a comparison of backward and forward spatial spans

This article was downloaded by: [Lycoming College]On: 10 October 2014, At: 22:24Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office:Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

The Quarterly Journal of ExperimentalPsychologyPublication details, including instructions for authors and subscriptioninformation:http://www.tandfonline.com/loi/pqje20

A comparison of backward and forwardspatial spansCesare Cornoldi a & Irene C. Mammarella aa Università di Padova , Padua, ItalyPublished online: 21 Apr 2008.

To cite this article: Cesare Cornoldi & Irene C. Mammarella (2008) A comparison of backward and forwardspatial spans, The Quarterly Journal of Experimental Psychology, 61:5, 674-682

To link to this article: http://dx.doi.org/10.1080/17470210701774200

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)contained in the publications on our platform. However, Taylor & Francis, our agents, and ourlicensors make no representations or warranties whatsoever as to the accuracy, completeness, orsuitability for any purpose of the Content. Any opinions and views expressed in this publicationare the opinions and views of the authors, and are not the views of or endorsed by Taylor &Francis. The accuracy of the Content should not be relied upon and should be independentlyverified with primary sources of information. Taylor and Francis shall not be liable for anylosses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantialor systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, ordistribution in any form to anyone is expressly forbidden. Terms & Conditions of access and usecan be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/pqje20

http://dx.doi.org/10.1080/17470210701774200

http://www.tandfonline.com/page/terms-and-conditions

Short article

A comparison of backward and forward spatial spans

Cesare Cornoldi and Irene C. MammarellaUniversita di Padova, Padua, Italy

The standard Corsi blocks task is frequently used to measure the capacity of visuospatial workingmemory, but the implications of the use of both forward and backward recall are still unclear. Inthe present study, we showed that the backward Corsi task is particularly powerful in discriminatingbetween low- and high-spatial-ability individuals and involves different processes from those involvedin the forward task. From a sample of 425 participants we selected one group of 20 high-spatial-abilityparticipants and one of 20 low-spatial-ability participants. The results demonstrated that a backwardspatial span offers specific information not available from a forward spatial span, and that there was nofacilitation due to a descending format. In particular, in the low-spatial-ability group, performancewas generally poorer, but backward Corsi recall was lower than forward recall, and participants didnot show any advantage following the descending presentation format—which in some contexts isconsidered to reduce proactive interference. We conclude that the backward Corsi task has specificvalue and that the assumption of fully parallel verbal and visuospatial working-memory systems canlead to a variety of misunderstandings.

Keywords: Corsi blocks task; Proactive interference; Visuospatial working memory.

The simple spatial span, in the Corsi blocks task orsimilar versions, represents the most direct way ofstudying visuospatial working memory (VSWM),similarly to the use of the simple word or digitspans for verbal working memory (VWM).However, the implications of the use of thespatial spans and their relationships with moregeneral spatial abilities are still not well studied.The present research examines these issues withparticular reference to a computerized presen-tation of the most popular spatial span—that is,

the Corsi blocks task, in its forward and backwardrecall.

Several cognitive psychology studies addressingVSWMhave examined the presence of effects par-allel to those detected in verbal working memory.For example, VWM studies have revealed thephonological similarity, unattended speech, andarticulatory suppression and length effects(Baddeley, 1986), while, analogously, studies onVSWM have revealed the visual similarity effect(e.g., Avons & Mason, 1999), the unattended

Correspondence should be addressed to Cesare Cornoldi, Dipartimento di Psicologia Generale, Universita degli Studi di Padova,

Via Venezia, 8, 35131, Padova, Italy. E-mail: [email protected]

We wish to thank Vera Zanella for her help in recruiting and testing participants in this study. We are also grateful to Josee

Turcotte and Gillian Rowe for an initial version of the computer program used in the present research.

674 # 2007 The Experimental Psychology Society

http://www.psypress.com/qjep DOI:10.1080/17470210701774200

THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY

2008, 61 (5), 674–682

Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

visual stimuli effect (e.g., FitzGerard &Broadbent, 1985), the spatial suppression effect(e.g., Salway & Logie, 1995), and the complexityeffect (e.g., Kemps, 2001). In other words,although it is not confirmed that VWM andVSWM have exactly the same properties or struc-tures, studying the analogies between them canfurther our understanding of the lesser knownvisuospatial system.

Verbal working memory has typically beentested using a series of different measures ofordered sequential recall with a self-terminatingspan procedure. The best known short-termmemory measure, the digit span, is widely usedin forward and backward recall directions.Comparison of performance in these two direc-tions has produced a number of assumptions; inparticular, performance in backward recall isalways poorer and has been associated to directinvolvement of the central executive (Baddeley,1986).

Recent research—using an exhaustive ratherthan self-terminating procedure—revealed anotherimportant implication of the working-memoryspans—that is, their susceptibility to proactiveinterference (PI). In particular, May, Hasher, andKane (1999; see also Lustig, May, & Hasher,2001) hypothesized that the classical spanprocedure creates a large amount of PI, becausethe series are presented in an ascendingpresentation format. Hence, the longest seriesbecome critical in the attribution of the spanscore. According to the authors, PI is built upacross the series that constitute the span tasks. Byconsequence, PI concerning the most critical longseries is greater for the ascending format (wheremany other series are presented before the longseries) than for the descending format, where thelong series are presented first. In their study,using a version of the reading-span task derivedfrom Daneman and Carpenter (1980), young andold adults were presented with a classical ascend-ing-span procedure—beginning with shortersequences and moving on to longer sequences—and a descending procedure—longer sequencesmoving on to shorter sequences. In two experi-ments, low-span participants (old adults)

showed poor performances with the classical spanprocedure but similar performance to young adultsfor the descending procedure. The susceptibilityto PI of VWM has had further confirmationusing the Rasch psychometric model (Bowles &Salthouse, 2003) and the operation span(Bunting, 2006).

Assuming an analogy between the VWM andVSWM would imply that the effects found withverbal span tasks would also apply to visuospatialspan tasks. First, visuospatial backward recall(compared to forward recall) would be assumedto produce a fall in performance, similar to thatobserved in verbal recall, in correspondence withhigher controlled processes. Secondly, perform-ance in typical spatial spans would be assumed tobe considerably affected by PI, especially inpeople with VSWM impairments associated withgreater susceptibility to interference.

On the other hand, the assumption of a paral-lelism between VWM and VSWM runs into anumber of difficulties, since many aspects of thetwo systems have different specificities. Asregards the recall direction, it has been observedthat order is a critical feature in verbal memory,but not in visuospatial memory: In fact, in thelatter a pathway can be recalled not only as asequential series but also as an overall pathwayincluding a series of positions. People either canhave a simultaneous representation of the overallpathway or else easily process different locationsalong a pathway in an order different from theoriginal (Mammarella & Cornoldi, 2005b): Inparticular, especially for very long series, recall oflocations could benefit by starting from the lastlocation. In fact, researches comparing forwardand backward recall of the Corsi blocks task gaveconflicting results, sometimes showing that thebackward recall direction does not necessarilyimply a decrease in performance with respect toforward recall, sometimes showing a decrease.An absence of decrease for backward recall, evenif observed under particular conditions (Wilde &Strauss, 2003), could raise the issue of whetherthe backward spatial span is measuring the sameVSWM component as the forward span or elsesimply adds new information. Currently specific

THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2008, 61 (5) 675

A COMPARISON OF BACKWARD AND FORWARD SPATIAL SPANS

Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

interpretations regarding the backward Corsiblocks task have been proposed in the literature,and the task has been associated with the centralexecutive component of working memory(Vandierendonck, Kemps, Fastame, & Szmalec,2004), with the binding between order andidentity locations (Vandierendonck & Szmalec,2005) and with spatial-simultaneous processes(Mammarella & Cornoldi, 2005b).

Also in the case of PI, spatial spans seem to haveimplications that differ from typical verbal spans.For example, typical verbal information has a highdegree of familiarity and to a larger extent involveslong-term memory and preexisting clusters ofinformation (Baddeley, Gathercole, & Papagno,1998). Instead, for spatial span tasks, locationsand their clusters—while perhaps supported by pre-existing knowledge (Kemps, 2001)—are not alreadystored in long-term memory and are graduallybetter learned and refined during the task: In thecase of VSWM, therefore, PI might be smallerthan for VWM but compensated by the advantageof greater knowledge about the stimuli.Furthermore, rehearsal processes seem to be differ-ent in verbal memory, where subvocal rehearsal isparticularly prone to interference. In VSWMtasks, no consensus has been reached to dateabout a description, but there is evidence(Kosslyn, 1990) that it may be based on a continu-ous image generation process involving an overallimage rather than single locations. It is thus possiblethat this type of rehearsal involving a limited seriesof representations is less sensitive to PI than oneinvolving a very high number of single items.Also Awh, Jonides, and Reuter-Lorenz (1998,Exp. 3) found that memory for locations has itsown characteristics and is not dependent on anorderly correspondence between the memorizedlocations and intervening stimuli. In particular,Fischer (2001) found no differences betweenascending and descending presentation formats,manipulating the encoding interval, retentiontime, and response alternatives (i.e., the items pre-sented remained visible during reproduction):However, the study did not involve low-abilityindividuals, as distinct from the studies of Mayet al. (1999).

In the present research we tested high- andlow-spatial-ability participants with the Corsiblocks task varying recall direction (forward vs.backward) and presentation format (ascending vs.descending). It should be noted that, as distinctfrom self-terminating procedures that halttesting when the participant repeatedly fails, theprocedure requires presentation of all the differentsequence lengths in both formats: This allowscomparison of the proportion of correct responseswith different lengths for both presentation formatand direction of recall. In fact, Vandierendonckand Szmalec (2005) have shown that sequencelength can, under specific conditions, affectforward and backward recall to different extents.In the present study, we deemed it relevant to con-sider the case of the supraspan lists where completerecall is very rare, and participants might benefit,in backward recall, from being able to immediatelypoint to the locations just presented.

We selected low- and high-spatial-abilitygroups on the basis of different performances ina typical standard spatial task used for measuringgeneral spatial abilities—that is, a paper-and-pencil mental rotation test (Vanderberg & Kuse,1978). Inclusion of the low-spatial-ability groupin the study allowed us to examine the sensitivityof a backward spatial span both in absolute termsand with respect to a forward spatial span.Furthermore, it allowed specific implications ofrecall direction and presentation format to betested. Indeed, Mammarella and Cornoldi(2005b) found recall direction to be critical forthe group of nonverbal learning disability children,but not for the control group, and an advantage ofdescending over ascending format seems presentonly with reduced span abilities such as theelderly, while nothing is known about implicationsof the presentation formats for other low-abilityindividuals. Inclusion of participants with lowspatial ability in our study also allowed furtherconfirmation that general spatial ability is criticallyassociated with impairment in VSWM. Althoughthe assumption is largely intuitive, the issue to datehas not undergone systematic study: Spatial abilityhas for the main part been examined within psy-chometric studies of components of intelligence,

676 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2008, 61 (5)

CORNOLDI AND MAMMARELLA

Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

while study of VSWM has almost exclusively beenlimited to the context of experimental research(but see the study of Kane et al., 2004).

Method

ParticipantsPhase 1. The first phase of this research was devotedto identifying participants with high and low spatialabilities. This was achieved by presenting 425students (mean age ¼ 19.96 years, SD ¼ 3.45; 83males and 342 females) in first-year psychology atPadua University with a demographic question-naire, followed by two group-screening tests: averbal comprehension task (MT Comprehensiontest, subtest number 4, Cornoldi, Rizzo, & PraBaldi, 1991) and a spatial task—that is, theMental Rotation Test (MRT; Vanderberg &Kuse, 1978). From the initial sample of 425 stu-dents we selected a high-spatial-ability (HSA)group of 20 participants and a low-spatial-abilitygroup (LSA), again of 20. The HSA groupincluded 18 females and 2 males (mean age ¼

19.05, SD ¼ 0.97) and the LSA group 19 femalesand 1 male (mean age ¼ 19.23, SD ¼ 0.94).Participants were included in the LSA groupprovided they obtained a MRT score lower thanthe 10th percentile computed for the entire group(MRT , 3), and in the HSA group if theirMRT score exceeded the 90th percentile(MRT . 13). The two groups were similar inage, t(38) ¼ –0.612, p ¼ .544, MTComprehension test performance, t(38) ¼ –0.710,p ¼ .48, and socio-cultural background U(Mann–Whitney) ¼ 192.5, p ¼ .85, but differedin MRT performance, t(38) ¼ 22.77, p , .001.

Phase 2. In the second phase, the HSA and LSAgroups were given a computerized version of theCorsi blocks task under a number of differentconditions.

Stimuli and procedureFor each condition (forward and backward recalldirections each in ascending and descending pres-entation formats), three block sequences werearranged for each sequence length (3 through 8),

corresponding to a total of 18 sequences pertype. The stimuli were presented on a 15-inchvideo monitor. The nine blocks were grey 2 �

2-cm squares placed at their relative standardpositions on a white background. Presentationof block sequences was computer monitored:Each block was highlighted by changing colourto black for 1,000 ms, with a blank interval of500 ms. A sequence (trial) started when the par-ticipant pressed the space-bar, and its conclusionwas indicated by a large square around the Corsidisplay for 500 ms. When the signal disappeared,the participant moved the cursor with the mouseand pressed the left mouse button to record theresponse locations. At the end of the responsephase, the participant pressed the space-bar tostart the next trial. The intertrial interval was1,200 ms.

The conditions tested were as follows:

. In the ascending presentation format, participantswere presented with three block sequences oflength 3 to 8 in both forward and backwardrecall. The presentation order of the forwardand backward recall was counterbalancedacross participants.

. In the descending presentation format, partici-pants were presented with exactly the sameblock sequences but starting with length 8 andconcluding with length 3. In this case too, thepresentation order of forward and backwardrecall was counterbalanced.

To summarize, participants in each groupwere randomly assigned to the between-subjectscondition (presentation format: ascending vs.descending), whereas recall direction (forwardvs. backward) was manipulated within subjects.

Results

Data scoring and data screeningA total of 2 participants (both belonging to theLSA group) who did not fully understand thetask were excluded; the following analyses aretherefore based on 38 participants (18 participantsperformed the ascending presentation format, and20 performed the descending presentation). We



Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

used a score (order score) based on the mean percen-tages of stimuli correctly recalled in terms of thepresentation order for each length, since thismeasure is more sensitive than the typical spanscore (see Fischer, 2001; Vandierendonck et al.,2004). For example, for a sequence “9–7–3–1–5–8” a response “9–7–4–1–2–8” received ascore of 66.67 (¼ 4/6 � 100). Statistics for eachgroup are reported in Figure 1.

In order to increase the observations for eachcell, we computed the mean percentages ofcorrect responses by pooling the responses foradjacent lengths (3–4, 5–6, and 7–8). A 2(groups: HSA vs. LSA) � 2 (presentationformat: ascending vs. descending) � 3 (sequencelengths: 3–4 vs. 5–6 vs. 7–8)� 2 (recall direction:forward vs. backward) mixed analysis of variance(ANOVA) showed a main effect of group, F(1,34) ¼ 6.94, MSE ¼ 397.96, p ¼ .013, hp

2 ¼ .17,resulting from the fact that HSA participant per-formance (77.41%) was better than that for theLSA group (70.41%). The main effect of recalldirection approached significance, F(1, 34) ¼

3.69, MSE ¼ 111.53, p ¼ .06, hp2 ¼ .10

(forward: 75.26%; backward 72.56%) whereas theinteraction recall direction by group was signifi-cant, F(1, 34) ¼ 4.56, MSE ¼ 111.53, p ¼ .04,hp2 ¼ .12 (see Figure 2). Post hoc analyses using

Tukey test revealed that forward and backwardrecall differed only in the LSA group (p , .05).

Moreover, results revealed a main effect ofsequence length, F(2, 68) ¼ 245.21, MSE ¼

125.79, p, .001, hp2¼ .88. A post hoc comparison

using Bonferroni’s correction showed that lengths3–4 (92.89%), 5–6 (76.31%), and 7–8 (52.53%)were all different (p , .001). The interactiongroup by sequence length was also significant,F(2, 68) ¼ 3.75, MSE ¼ 125.79, p ¼ .028, hp

2 ¼

.09, owing to the fact that the HSA group per-formed better than the LSA group only forlengths 5–6 and 7–8 (p , .01 in both cases) asdemonstrated by a Tukey test. Finally, the signifi-cant interaction recall direction by sequencelength, F(2, 68) ¼ 6.04, MSE ¼ 99.02, p ¼

.004, hp2 ¼ .15, resulted from the fact that the

forward and backward Corsi differed only inlength 5–6 (p , .01). We failed to find any

significant effect related to presentation format,F(1, 34) , 1, or indeed significant interactions.

Discussion

A main result of the study concerns the differencebetween low- and high-spatial-ability groups inoverall Corsi task performance. This result addsfurther evidence to the assumption that a lowVSWM is a critical feature of LSA participants.The mechanism underlying the relationshipbetween spatial ability and VSWM is still notclear, but—in view of mounting evidence ofthe critical role of working memory in tradi-tional intelligence measures—VSWM could beassumed to be critical for spatial reasoning, as itoffers the resources for maintaining and operatingon the spatial information that has to be manipu-lated in spatial intelligence tasks. However, thepresent results suggest that differences in spatialability are more likely to be observed in backwardrecall of the Corsi blocks task.

As regards comparison between the two recalldirections, there was no clear advantage offorward recall: This is in clear conflict with obser-vations using verbal spans measures (Wilde &Strauss, 2003). In fact, for the HSA group, per-formance for the two recall directions was almostidentical, with backward recall even slightlysuperior (77.56%) to forward recall (77.26%).Although the gap between forward and backwardverbal span might possibly be reduced also in ahigh-verbal-ability group, the general advantageof a forward order remains stable across studies(see Wilde & Strauss, 2003). The fact that thisadvantage is absent for the Corsi blocks taskshows that the classical interpretation of backwardrecall as a task relying largely on highly controlledprocesses (e.g., Vandierendonck et al., 2004)cannot be applied to the case of a VSWM span.Results are in line with the suggestion of Smythand Scholey (1992) that, whereas the centralexecutive is involved when the order of presen-tation of verbal items is reversed, additionalexecutive resources are not required to reversethe order of presentation of the Corsi blocks items.



Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

Figure 1.Mean percentage of correct responses for lengths 3–4 (L 3–4), 5–6 (L 5–6), and 7–8 (L 7–8) in forward and backward recall, by

high-spatial-ability (HSA) and low-spatial-ability (LSA) groups, for ascending (upper panel) and descending (lower panel) presentation

format. Error bars represent standard errors.



Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

In the present study, the contrast between thetwo recall directions is highlighted by the resultof the LSA group. In fact, this group showedspecific impairment in the backward recall direc-tion. The present study not only extends theresults obtained by Mammarella and Cornoldi(2005b) with nonverbal-learning-disabled chil-dren to a group of young adults with low spatialabilities, but also offers further support to thehypothesis that backward recall involves processesthat differ from those of forward recall (presum-ably more simultaneous and less sequential) andrelies to a greater extent on specific spatial pro-cesses. In the present study, the analysis of thesequence lengths provides a better understandingof the difference between the two recall directions;forward recall in fact produced significantly betterperformance with sequence lengths of 5–6 items.This result can be explained by considering thespecific implications of the different recall direc-tions. In particular, for short lists representationof the entire pathway is easy and can be used forrecall in the backward direction. In contrast, longsequences are particularly difficult, and recall of

the last locations can ensure that at least somelocations are remembered. However, for inter-mediate sequence lengths, item recall (locations)is also critical: This combined requirement ofitem recall and order could lead to specific disad-vantages in backward recall. The importance ofbinding order and locations has also been stressedby Vandierendonck and Szmalec (2005).

As regards presentation format, our resultsconfirm that this variable may have severaleffects, depending on a number of boundary con-ditions. First, an advantage of descending overascending format in the elderly has been repeatedlyobserved for verbal working memory (May et al.,1999; Lustig et al., 2001), whereas this effect wasnot present in this study when testing youngadults with low spatial abilities, even though chil-dren with nonverbal-learning disability also seemsusceptible to interference (Mammarella &Cornoldi, 2005a). Second, it should be notedthat verbal material employed for verbal spans ishighly familiar and potentially susceptible tomeaningful groupings (Baddeley et al., 1998). Incontrast, the poor familiarity of all participants

Figure 2.Mean percentage of correct responses for forward and backward recall by high-spatial-ability (HSA) and low-spatial-ability (LSA)

groups. Error bars represent standard errors.



Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

with spatial tasks could produce a more obviousadvantage resulting from practice (which was par-ticularly high in the present study), with its poss-ible implications such as increased taskfamiliarity, confidence, and development of strat-egies like clustering locations; these are variablesthat typically strongly affect memory for spatialconfigurations.

In a recent study (Rowe, Hasher, &Tourcotte, in press)—not known to us whenthe present study was conducted—a memoryadvantage due to the descending presentationformat was found in the elderly also for theCorsi blocks task, but an opposite advantage ofthe ascending format was found in young indi-viduals. The authors concluded that two factorscan contribute to performance on working-memory span tasks—that is, practice andinterference. The benefits of practice lead tooverestimation of span size for young adults,while the costs of interference lead to span sizeunderestimation for older adults. This hypothesismight also apply to our low-spatial-ability indi-viduals, poor in spatial spans but with higherability—compared with elderly people—forrapid development of effective strategies for per-forming the task. This practice advantage mayhave compensated the susceptibility to PI dueto the increase in the number of already highlyactivated locations, at least in our participants,including people with visuospatial weaknesses,thus producing a similar performance for thetwo presentation formats. On the contrary, prac-tice was not sufficient for compensating thedifficulties met by the elderly group studied byRowe et al. (in press).

In conclusion, the present research showed thatthe Corsi blocks task is not prone to the effects ofdirection of recall and presentation format in thesame way and to the same extent as verbalworking-memory span tasks; rather, HSA andLSA differed in performance on the backwardCorsi blocks task, demonstrating that forwardand backward recall directions involved differentspatial processes. Moreover, LSA did not benefitfrom the descending condition of the VSWMspan task, revealing that LSA is not equivalent

to other groups of low-span individuals—forexample, the elderly.

Original manuscript received 22 June 2007

Accepted revision received 3 October 2007

First published online 17 December 2007

REFERENCES

Avons, S. E., & Mason, A. (1999). Effects of visualsimilarity on serial report and item recognition.Quarterly Journal of Experimental Psychology, 52A,217–240.

Awh, E., Jonides, J., & Reuter-Lorenz, P. A. (1998).Rehearsal in spatial working memory. Journal of

Experimental Psychology: Human Perception and

Performance, 24, 780–790.Baddeley, A. (1986). Working memory. Oxford, UK:

Oxford University Press.Baddeley, A., Gathercole, S. E., & Papagno, C. (1998).

The phonological loop as a language learning device.Psychological Review, 105, 158–173.

Bowles, R. P., & Salthouse, T. A. (2003). Assessing theage-related effects of proactive interference onworking memory tasks using the Rasch model.Psychology and Aging, 18, 608–615.

Bunting, M. (2006). Proactive interference and itemsimilarity in working memory. Journal of

Experimental Psychology: Learning, Memory, and

Cognition, 32, 183–196.Cornoldi, C., Rizzo, A., & Pra Baldi, A. (1991). Prove

avanzate MT di comprensione nella lettura.

[Advanced MT reading comprehension tests].Firenze, Italy: Organizzazioni Speciali.

Daneman, M., & Carpenter, P. A. (1980). Individualdifferences in working memory and reading.Journal of Verbal Learning and Verbal Behavior, 19,450–466.

Fischer, M. H. (2001). Probing spatial working memorywith the Corsi blocks task. Brain and Cognition, 45,143–154.

FitzGerard, P., & Broadbent, D. E. (1985). Memory forattended and unattended visual stimuli. Quarterly

Journal of Experimental Psychology, 37A, 339–365.Kane, M. J., Hambrick, D. Z., Tuholski, S. W.,

Wilhelm, O., Payne, T. W., & Engle, R. W.(2004). The generality of working memory capacity:A latent variable approach to verbal and visuospatial



Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

memory span and reasoning. Journal of ExperimentalPsychology: General, 133, 189–217.

Kemps, E. (2001). Complexity effect in visuo-spatialworking memory: Implications for the role of long-term memory. Memory, 9, 13–27.

Kosslyn, S. M. (1990). Mental imagery. InS. M. Kosslyn & D. N. Osherson (Eds.), Visual cog-nition and action: An invitation to cognitive science

(Vol. 2; pp. 73–97). Cambridge, MA: MIT Press.Lustig, C., May, C. P., & Hasher, L. (2001). Working

memory span and the role of proactive interference.Journal of Experimental Psychology: General, 130,199–207.

Mammarella, I. C., & Cornoldi, C. (2005a). Difficultiesin the control of irrelevant visuospatial informationin children with visuospatial learning disabilities.Acta Psychologica, 118, 211–228.

Mammarella, I. C., & Cornoldi, C. (2005b). Sequenceand space. The critical role of backward spatialspan in the working memory deficit ofvisuospatial learning disabled children. Cognitive

Neuropsychology, 22, 1055–1068.May, C. P., Hasher, L., & Kane, M. J. (1999). The role

of interference in memory span. Memory and

Cognition, 27, 759–767.

Rowe, G., Hasher, L., & Tourcotte, J. (in press). Agedifferences in visuospatial working memory.Psychology and Aging.

Salway, A. F. S., & Logie, R. H. (1995). Visuospatialworking memory, movement control and executivedemands. British Journal of Psychology, 86, 253–269.

Smyth, M. M., & Scholey, K. A. (1992). Determiningspatial span: The role of movement time and articu-lation rate. Quarterly Journal of Experimental

Psychology, 45A, 479–501.Vanderberg, S. G., & Kuse, A. R. (1978). Mental

rotations, a group test of three-dimensional spatialvisualization.Perception andMotor Skills, 47, 599–604.

Vandierendonck, A., Kemps, E., Fastame, M. C., &Szmalec, A. (2004). Working memory componentsof the Corsi block task. British Journal of Psychology,95, 57–79.

Vandierendonck, A., & Szmalec, A. (2005). An asym-metry in the visuo-spatial demands of forward andbackward recall in the Corsi blocks task.Imagination, Cognition and Personality, 23, 225–231.

Wilde, N., & Strauss, E. (2003). Functional equivalenceof WAIS-III/WMS-III digit and spatial span,under forward and backward recall conditions.Clinical Neuropsychologist, 16, 322–330.



Dow

nloa

ded

by [

Lyc

omin

g C

olle

ge]

at 2

2:24

10

Oct

ober

201

4

a comparison of backward and forward spatial spans

Documents