Standard Deviations of Number of Words
Recalled in the Updating Task in Experiment 1 volving central control (see Daneman & Merikle, 1996;
Words Recalled
De Beni et al., 1998). In other words, the control functions
of working memory seem related to text comprehension
Number of Words M SD
ability. This relation was extended to a working memory
4 (no updates) 14.90 1.95
6 (two updates) 13.2 2.46
test that was not originally devised for examining the re-
8 (four updates) 12.63 2.14 lation between working memory and comprehension. The
10 (six updates) 11.47 2.11 updating test we used is considered a measure of the cen-
tral executive component of working memory (Morris &
Jones, 1990; Van der Linden, Bredart, & Beerten, 1994).
The fact that poor comprehenders had a lower performance
score was based on the greatest list length a participant was able to than good comprehenders in this task, but not in a standard
recall completely in the correct order (span). The mean span was
4.94 (SD = 1.44) for poor comprehenders and 5.14 (SD = 1.10) for
span task, confirms that, within Baddeley’s (1990) tripar-
good comprehenders, with a slight nonsignificant difference be- tite working memory model, the comprehension process
tween the groups (t = 0.043, p = .67). engages the central executive more than it does the artic-
Materials and Procedure. Sixteen lists of words were pre- ulatory loop. It should be noted that we found a significant
sented. The words were bisyllabic, highly familiar nouns taken effect of list length, but no interaction between groups
from the Cornoldi (1974) repertoire. The 16 lists were divided into and list length. These data demonstrate that an increase in
four groups, varying by number of words (4, 6, 8, or 10). An exam-
list length produces a more complex request, thus impair-
ple of an 8-word list was the following: forno (oven), latte (milk),
duomo (cathedral), fiato (breath), vasca (bath), zucca (pumpkin), ing performance. However, although this aspect is impor-
succo (juice), cassa (box), and baffo (mustache). The lists were pre- tant in principle, it is not necessarily crucial to reading
sented orally, in a fixed order, using a tape recorder and balancing comprehension ability. The two reading comprehension
the presentation order of lists of different lengths. Within each list, ability groups were similarly affected by an increase in
the words were presented at the rate of 1 word per second. The end list length. In particular, it should be noted that poor com-
of the list was signaled by a beep, and the participant was immedi- prehenders also performed worse with lists of four items
ately required to write the last 4 words in the list. If he or she could
not recall a word, it had to be omitted (in contrast to Morris & in which no substitution–updating was apparently re-
Jones’s [1990] procedure, the participants were not required to re- quired. This result suggests that memory updating does
member four items). The participants were tested in small groups not involve only simple maintenance and substitution, but
of 3 or 4. The test was preceded by a warm-up session with two lists probably more general control and coordination of infor-
of each length. The experiment took approximately 15 min. mation activation/suppression in working memory.
The results also offer an insight into the role of updat-
Results ing ability in reading comprehension. In fact, both the up-
Table 1 presents the mean numbers of correctly repro- dating and the reading comprehension tasks require the
duced words for the different length lists. A word was participant to continuouslyupdate information and, when
considered correctly reproduced only if it was recalled in required, find appropriate information. This requirement
the correct position. When a list was only partially re- cannot be one-way only but, rather, takes place through a
called, a recalled word was considered ordered if it was flexible and strategic use of working memory. In fact, only
not presented before the word just recalled. A 2 3 4 a few participants reported that they were simply main-
analysis of variance (ANOVA) for a mixed design re- taining the last four words, continuously substituting the
vealed significant effects of group [F(1,28) = 4.98, first ones. As the presentation rate was very rapid and the
MS e = 7.86, p = .034; good comprehenders, M = 54.64, quantity of material was unknown and could have been
SD = 5.8; poor comprehenders: M = 50.06, SD = 5.43] small, many participants reported that they had partially
and of list length [F(3,84) = 18.27, MSe = 3.35, p , .001]. selected and updated the material during the presentation
We also computed the errors, but the two groups did not and partially completed the updating operation when re-
differ significantly in the mean number of either intru- quired to recall the words.
sion errors (M = 3.31, SD = 1.70 for the poor compre-
henders; M = 2.64, SD = 2.56, for the good comprehen- EXPERIMENT 2
ders; p = .40) or order errors (M = 1.50, SD = 1.41, for
the poor comprehenders vs. M = 1.00, SD = 1.36, for the The working memory updating task used in Experi-
good comprehenders; p = .33). Overall, the number of re- ment 1 had the advantage of requiring a continuous pro-
sponses of both good and poor comprehenders (58.28 and cess of refreshing relevant information with semantically
54.87, respectively) was below the maximum possible significant items, but it had the disadvantage of simulat-
value of 64. ing rather superficially the updating processes involved
in real comprehension tasks. In fact, reading comprehen-
Discussion sion requires the reader to select and update on the basis
The results of Experiment 1 confirm that even when of the semantic relevance of information, not just on the
poor comprehenders are matched with good comprehen- basis of information position. For this reason, we devel-
ders in a series of abilities, including short-term memory, oped a task in which updating had to be made on the basis
 WORKING MEMORY AND UPDATING IN READING COMPREHENSION 347

of a semantic criterion (item size), rather than simply by Table 2

using a presentation order criterion. In this version of the Means and Standard Deviations for Poor
and Good Comprehenders in Indices of
task, the length of the list was always the same, but the Updating and Final Recall Tasks (Experiment 2)
number of relevant elements varied within the list, as did
Poor Good
the number of elements that had to be remembered. We Comprehenders Comprehenders
also introduced a final recall test in order to test the hy- Task M SD M SD
pothesis that the memory deficit of poor comprehenders
Updating
is not general, but specifically involves target items. Words recalled 80.16 6.28 85.95 4.79
Intrusions 10.67 5.49 6.35 4.19
Method Unexpected final recall
Participants. Eighteen poor comprehenders and 20 good com- Target words 26.05 5.52 30.30 6.57
prehenders, with characteristics similar to those of the participant Relevant items 14.55 5.04 16.75 6.46
groups in Experiment 1, participated in this experiment. Filler items 3.39 3.07 2.50 2.15
Materials and Procedure. Twenty-four lists of 12 words were
presented. Each list contained familiar nouns that referred to either
objects/animals measurable by size (relevant information; in a num- items (30.76%), but much greater than the percentage of
ber varying from 5 to 10) or abstract entities (filler items). Nouns filler items [2.73%; F(2,72) = 261.33, MSe = 28.66, p ,
of sizable objects/animals were selected in a pilot study. In the pilot, .001].
five judges rated the discriminability of the item sizes in the lists.
From the ratings, it appeared that the sizes of the items in each list
were immediately evident and clearly different from each other. Discussion
The instructions emphasized that the participant would be pre- The present experiment confirms that poor compre-
sented with lists including filler abstract items and concrete items, henders have difficulty with verbal working memory tests
the size of which had to be considered. Only for concrete items did and with updating tasks in particular. In this case, in con-
he or she have to select the smallest. For half the lists, the partici- trast to Experiment 1, updating had to be done on the basis
pants had to remember the three smallest items, and for the other
half, the five smallest. The words were presented orally using a tape
of a semantic criterion. The updating process therefore re-
recorder, at the rate of one word per second. Each participant was flected the updating operations more directly involved in
tested individually and gave a verbal response after each list. An the comprehension, since text comprehension requires
example of a list was incontro (meeting), senso (sense), picchio that relevant information is extracted and made highly ac-
(woodpecker), passione (passion), legge (law), mucca (cow), felicità cessible in working memory. The pool of accessible in-
(happiness), quantità (amount), bruco (caterpillar), agnello (lamb), formation is then updated in favor of new, more relevant,
festa (feast), and rana ( frog), and the three words to be remembered
incoming information. It must be noted that poor com-
were woodpecker , caterpillar , and frog. At the end of the experi-
ment, the participant was invited unexpectedly to remember all the prehenders recalled a lower number of target items in both
words which had been presented. The recall test was written, and the immediate and the final tests, suggesting that appro-
the participant had 7 min to complete the test. priate selection and updating of information had long-
term effects.
Results In the present case, we also found a difference between
Table 2 presents the mean numbers of words recalled groups in the mean number of intrusion errors. Poor com-
correctly for the two groups, their overall intrusion errors prehenders made more intrusions when the intrusive in-
(nouns presented in the list, but not the smallest ones), formation initially had to be processed deeply—that is,
and the mean numbers of words that were recalled in the for concrete (sizable) nouns. The results of this experi-
final test divided by category. The poor comprehenders ment replicate De Beni et al.’s (1998) observations, which
gave a significantly lower performance in the updating showed that poor comprehenders make more intrusion er-
task, recalling 80.16 words (out of a maximum of 96), rors than good comprehenders in listening span tests and
versus a mean recall of 85.95 for good comprehenders, that these errors are more frequent for more extensively
a highly significant difference [t (36) = 3.21, p = .003]. processed information. However, differences in listening
The difference in intrusion errors was significant, being and updating task procedures do not allow us to attribute
nearly double for the poor comprehenders [t (36) = 2.74, the effects to exactly the same factors. Those participants
p , .01]. interviewed after the experiment reported that they main-
Regarding final recall (Table 2, bottom), the only sig- tained many of the relevant items activated to a different
nificant difference between the groups concerned the extent. Intrusion errors could have been due to memory
target words [t(36) = 2.14, p = .039], a result that substan- noise, wrong levels of activation, inappropriate selection
tially reflects the good comprehenders’ better working when the response was required, or other factors. In par-
memory recall of target items. In the other final recall ticular, the participants’ difficulty could have been due to
measures, the two groups gave a similar performance, an increase either in the number of relevant items (lead-
which suggests that in other respects, the memory of the ing to an increase in the request to suppress nontarget
two participant groups was similar. Overall, the percent- items) or in the memory request (leading to an increase in
age of sizable, but nontarget, items (i.e., the larger items) the number of items to be recalled). A further experiment
recalled (18.63%) was lower than the percentage of target tested between these possible explanations more directly.
348 PALLADINO, CORNOLDI, DE BENI, AND PAZZAGLIA

EXPERIMENT 3 practice, modality of item presentation, and the items themselves

were the same as those in Experiment 2. The participants were in-
In Experiment 3, we examined the consequences of dividually tested in a quiet room in their school.
different memory and suppression requests in the updat-
ing task administered in Experiment 2. The task was pre- Results
sented to new groups of good and poor comprehenders. Figure 1 presents the mean percentages of words from
Furthermore, the larger size of the initial available sam- the four categories of the lists that were remembered by
ple made it possible to select a larger group of poor com- the two groups (percentages, rather than absolute values,
prehenders than in the other experiments. In this experi- were considered because the sets of items to be recalled
ment, we tested the hypothesis that updating is affected by varied according to the category of lists). A 2 3 2 3 2
increases in suppression and memory requests. The sup- mixed design ANOVA (group 3 suppression request 3
pression request was considered higher when the partic- memory request) revealed a significant group effect
ipant had to extract target information from a larger pool [F(1,58) = 21.32, MSe = 430.63, p , .001], a significant
of relevant information. Consequently, more information effect of suppression request [F(1,58) = 20.48, MSe =
must be activated, examined, and temporarily maintained 91.64, p , .001], and a significant effect of memory re-
in memory, and in turn, more information must be sup- quest [F(1,58) = 343.72, MSe = 73.23, p , .001]. Poor
pressed. The memory request was considered higher when comprehenders gave a poorer performance, and increased
the participant had to remember a larger number of tar- suppression and memory requests impaired recall.
get items. Consequently, he or she had to hold in memory Figure 2 presents the mean numbers of intrusion errors
and/or extract a greater number of elements, with an in- for the two groups divided according to list categories. A
creasing amount of activity required. Once again, we also 2 3 2 3 2 ANOVA revealed that all the main effects
examined the effects of memory and suppression requests were significant [groups, F(1,58) = 6.50, MSe = 9.54, p =
in a working memory updating task (performance and er- .013; suppression request, F(1,58) = 15.15, MSe = 2.54,
rors) for two groups that differed in reading comprehen- p , .001; memory request, F(1,58) = 15.76, MSe = 1.78,
sion ability. We hypothesized that higher memory and p , .001]. Furthermore, the group 3 suppression request
suppression requests would tend to overload the working interaction was significant [F(1,58) = 5.15, MS e = 2.54,
memory system of both groups, impairing performance. p = .027]. From Figure 2, it can be seen that increases in
Furthermore, if poor comprehenders have difficulty main- suppression and memory request produced a higher num-
taining relevant information and rejecting what has be- ber of intrusions and that the interaction was due to the
come irrelevant, they should also have a poorer perfor- fact that poor comprehenders were particularly disturbed
mance and present more intrusion errors with a higher by the high-suppression lists. A closer look at the number
suppression request. of intrusions produced by the poor comprehender group
revealed a high level of variability. This could be due to
Method the fact that poor comprehenders used a greater variety of
Participants. Thirty poor and 30 good comprehenders, between strategies than did good comprehenders.
11 and 15 years of age, attending a junior high school in Lodi, a The number of recalled filler items was particularly low
town in Northern Italy, participated in this experiment.
Materials and Procedure. As in the preceding experiment, in the final recall test, so we only considered the number
24 lists of 12 words were presented. The words in the lists were of relevant items recalled, divided according to whether
mostly the same as those used in the preceding experiment; the new they were a target or had to be suppressed. A 2 3 2
words had similar characteristics. Relevant information was made ANOVA revealed that good comprehenders remembered
up of animal and object nouns distributed in the different lists, bal- a significantly higher number of relevant nouns than did
anced within-subjects. The 24 lists were divided into four cate- poor comprehenders [F(1,58) = 12.91, MSe = 25.70, p ,
gories, according to the number of relevant items (high and low
suppression required) and the number of items to be recalled (high
.001] and that target items were remembered better than
and low memory request in loading and retrieval operations). In the nontarget items [F(1,58) = 33.51, MSe = 21.00, p , .001].
low-suppression lists there were two irrelevant items, whereas there The absence of an interaction was due to the fact that good
were five in the high-suppression lists. The participant had to re- comprehenders remembered not only the target items
member three items in the case of low memory load and five items better (M = 18.92, SD = 6.37, vs. M = 15.21, SD = 4.93),
for high memory load. Lists for each category were presented in a but also the relevant nontarget items (M = 13.69, SD =
block. The following is a list with high suppression and memory re-
4.17, vs. M = 10.75, SD = 3.34).
quests: penna (pen), relazione (relationship), portafogli (wallet),
culla (cot), scala (ladder), occhiali (glasses), dolore (pain), quader-
none (exercise book), gomma (rubber), treno (train), sedia (chair), Discussion
and violino (violin). The participants were required to remember The results of Experiment 3, on the whole, replicated
the five smallest items: pen, wallet, glasses, exercise book, and rub- the results of Experiment 2 with a different group of par-
ber. The order of the blocks of lists was balanced between subjects. ticipants. It appears that poor comprehenders have diffi-
There were four practice trials for each category. In order to avoid
recency effects, at the end of the working memory test, the partici-
culty in the updating task, producing both a lower recall
pants were asked to do an interpolated task, consisting of counting and a greater number of intrusions. Furthermore, success
backward by two from 20 to 0 before doing an unexpected written in the task is related to the operations of suppressing rel-
free recall test of all the items that had been presented. Instructions, evant information and storing and retrieving target infor-
 WORKING MEMORY AND UPDATING IN READING COMPREHENSION 349

Figure 1. Mean percentages of words correctly recalled by good and poor comprehenders in the Ex-
periment 3 updating task under high or low maintenance (Maint) and suppression (Supp) requests.

mation. Memory and suppression requests affected both ally different from the pattern we found in the preceding
recall and intrusion errors, in both groups. However, an experiment, since the absolute values of target items re-
interaction between groups and suppression requests was called were lower in this experiment than in the preced-
observed for intrusion errors. This result suggests that ing one. It is possible that the interpolated task introduced
poor comprehenders have a specific difficulty suppress- in the present experiment had the effect of impairing re-
ing relevant information when there is a high number of call of items and increasing the role of long-term seman-
items to be suppressed. In fact, good comprehenders made tic abilities that may be superior in good comprehenders
a similar number of intrusion errors with high- and low- (Cornoldi & Oakhill, 1991).
suppression lists. On the contrary, poor comprehenders’
intrusion errors in high-suppression lists were twice as EXPERIMENT 4
many (6.27 vs. 3.73), as compared with the low-suppression
lists. This effect was specific to intrusions. By contrast, The aim of Experiment 4 was to obtain further evi-
the two groups appeared to be similarly affected by in- dence concerning the effects found in the preceding ex-
creased memory request (see Figure 2). These data sug- periments and to explore the characteristics of intrusion
gest that poor comprehenders’ difficulty in avoiding in- errors. In fact, an intrusion error could be due to an item
trusion errors is not due to an increased memory load or either in a preceding list or in the same list. Intrusions of
retrieval effort but is related to the quantity of relevant in- items in the same list may be considered as working
formation that must be suppressed. An increase in the memory errors, because the intrusive elements are pre-
number of items to be selected affects the two groups sim- sumably maintained in working memory. Poor compre-
ilarly; an increase in items to be suppressed specifically henders, who have a poorer working memory system,
adds to the difficulty poor comprehenders have in avoid- should be differentiated by this latter measure, but not
ing intrusion errors. by the former. Furthermore, De Beni and Palladino
The data on final recall again showed that good com- (2001) have argued that it may be possible to distinguish
prehenders had a better memory for target items, which between two types of intrusions: intrusions of relevant
substantially mirrored the difference already found in items that can be immediately excluded from the pool of
working memory recall, as in the preceding experiment. target items (immediate intrusions) and those of relevant
However, in this case, we also found an unexpected dif- items that can only later be excluded from the set of tar-
ference between groups in the recall of relevant nontar- get items (delayed intrusions). Let us consider the list in-
get information. It should be noted that the performance troduced previously as an example. The items were pre-
pattern of the participants in this experiment was gener- sented in the following order: pen, relationship, wallet, cot,
350 PALLADINO, CORNOLDI, DE BENI, AND PAZZAGLIA

Figure 2. Mean numbers of intrusions in the Experiment 3 updating task made by

good and poor comprehenders under high or low maintenance (Maint) and suppres-
sion (Supp) requests.

ladder, glasses, pain, exercise book, rubber, train, chair, Materials and Procedure. The materials and procedure were
and violin, where pen, wallet, eyeglasses, exercise book, the same as those in Experiment 3. In this case, final recall was not
and rubber were the target items. The other relevant items, required.
cot and ladder, had to be considered for a while within
the pool of relevant items and then excluded (delayed ex- Results
clusions). Although larger than the target items, cot and Table 3 presents the mean percentages of nouns con-
ladder were presented when the working memory was not tained in the four types of lists that were recalled by the
yet occupied by smaller items. By contrast, the items train, two groups. A 2 3 2 3 2 ANOVA revealed significant
chair, and violin could immediately be considered as being effects of group [F(1,28) = 6.20, MSe = 208.70, p = .019]
nonappropriate(immediate exclusions), because they were and memory request [F(1,28) = 59.60, MSe = 87.76, p ,
presented at the end of the list when smaller items were al- .001]. The effect of suppression request only approached
ready in memory. significance [F(1,28) = 3.80, MS e = 82.48, p = .061].
The other type of intrusion concerned the items in pre- In the analysis of intrusion errors, we considered a
ceding lists that we assume no longer pertain to working source of error factor by considering intrusions from the
memory but, rather, to long-term memory. May, Hasher, same list, as well as those from preceding lists (see Ta-
and Kane (1999), in a study with older participants, found ble 4). A 2 3 2 ANOVA (groups 3 source of error) showed
that they were more affected by proactive interference a main significant effect of type of intrusion [F(1,28) =
produced by items from preceding lists. If intrusion er- 85.89, MSe = 20.70, p , .001], owing to the fact that in-
rors generically involve all the memory systems and, there- trusions were mainly items from the same list, and a sig-
fore, all kinds of information, the distinction between nificant interaction between groups and type of intrusion
items belonging to the same and those belonging to pre- [F(1,28) = 4.93, MSe = 20.70, p = .035]. Single-cell com-
ceding lists should not have any influence. If, on the con- parisons revealed that poor comprehenders had signifi-
trary, the poor comprehenders’ intrusion problem is re- cantly more intrusions of items in the same list [t(28) =
lated to the manipulation of highly accessible information, 2.13, p = .042]. On the contrary, the two groups were not
intrusions would be specifically due to items from the lat- significantly different for intrusions from items in preced-
est list, but not those from preceding lists. ing lists, and good comprehenders made an even higher
number of errors.
Method For intrusions from items in the same list, we consid-
Participants. Thirteen poor comprehenders and 17 good com- ered further the immediate intrusions (sizable objects that,
prehenders, similar in characteristics to the groups of participants owing to their size, can be immediately excluded) and
tested in Experiment 1, participated in the experiment. the delayed intrusions (items that can be excluded only
 WORKING MEMORY AND UPDATING IN READING COMPREHENSION 351

Table 3
Mean Percentages and Standard Deviations ory interference is less crucial than working memory in-
of Correctly Recalled Words for Poor terference in the difference between groups in the updating
and Good Comprehenders (Experiment 4) task.
Poor Good
Comprehenders Comprehenders EXPERIMENT 5
Type of List M SD M SD
High load, high suppression 67.69 13.08 68.82 9.71 The data of the last three experiments were coherent in
High load, low suppression 70.51 9.31 78.04 8.17 showing that poor comprehenders have difficulty updat-
Low load, high suppression 80.77 11.70 87.91 10.25 ing information in working memory. The task we devised
Low load, low suppression 79.49 14.94 90.20 7.99
to test this assumption—that is, the request to remember
only the smallest items—was very simple and easy to ac-
complish not only for young adults, but also for children.
after a while). This contrast must be considered with cau- However, it is possible that the same task became difficult
tion, since the items could not be balanced. We did not in the context of the updating task. In other words, it is
find a greater increase in delayed intrusions in the poor possible that poor comprehenders, having difficulty in
comprehender group; in fact, poor comprehenders made the updating task, became disoriented even with the sim-
37% more delayed intrusions, a proportion that was even ple task of evaluating item size. In this case, low perfor-
lower than that observed for immediate intrusions (61% mance in memory for the smallest items would not be due
increase). However, in both groups, delayed intrusions to a lack of updating, but to a failure in recognition of the
were more frequent than immediate intrusions. smallest items owing to attention lapse in the complex
double task. In order to test this possibility, we ran a fur-
Discussion ther experiment that explicitly required the participants to
The results of Experiment 4 substantially replicated indicate the size of each item presented. If poor compre-
the results of the preceding experiments. Poor compre- henders have difficulty in this evaluation, it may be an
henders had poorer working memory, as is manifested by indication that their poorer performance in the updating
their performance in an updating memory task. In the pre- task is related to an inability to correctly identify the tar-
sent case, the differences between groups were less dra- get items, rather than to updating per se.
matic, since the poor comprehenders were selected on It must be noted that this new request emphasizes the
the basis of a less strict criterion. importance of the smallest items, and this manipulation
The main goal of this experiment concerned the pos- could also affect overall performance in the updating task.
sibility of finding different intrusion error patterns in the In fact, we assumed that poor comprehenders’ difficulty
groups of poor and good comprehenders. In fact, we in updating was due to the fact that they were also paying
found that poor comprehenders made more intrusion er- attention to items that were irrelevant with respect to the
rors only with respect to items in the latest list and inde- critical feature of being small. Here, this inappropriate fo-
pendent of the exclusion delay within the list. This result cusing should be reduced, with the probable consequence
confirms the hypothesis that poor comprehenders’ intru- of a reduction in poor comprehenders’ difficulty.
sion errors are therefore specifically related to the work-
ing memory system. Since the critical variable differen- Method
tiating poor and good comprehenders is comprehension Participants. Fourteen poor comprehenders and 15 good com-
ability, the results suggest that a similar pattern of intru- prehenders similar in characteristics to the groups tested in Exper-
sion errors is also related to text comprehension. The ef- iments 1, 2, and 4 were selected.
fect seems to be due to high memory strength informa- Materials and Procedure. The materials were the same as those
tion, rather than to a weakened trace and loss of memory in Experiments 3 and 4, and the procedure was the same as that in
Experiment 4, except that the participant had to give a size estima-
discriminability, as was found by De Beni et al. (1998). tion for each item. After presentation of each item, the participant
The present results suggest that task difficulty in the had to indicate, for the sized items, its size on a scale from 0 to 100,
present study was not related to time of permanence in
memory before exclusion but, rather, to a more complex
series of strategies that may include not only the possi- Table 4
bility of immediate exclusion, but also the possibility of Means and Standard Deviations of Different
maintaining a large part of the material at different levels Types of Intrusion Errors Produced by Poor
and Good Comprehenders (Experiment 4)
of activation.
The absence of a difference between groups in number Poor Good
Comprehenders Comprehenders
of intrusions from a previous list does not necessarily
Intrusions M SD M SD
mean that the two groups (poor and good comprehenders)
Immediate 5.38 3.73 3.35 2.21
were affected to a similar extent by proactive interfer-
Delayed 9.77 5.13 7.12 3.30
ence. It has been shown that only a small part of the long Total from the same list
term memory interference effect is emphasized by an in- (immediate + delayed) 15.15 7.58 10.47 4.37
crease in the number of intrusions (Barnes & Underwood, Total from
1959). However, this result suggests that long-term mem- preceding lists 1.54 1.66 2.12 1.11
352 PALLADINO, CORNOLDI, DE BENI, AND PAZZAGLIA

Table 5
Mean Numbers of Correctly Individuated Items, Mean Percentages
of Correctly Recalled Words in the Different Lists, and Mean Numbers
of Intrusion Errors by Poor and Good Comprehenders (Experiment 5)
Poor Good
Comprehenders Comprehenders
M SD M SD
Correct identifications (max = 96) 93.71 3.32 94.00 1.89
High load, high suppression (%) 66.19 8.56 71.11 9.97
High load, low suppression (%) 65.48 9.92 75.56 8.03
Low load, high suppression (%) 82.14 13.64 85.19 7.76
Low load, low suppression (%) 85.32 9.65 89.63 6.25
Intrusions 4.57 4.35 5.13 3.36

whereas for the filler items he or she just had to say “unsized.” The intrusions in the task is confirmed by the significant cor-
participants were instructed on how to do the task and were pre- relation between the number of immediate intrusions and
sented with examples of items of different sizes in order to become the number of correct recalls (Pearson’s r = 2.56, p ,
familiar with the 0–100 scale. They were also instructed to give an
immediate response, because the presentation rate could not be .01) found when considering all the participants, whereas
changed. At the end of the list, the task proceeded in the same way the corresponding correlation between delayed intrusions
as in the preceding experiments. and correct recalls did not reach significance (r = 2.04).

Results Discussion
The participants did not have difficulty with the task, The results of Experiment 5 clearly show that the poor
since they were able to give immediate responses for the comprehenders’ difficulty in the updating task proposed
size request. The fact that neither the good nor the poor in the preceding experiments was not due to an inability
comprehenders had difficulty with the size evaluation is to evaluate the size of items and then to identify the small-
confirmed by the very high numbers of smallest items est ones. Despite the accuracy of their ratings, the poor
that actually received the smallest evaluations. This oc- comprehenders had a lower recall of the target (i.e., small-
curred indifferently for all the types of lists, including the est) items. Therefore, the difficulty of the poor compre-
most difficult ones. Table 5 presents the main results of henders was not due to inappropriate identification but
the study. seems to have been due to an inability to select the ap-
The number of smallest items correctly identified was propriate items during recall. However, in this case, the
very close to the maximum number of 96 in both groups, poor comprehenders’ working memory difficulty was not
and the difference between groups was not significant related to a general increase in the number of intrusion
( p = .78). However, poor comprehenders still had diffi- errors but, possibly, to an inability to immediately sup-
culty in the updating task—that is, in recalling the small- press large items (immediate intrusions). The two groups
est items. A 2 3 2 3 2 mixed design ANOVA (group 3 made a very low number of intrusions. This dramatic de-
suppression request 3 memory request) on the percent- crease appears to have been due to the request to evaluate
ages of correct responses revealed a significant group ef- the size of the items, which presumably caused the partic-
fect [F(1,27) = 4.39, MSe = 206.19, p , .05]. Furthermore, ipants’ attention to be focused on the items that received
the main effects related to the nature of the lists were also absolute low ratings, reinforcing the appropriate selection.
significant for both suppression request [F(1,27) = 5.27, In fact, many of the participants reported that sometimes
MS e = 44.27, p , .05] and memory request [F(1,27) = during recall they were able to remember the numbers
135.53, MS e = 54.62, p , .001]. given in the ratings even better than the related items. This
Despite a difference in correct recalls, the two groups occurred in particular for the items that received the low-
did not differ in the number of intrusion errors ( p = .70). est ratings. For example, a participant could say “Well, I
In fact, the new task requirement had a general effect of remember there was an item I gave a rating of 5 to, now
reducing the tendency to recall large objects. Intrusion er- I will try to retrieve it.” In our opinion, this focus on the
rors in both groups were lower than those in the preced- smallest items contributed further to reducing the focus
ing experiments, with reductions even greater than 50%. on the larger items, thus reducing the probability that the
The last line of Table 5 presents the mean numbers of latter would be remembered.
intrusion errors owing to the recall of items from the same
list that were not the smallest ones. The data concern im- GENERAL DISCUSSIO N
mediate and delayed suppressions considered together,
since the pattern of errors was similar for the two. Poor A large body of literature (for a meta-analysis, see Dane-
comprehenders had a slight, but not significant, tendency man & Merickle, 1996) has shown that working memory
to make more immediate intrusions (2.43 vs. 2.14), and is associated with reading comprehension ability. In fact,
good comprehenders a tendency to make fewer immediate reading comprehension requires that parts of the text are
intrusions (1.80 vs. 3.33). The importance of immediate temporarily held in a working memory system to be ana-
 WORKING MEMORY AND UPDATING IN READING COMPREHENSION 353

lyzed and integrated. A critical working memory opera- Jones, 1990) assumes that the participant holds the tar-
tion involved in reading seems to be updating relevant get items in working memory, and then, when new target
information. However, no previous research has shown items arrive, those that are no longer targets are dropped.
any relation between reading ability and the ability to For example, it could be assumed that in the updating
carry out a working memory updating task. The present tasks of Experiments 2, 3, 4, and 5, the participant selects
study offers considerable evidence in favor of this rela- the three (or five) smallest items, and if another small item
tion. In fact, five different groups of poor comprehenders was presented, it was added to his or her memory, and the
presented a similar pattern of difficulty in updating tasks. largest one maintained was then dropped. However, our
Experiments 2, 3, and 4 also show that poor compre- data suggest that the process is more complicated. For ex-
henders, under certain conditions, tend to make more in- ample, we found that in Experiment 1, although they had
trusion errors—that is, they remember more items that are a span as high as good comprehenders, poor compre-
relevant but not target items. This last result is intriguing, henders were poorer also with the four-item list updating
because it adds evidence to the hypothesis that success in task—that is, when no drop–substitution process was re-
working memory and comprehension tasks is related not quired [good comprehenders: M = 15.64, SD = 1.08;
only to the ability to recall selected information, but also poor comprehenders: M = 14.25, SD = 2.32; t(28) = 2.15,
to the ability to suppress other potentially relevant but p = .043]. This result is against the simple view of up-
not target information (e.g., De Beni et al., 1998; Gerns- dating as an “all or nothing” process of maintenance–
bacher, Varner, & Faust, 1990). The fact that the same ef- substitution. In Experiments 4 and 5, we found that poor
fect was not found in Experiments 1 and 5 confirms that comprehenders’ difficulty was not related more to delayed
the two updating tasks did not measure exactly the same exclusions. This suggests that their difficulties lie not
process. Experiments 1 and 5 highlighted, in particular, only in substitution, but also in selection/maintenance.
the target items through either final position (Experi- It must also be noted that, if reduction of the activation/
ment 1) or the request to explicitly evaluate the size of elimination from memory of items no longer included in
the objects. Since the intrusion error appears to be af- the target pool is immediate, it would be difficult to ex-
fected by the stress given to nontarget items (De Beni plain why good comprehenders, who have better updat-
et al., 1998; Passolunghi, Cornoldi, & De Liberto, 1999), ing ability, should also have a better long-term memory
appropriate emphasis given to target items should have for all the items presented.
the consequence of reducing the emphasis on, and ten- Altogether, these findings suggest that updating is not
dency to recall, nontarget items. a simple inclusion/exclusion process. It is probably a
Furthermore, we found there to be general effects more complex process that attributes different levels of
owing to poorer performance in the updating task with in- activation to the items presented and continuously up-
creases in suppression and memory requests. These data dates that level while maintaining a larger set of elements
offer a first insight into the nature of the updating process. activated. In reading comprehension, the reader should
The memory request effect shows that updating is more keep various pieces of information available in a similar
complex when more information must be selected, prob- way, tuning activation level to their relevance and impor-
ably because of limitations in the working memory sys- tance, until a final interpretation can be made and a clear
tem, which becomes overloaded by the memory load in text representation or mental model built. Obviously, this
addition to the operations required by the task. The sup- theoretical perspective is purely speculative (even if con-
pression request effect shows that success in the updat- sistent with some literature and reading comprehension
ing task is related not only to the memory request—that models; see, e.g., Gernsbacher, 1990; van Dijk & Kintsch,
is, the number of elements that must be remembered— 1983) and requires experimental evidence.
but also to the number of elements that must be excluded. In explaining why poor comprehenders fail our updat-
Obviously, these elements were also potential targets, and ing tasks, we must consider the hypotheses advanced for
a higher suppression request also implies an increase in the similar failure of poor comprehenders in other work-
memory load, which cannot easily be dissociated from the ing memory tasks. The ability to process sentences has
suppression request per se. not been ruled out as a factor explaining the relation of the
Our observations apparently do not agree with those listening span test to reading comprehension (Daneman
of Morris and Jones (1990), who found that the number & Carpenter, 1980, 1983). However, this explanation can
of updates and a concurrent task do not depress updating be ruled out, because no sentence processing was required
performance. However, it should be noted that, as well as in these updating tasks. Similarly, we can reject the hy-
involving differences in materials and procedure, Morris pothesis that the relation between reading comprehension
and Jones’s concurrent task concerned the articulatory and updating ability is simply due to very general intel-
loop, rather than the more active central components of lectual ability, since the participants were matched for at
working memory, as did our suppression and memory least some intelligence components. In Experiment 1,
request manipulations. the participants were also matched for their short-term
In order to understand completely the psychological memory ability, confirming that working memory differ-
mechanisms involved in updating, new and further evi- ences are not necessarily related to immediate memory
dence is necessary. A simple view of updating (Morris & differences (see De Beni et al., 1998).
354 PALLADINO, CORNOLDI, DE BENI, AND PAZZAGLIA

Engle and coauthors (e.g., Conway, Tuholski, Shisler, & di comprensione nella lettura [Advanced MT reading comprehension
Engle, 1999; Engle, Cantor, & Carullo, 1992) have argued tests]. Florence: Organizzazioni Speciali.
Daneman, M., & Carpenter, P. A. (1980). Individual differences in
that success in remembering relevant, and in suppressing
working memory and reading. Journal of Verbal Learning & Verbal
irrelevant, information in working memory tasks is related Behaviour, 19, 450-466.
to the quantity of cognitive resources available to the Daneman, M., & Carpenter, P. A. (1983). Individual differences in
working memory system. This perspective seems attrac- integrating information between and within sentences. Journal of Ex-
tive, since it is compatible with the view that the updating perimental Psychology: Learning, Memory, & Cognition, 9, 561-584.
Daneman, M., & Merikle, P. M. (1996). Working memory and lan-
task requires complex management of the activation pro- guage comprehension: A meta-analysis. Psychonomic Bulletin & Re-
cesses, modulating at each moment the status of all the view, 3, 422-433.
items. It also takes into consideration the suppression re- De Beni, R., & Palladino, P. (2001). Decay in working memory per-
quest involved in this task. formances with aging: Intrusion errors analyses. Manuscript sub-
mitted for publication.
The particular results of Experiments 1 and 5 (differ- De Beni, R., Palladino, P., Pazzaglia, F., & Cornoldi, C. (1998).
ence between groups in recall, but not in intrusions) show Increases in intrusion errors and working memory deficit of poor
that, in our task, the participants did not adopt the strategy comprehenders. Quarterly Journal of Experimental Psychology,
of using any item at all to make up the expected number 51A, 305-320.
of items to be recalled. The higher number of intrusions de Vega, M. (1995). Backward updating of mental models during con-
tinuous reading of narratives. Journal of Experimental Psychology:
for poor comprehenders in Experiments 2, 3, and 4 seems Learning, Memory, & Cognition, 21, 373-385.
to be due to poor suppression mechanisms, rather than to Ehrlich, K., & Rayner, K. (1983). Pronoun assignment and semantic
a particular task artifact. In our view, the exclusion of integration during reading: Eye movements and immediacy of pro-
possible intrusions occurs during both the processing of cessing. Journal of Verbal Learning & Verbal Behavior, 22, 75-87.
Engle, R. W., Cantor, J., & Carullo, J. J. (1992). Individual differ-
incoming information and the retrieval process involved
ences in working memory and comprehension: A test of four hypothe-
in giving a response. During presentation, the participant ses. Journal of Experimental Psychology: Learning, Memory, & Cog-
processes each incoming item, attributes appropriate ac- nition, 18, 972-992.
tivation to the item, and updates its activation. If an item Garner, R. (1987). Metacognition and reading comprehension. Nor-
is wrongly activated or its activation is not updated, it will wood, NJ: Ablex.
Gernsbacher, M. A. (1990). Language comprehension as structure
disturb the process of retrieval of pertinent information building. Hillsdale, NJ: Erlbaum.
and, at worst, will even be remembered erroneously. In Gernsbacher, M. A., Varner, K. R., & Faust, M. E. (1990). Investi-
fact, an insufficient suppression mechanism makes the gating differences in general comprehension skill. Journal of Exper-
number of possible candidates greater, producing a greater imental Psychology: Learning, Memory, & Cognition, 16, 430-445.
Glenberg, A. M., & Langston, W. E. (1992). Comprehension of il-
probability of remembering nontarget elements. lustrated text: Pictures help to build mental models. Journal of Mem-
In conclusion, our data show consistently that the abil- ory & Language, 31, 129-151.
ity to update information in working memory is strictly May, C. P., Hasher, L., & Kane, M. J. (1999). The role of interference
related to reading comprehension ability. This connection in memory span. Memory & Cognition, 27, 759-767.
appears to be due to a common underlying control mech- McKoon, G., & Ratcliff, R. (1992). Inferences during reading. Psy-
chological Review, 99, 440-466.
anism, the characteristics of which should be investigated Morris, N., & Jones, D. M. (1990). Memory updating in working
further in future research. memory: The role of central executive. British Journal of Psychology,
81, 111-121.
REFERENCES Morrow, D. G., Bower, G. H., & Greenspan, S. L. (1989). Updating
situation models during comprehension. Journal of Memory & Lan-
Anastasi, A. (1988). Psychological testing (6th ed.). New York: Macmil- guage, 13, 441-469.
lan. Orsini, A., & Laicardi, C. (1997). WAIS–R: Contributo alla taratura
Baddeley, A. D. (1986). Working memory. Oxford: Oxford University italiana [A contribution to Italian standardization]. Florence: Orga-
Press. nizzazioni Speciali.
Baddeley, A. D. (1990). Human memory: Theory and practice. Hove, Passolunghi, M. C., Cornoldi, C., & De Liberto, S. (1999). Working
U.K.: Erlbaum. memory and intrusions of irrelevant information in a group of spe-
Barnes, J. M., & Underwood, B. J. (1959). “Fate” of first-list associ- cific poor problem solvers. Memory & Cognition, 27, 779-790.
ations in transfer theory. Journal of Experimental Psychology, 58, Pollack, I., Johnson, L., & Knaft, P. (1959). Running memory span.
97-105. Journal of Experimental Psychology, 57, 137-146.
Brown, A. L., Armbruster, B. B., & Baker, L. (1986). The role of Robertson, R. R. W., & Gernsbacher, M. A. (1996, May). Suppres-
metacognition in reading and studying. In J. Orasanu (Ed.), Reading sion during narrative comprehension. Paper presented at the 68th
comprehension: From research to practice (pp. 49-75). Hillsdale, NJ: Annual Meeting of the Midwestern Psychological Association,
Erlbaum. Chicago.
Clark, H. H., & Sengul, C. J. (1979). In search of referents for nouns Van der Linden, M., Bredart, S., & Beerten, A. (1994). Age-related
and pronouns. Memory & Cognition, 7, 35-41. differences in updating working memory. British Journal of Psy-
Conway, A. R. A., Tuholski, S. W., Shisler, R. J., & Engle, R. W. chology, 85, 145-152.
(1999). The effect of memory load on negative priming: An individ- van Dijk, T. A., & Kintsch, W. (1983). Strategies of discourse com-
ual differences investigation. Memory & Cognition, 27, 1042-1050. prehension. New York: Academic Press.
Cornoldi, C. (1974). Imagery values for 310 Italian nouns. Italian Wechsler, D. (1987). Manual for Wechsler Memory Scale–Revised.
Journal of Psychology, 1, 211-225. New York: Psychological Corporation.
Cornoldi, C., & Oakhill, J. (1991). Reading comprehension difficul-
ties: Processes and interventions. Hove, U.K.: Erlbaum. (Manuscript received October 13, 1998;
Cornoldi, C., Rizzo, A., & Pra Baldi, A. (1991). Prove avanzate MT revision accepted for publication June 26, 2000.)

View publication stats