Assessing negative priming by attended distractors in a ... · 1131 Braz J Med Biol Res 37(8) 2004...

1131

Braz J Med Biol Res 37(8) 2004

Assessing negative priming by attended distractorsBrazilian Journal of Medical and Biological Research (2004) 37: 1131-1153ISSN 0100-879X

Assessing negative priming by attendeddistractors in a paper-and-pencil task

Departamento de Psicologia e Educação, Faculdade de Filosofia,Ciências e Letras de Ribeirão Preto, Universidade de São Paulo,Ribeirão Preto, SP, Brasil

F.M. Rosin

Abstract

The paper-and-pencil digit-comparison task for assessing nega-tive priming (NP) was introduced, using a referent-size-selectionprocedure that was demonstrated to enhance the effect. NP isindicated by slower responses to recently ignored items, and pro-posed within the clinical-experimental framework as a major cogni-tive index of active suppression of distracting information, criticalto executive functioning. The digit-comparison task requires cir-cling digits of a list with digit-asterisk pairs (a baseline measure fordigit-selection), and the larger of two digits in each pair of theunrelated (with different digits in successive digit-pairs) and re-lated lists (in which the smaller digit subsequently became a tar-get). A total of 56 students (18-38 years) participated in two experi-ments that explored practice effects across lists and demonstratedreliable NP, i.e., slowing to complete the related list relative to theunrelated list, (F(2, 44) = 52.42, P < 0.0001). A 3rd experimentexamined age-related effects. In the paper-and-pencil digit-com-parison task, NP was reliable for the younger (N = 8, 18-24 years)and middle-aged adults (N = 8, 31-54 years), but absent for theolder group (N = 8, 68-77 years). NP was also reduced with aging ina computer-implemented digit-comparison task, and preserved in atask typically used to test location-specific NP, accounting for thedissociation between identity- and spatial-based suppression ofdistractors (Rao R(3, 12) = 16.02, P < 0.0002). Since the paper-and-pencil digit-comparison task can be administered easily, it can beuseful for neuropsychologists seeking practical measures of NPthat do not require cumbersome technical equipment.

CorrespondenceCorrespondence

F.M. Rosin

Instituto de Investigaciones

Facultad de Psicología

Universidad de Buenos Aires

Av. Independencia, 3065, 3º Piso

(C1225AAY) Buenos Aires

Argentina

E-mail: [email protected]

Research supported by a CAPES/SPU

fellowship.

Publication supported by FAPESP.

Part of a Doctoral thesis presented

by F.M. Rosin to the Departamento

de Psicologia e Educação, Faculdade

de Filosofia, Ciências e Letras de

Ribeirão Preto, Universidade de São

Paulo, Ribeirão Preto, SP, Brazil.

Available on-line: [http://

www.teses.usp.br/teses/disponiveis/59/59134/tde-07052002-112218].

Presented at the

XXVII International Congress of

Psychology, Stockholm, Sweeden,

July 23-28, 2000, and the XIX

Brazilian Congress of Neurology,

Salvador, BA, Brazil, October 7-12,

2000.

Received February 19, 2003

Accepted April 7, 2004

Key words• Negative priming• Selective attention• Paper-and-pencil tasks• Aging• Executive function

Introduction

During the last few decades, there hasbeen an increasing interest in the inhibitorycontrol of the flow of visual processes intoactions, which may be evoked even whenthere is little or no conscious intention to act.Measures of the capacity to resist a pre-potent response of the moment are difficultto be estimated directly from performance in

normal human subjects, and therefore it hasbeen necessary to develop indirect measuresof performance such as priming procedures.An increasing attempt has been made todevelop pencil-and-paper neuropsychologi-cal tests which do not require cumbersometechnical equipment, in order to simplify theirday-to-day clinical use and to permit stan-dardization (1).

The objective of the present study was to

1132


F.M. Rosin

introduce the digit-comparison task, a newpaper-and-pencil task from which the nega-tive priming (NP) effect has been reliablyindexed in a rapid and easy way, and todocument its suitability for detecting indi-vidual differences within the clinical-experi-mental neuropsychology setting.

There is evidence that NP may reflect amechanism of control, crucial for coherentbehavior, expressed when directing responsestoward a target stimulus in the presence of aconcurrent distractor which competes forthe control of action. NP, also known asdistractor suppression effect, has been typi-cally observed in selective attention tasksthat require selecting and responding to tar-gets while disregarding distracting informa-tion (2). The statement that attention theo-rists have argued about is that there exists alimitation of processing capacity, so thatselection of and responses to the relevantaspects of the environment may entail theactive ignoring of that which does not servethe current goals. The implication is that thisincredibly efficient processing of selectiveattention produces a normal slowing in timeor is more error-prone in selecting an item,when it was selected against in the precedingtrial. This effect was called NP, and it hasbeen defined as a significant time cost inresponding to a target that shared featureswith the distractor recently ignored in theprevious trial, when compared with responsetimes to a new target (3). For example,people usually take more time to report theidentity or location of an item when in thepreceding trial they were supposed to ignorean identical distractor or a distractor pre-sented at the same site, respectively.

Slowing for recently ignored targets inpriming tasks generally reaches modest mag-nitude, with values of less than 20 ms (4).Therefore, computerized procedures are usu-ally implemented to measure NP more pre-cisely. It was demonstrated that tasks requir-ing overt attention to distractors providedenhanced measures of NP (5). Thus, for the

paper-and-pencil digit-comparison task, areferent size-selection procedure was cho-sen, which required attending to both targetand distractors, i.e., the subject was asked tocompare digit pairs and select the highestdigit. Because of both simplicity in its mate-rial and procedure and the larger magnitudeof NP that is obtained by means of its referentsize-selection procedure, a task such as thepaper-and-pencil digit-comparison task wouldbe of interest to neuropsychologists seekingpractical ways to measure NP, which hasevolved into an important research area. NPhas been regarded as the best available indexof active suppression of information irrel-evant to the current goals in selective atten-tion tasks and has been proposed for thedetection of syndromes that involve cogni-tive impairment (6).

NP has been broadly investigated be-cause of its relevance to theories of selectiveattention and its promising clinical applicabil-ity, since various clinical and developmentalpopulations with increased interference fromintrusive irrelevant information often fail topresent NP. Absent or diminished NP wasreported from neuropsychological studieswith people who are supposed to showimpaired inhibitory control in selective atten-tion, although not all were more susceptibleto interference from a distractor, e.g., adultswith schizophrenia (7), focal cerebral dam-age (8,9), Alzheimer’s disease (10), andHuntington’s diseases (11). Developmentalstage studies also demonstrated that youngchildren and older adults often do not demon-strate reliable NP in selection tasks thatrequire a response to the identity of a target(6,12-15). The findings were interpreted asa deficiency in inhibitory attentional mechan-isms and were taken as evidence consistentwith the view that NP is a marker of theresidual inhibition associated with the previ-ous irrelevant stimuli (16). According to thistheory, response times to the recently ig-nored targets should be delayed by suppress-ing their internal representations, or by block-

1133


Assessing negative priming by attended distractors

ing from potential effectors, suggesting thatdistractors are processed beyond physicalfeatures.

Much of the literature is consistent withselective attention processes underlying NP,but evidence from several studies challengesthis theory and alternative and/or comple-mentary theoretical accounts for NP involv-ing memory processing have been proposed(13,17). Memory-based models explain thatNP may be caused by a conflict in theretrieval episode, such that the current stimuluscues the retrieval of most recent processinginstances involving that stimulus and con-flicts with the current response, rather thanbeing directly caused by a process of activesuppression of the distractor in the precedingtrial (18). The memory-based construct alsoshould predict reduced NP for older adults,because they should be less likely to engagein retrieval of previous episodes of the stimuli.

In contrast with the hypothesis of a pureinhibition mechanism applied to distractorsduring the prime trial causing NP, an episodicretrieval mechanism and item informationmismatching might predict a positive relationrather than an inverse relation between NPand interference. In fact, a number of studieshave indeed found that conditions that in-duced a greater magnitude of interferencealso induced a greater magnitude of NP, eventhough these two distractor processing indi-ces can be dissociated in the individual per-formance (8; for review, see Ref. 4). Thus,the efficiency in selecting (lesser interfer-ence) by increased inhibition of a distractormight be related to higher NP in the subse-quent trial, supporting an inhibitory effect,whereas a positive relation between higherinterference by more in-depth processing ofthe distractor and increased NP could beexplained by other mechanisms accountingfor NP. According to an episodic-retrievaltheory, stronger distractors may be morelikely to be encoded as “to be ignored”, andconsequently they would be less likely to beselected in the subsequent trial; hence inter-

ference and NP might correlate positively.However, a positive relation between NP andinterference does not exclude the inhibitoryexplanation. From the inhibitory perspective,inhibition is usually viewed as reactive to thesource of interference; therefore, a positivecorrelation between NP and interference maybe explained by a reactive inhibition mechan-ism implemented when distractor interfer-ence is high, so that distractors intrude morein the control of action. On the other hand,when interference is treated as an individual-difference variable, individuals who are moresusceptible to interference are predicted toshow diminished NP because they may beless able either to inhibit the distractor (suchas the inhibitory account proposes) or toencode the distractor and retrieve the prim-ing episode (according to the episodic-re-trieval account). Furthermore, NP may in-volve the operation of both attention-basedand memory-based mechanisms (19,20)during selection of the target from thedistractor at encoding and during memoryretrieval at analysis of the subsequent selec-tion, and a variety of conditions in the exper-imental design may result in NP tasks thatfavor one type of processing or the other.

Although the review of the empiricalevidence broadens the scope of the presentstudy, the relation between distractor inter-ference and NP and its possible dissociationunder some circumstances have been ofspecial relevance for neuropsychologicalstudies over the last decades. NP may con-stitute an executive measure that interfacesperception and action depending upon taskdemands and directs attention resources to-ward goal-relevant activity, which discrimi-nates individuals who do not seem to differ intheir pure measures of efficiency of selectiveattention. Thus, NP may be an expression ofthe executive role of attention in suppressingan unattended object to prevent responseconflict and in providing coherence betweenthe conscious experience and action streams(19).

1134


F.M. Rosin

NP has been shown to be reliable under avery wide range of stimuli and proceduresand different response modalities (21), andacross practice (22). NP has been reportedfrom a variety of computer-implementedtasks such as color naming tasks (23), iden-tity-based tasks which require attending anitem specified by a perceptual cue (e.g.,symbol identity, color, location) while ignor-ing a distractor, location-based tasks whichrequire selecting the target on the basis of aphysical attribute but reporting the locationof the target (24), same-different tasks thatask participants to match and judge items asthe same or different (25), and size-selectiontasks that require attending to both target anddistractor by means of a size comparisonbefore selection and response can occur(26).

The referent size-selection procedure hasbeen shown to enhance NP. This procedurerequires subjects to compare the semanticreferents of two item names, e.g., which oftwo nouns weighed the most or which oftwo animal names corresponded to the larg-est animal (5,27), and has evidenced that NPmay also occur when the distractor is pro-cessed deeply and reaches awareness. Al-though the status of a distracting stimulusrelative to conscious awareness has not beenexplicitly described in the inhibition-basedmodel, lack of awareness of the prime’sidentity has been traditionally one of themarkers of NP (3,16). The finding of NPfrom performance in these decision-typetasks, which favor post-lexical processing inthe prime display with overt access to themeaning of both target and distractor, hasbeen considered a challenge to the theory bywhich NP should result just from earlierinhibition mechanisms in selective attention-preventing distractors to attain consciousprocessing. Slowing in response times maybe explained by a conflict in the retrievalepisode, as a result of ambiguity in informa-tion about the target, because the item previ-ously encoded as a distractor is coded as

relevant in the current trial and requires aresponse. The memory-based model pre-dicts more NP when prime selection condi-tions afford more in-depth processing be-cause the distractor attributes (e.g., identity,spatial location, their status as “relevant” or“irrelevant”, and the response “respond” or“do not respond” they require) are likely to beretrievable when the distractor appears as atarget in the subsequent trial.

It appears that NP from tasks whoseresponses are based on identity and locationattributes may rely on separate mechanisms(19). Several developmental studies havereported that NP failed to occur in tasksrequiring suppression of the identity ofdistractors in older adults and children,whereas location suppression was sparedwith age (12,14). These reports from cogni-tive psychology research were taken as con-vergent evidence in favor of the existence ofat least two separate inhibitory systems in-volved in identity and location suppression ofdistractors from visual selective attentiontasks, which may be selectively affectedacross the life span. The hypothesis of a dualnature of the visual system has derived fromneurophysiological studies in which two spe-cialized cortical visual pathways that reachthe frontal cortex were proposed. The ven-tral or occipitotemporal pathway seems to bespecialized in object shape and in attenuatingresponses to stimuli previously presented,and the dorsal or occipitoparietal pathwaywas proposed for processing of spatial loca-tion (28).

NP was initially indexed from the totaltime to name ink-colors in a list of Stroopcolor-words whose ink-color correspondedto the preceding word that had been recentlyignored (2). Experiments recording vocalnaming (23) and key-press (29) responselatencies to randomized and individually pre-sented Stroop words replicated the effect ofdistractor suppression, which allowed disre-garding the idea that a potential artifact of themassed-list procedure had been the cause of

1135



slowing in responses. Although the readinglist paradigm was criticized because of ex-perimenter effects and insensitivity factors,such as manual operation of a stopwatch andsubjects’ movements while reading, NP wasalso reliable in list-reading tasks that requiredparticipants to name lists of letters specifiedby a color cue in letter pairs (30). By includ-ing lists of letters alone or paired with unre-lated or related distractor letters, increasedinterference from a concurrent distractorand reliable NP were reported in older adultsand adults with Alzheimer’s disease (31).

The NP effect on the massed-list proce-dure from the Stroop color-word task seemsto be robust. However, the Stroop task istraditionally implemented in the clinical set-ting to assess the effect of interference withcolor naming by the conflicting word,whereas the NP effect derived from theadditional cost in naming colors that had beenpreviously ignored, e.g., to name the inkcolor ‘blue’ when the previous ignored wordhad been ‘blue’, is usually limited to experi-mental frameworks. While the interferenceindex is good, NP appears to be unreliableand less is known about its psychometricproperties (32). As explained above, therehas been much debate over the causes of NP,and therefore this effect has been extensivelystudied in order to clarify its construct valid-ity and to specify which mechanisms may beinvolved in the effect, in order to allowcognitive psychologists to estimate how theimpairment in NP could contribute to thecognitive deficits of several clinical groupsand to construct reliable NP indices. This istrue for all attempts to implement any NPtask as a diagnostic tool, given that NP isusually a small effect which is not reliablyobserved in the performance of all partici-pants. When the list procedure of the Strooptask is implemented, there is reason to sus-pect that practice effects across the color-word lists may affect NP scores. In fact,repeated exposure to the Stroop task hasbeen shown to affect the magnitude of the

interference by conflicting words with colornaming, especially among older subjects.Once the effect of practice was controlled byimplementing an additional unrelated color-word list, reliable NP scores could be ob-tained from the reading-list procedure of theStroop task in younger and older people(Rosin FM, unpublished data). This proce-dure is also relatively fast and inexpensive,but it relies on reading ability, so that themeasures of performance of people of loweducational level and children are usuallyaffected. Also, it cannot be administered tocolor-blind people.

In the present study, a digit-comparisontask was developed in a paper-and-pencilversion in order to obtain a reliable measureof NP in a practical way. This task involvesthe referent size-selection procedure (i.e.,subject is asked to compare pairs of itemsand select the target specified by a semanticcue), ensuring that both target and distractorare attended, since this procedure has beendemonstrated to enhance NP. The selectionof digits as stimuli obeyed to the fact that thetask was developed with the aim to be admin-istered in a clinical setting. In fact, at the timea simultaneous study with patients with ce-rebral damage was being carried out. Someof them were non-readers or with low degreeof formal education, and from heteroge-neous cultural background. So, it was moredifficult to employ stimuli with highest se-mantic weight, and letter stimuli were notsuitable for the assessment of the wholesample of patients. By contrast, all patientsknew digits and were able to perform thepaper-and-pencil digit-comparison task. Inthis task, subjects are asked to circle thetarget, which reduces somewhat other mo-tor activities that could affect the measure ina reading-list procedure and enables moreaccurate assessment of errors. The stimuluspairs are digit-asterisk and digit pairs. Thetask requires circling the digit of the digit-asterisk or the greater of two digits in a seriesof stimulus pairs listed on a sheet of paper.

1136


F.M. Rosin

Digit stimuli were selected in order to de-velop a suitable task for the clinical setting,ensuring that people with different educa-tional backgrounds would know the stimuliand could manage them. There are threeseparate lists: the first is the asterisk-digitcondition (asterisk list), in which the distractoris a non-digit stimulus and remains the samethroughout the list; the second is the unre-lated digit-pair condition (unrelated list) inwhich distractor and target digits, in succes-sive digit pairs, are never the same; and thethird list is the related digit-pair condition(related list), in which in some pairs thedistractor in one digit pair is the target in thenext pair. Figure 1 shows examples of thestimuli. Marking times are recorded for eachlist. The asterisk list serves as a baselinemeasure to estimate the additional time takenfor selection between the digits. The slowingin marking time for the related list relative tothe unrelated list condition is an index of NP.

The first experiment was run as a pilotdemonstration to assess whether the paper-and-pencil digit-comparison task indexed NPand whether the NP measure was affectedby the order of administration of the lists.The second experiment included the asterisk

list, which provided a baseline measure ofspeed of performance in a simpler condition,representing a control measure to estimatethe difference in performance when com-parison between conflicting digits was re-quired. In addition, performing the asteriskcondition may have included initial practiceeffects on marking times. This experimenttested the effect of practice on the time tocomplete the unrelated and related lists byadding an additional unrelated list after theunrelated and related lists were completed.

A third experiment was carried out todetermine whether aging affects NP as as-sessed by the paper-and-pencil digit-com-parison task, and included a computer-imple-mented version of the digit-comparison taskwith recording of vocal naming responselatencies to randomized and individually pre-sented stimulus pairs. This experiment alsoincluded a target location-based task to as-sess the possibility that older adults might beable to show NP when the location of thetarget was the focus of the response, but notwhen its identity was the basis of selectionand response.

Experiment 1

Method

Participants. Thirty-three university stu-dents, 17 women and 16 men (range = 18-31years, mean = 20.52), participated in thisstudy. All participants had normal or cor-rected-to-normal near visual acuity as meas-ured with an ophthalmic vision tester (Stan-dard number 1, Bausch & Lomb Occupa-tional Vision Tests, New York, NY, USA).All participants gave informed consent toparticipate in the study which was approvedby the Ethics Committee of the Hospital dasClínicas, Faculdade de Medicina de RibeirãoPreto, USP (Project HCRP No. 2825/98).

Experimental design. Participants per-formed individually the paper-and-pencil digit-comparison task, which required to use a felt

Figure 1. Examples of the stimuli used in the digit-comparison task. In the digit-asteriskcondition, the distractor is a non-digit stimulus and remains the same throughout the list.In the unrelated condition, distractor and target digits vary in successive digit-pairs. In therelated condition, the distractor digit in a pair is the same as the next target. The asteriskand unrelated lists were preceded by a sample sheet to ensure that subjects hadunderstood the instructions. The time to complete each list and errors were recorded. Theasterisk condition was considered to be a baseline measure of performance. The negativepriming effect was indexed from the difference in marking times between the unrelatedlist and the related list.

1137



pen to circle the digit of digit-asterisk pairs orthe greater digit of two digits in a series ofstimulus pairs listed on a sheet of paper.Separate lists (unrelated and related lists)contained 54 stimulus pairs each, whichwere distributed across three columns laidout on A-4 size sheets (see example of stimulifor unrelated and related lists in Figure 1).Stimuli in a pair were spaced 1 cm apart andpairs down the list were spaced 1 cm apart.Columns were spaced approximately 4.7 cmapart. Digits were 4 mm high x 2 mm widebold font characters. Five disyllabic (for thePortuguese language) digits were used for allconditions, i.e., 4, 5, 7, 8, and 9. Target anddistractor locations, i.e., to the left or right ina pair, were evenly balanced and randomlydistributed. The digit 4 appeared as a distractormore frequently than other digits and the digit9 appeared as a target more frequently thanother digits. Target and distractor locationsacross pairs of successive trials were equallydistributed across lists.

Participants completed a sample sheet toensure that they had understood the instruc-tions and then, they completed the unrelatedand related lists. The unrelated list containedunrelated digit pairs in which target anddistractor digits, in successive pairs, werenever the same. The related list consisted of 20unrelated digit-pairs and 34 related digit-pairs,i.e., the distractor in one digit pair was thetarget in the next digit pair. Eighteen partici-pants performed the unrelated list first and 15participants performed the related list first.Eight participants were asked to complete anadditional unrelated list at the end of the task toassess additional practice effects.

Participants were shown the sample andtold to circle the greater digit of the digit pairsfor the lists, as fast as possible. Subjectswere also told that if they made a mistakethey should correct it by crossing out thewrong digit and circling the correct digit.A stopwatch was used to measure markingtimes. For each list, the experimenter said“Ready”... “Go”, and started the stopwatch

on the word “Go”, stopping it as the subjectcircled the last digit. To avoid possible ex-perimenter effects, a neutral experimenterthat ignored the order of the list conditionsrecorded the marking times. Errors and omis-sions were also recorded. The number oferrors showed to be minimal in healthy youngpeople and they tend to correct them on thego. However, a simultaneous study revealedan increase in errors and omissions in theperformance of some patients with cerebraldamage. Since speed of performance mayhide a differential cost in accuracy acrosssubjects, a formula that takes into accountomissions and the commission of errors thatwere not spontaneously corrected could beuseful to obtain a more accurate measure ofperformance when comparisons between-groups are required. For that purpose, it wasestimated the corrected marking time (cMT)for all list conditions, which was computedas follows: cMT = (marking time/total ofpairs marked) x (non-corrected errors +omissions) + marking time. In the presentstudy, conducted on healthy young and olderpeople, all analyses carried out on the cMTsyielded similar pattern of significant effectsto the analyses conducted on the markingtimes without the inclusion of errors, so thislast and most simple measure was used.

In order to assess NP, i.e., the delay tocomplete the related list relative to the unre-lated list, the marking times for each subjecton lists in each condition were submitted toanalysis of variance (ANOVA), with re-peated measures for the list conditions.Duncan’s multiple range test was used forpost hoc comparisons. Wilcoxon’s matched-pair test was used for within-subject com-parison of mean errors and Mann-Whitney’sU-test was used for between-subject com-parisons. An alpha level of 0.05 was used forall statistical tests.

Results and Discussion

Figure 2 shows the mean marking times

1138


F.M. Rosin

for the unrelated and related lists accordingto the order of presentation of the lists. A 2(list condition: unrelated vs related) x 2 (orderof list) ANOVA revealed that subjects wereslower in completing the related list, withdigit targets that had been previously ignoredas distractors, than they were in completingthe list with unrelated digit pairs, F(1, 31) =15.71, P = 0.0004. These higher markingtimes for the related condition indicated reli-able NP. The main effect of list order wasabsent, F(1, 31) = 0.36, P < 0.551, but theinteraction between order and list was sig-nificant, F(1, 31) = 6.16, P < 0.019. Post hoccomparisons revealed that the marking timesfor the related list were significantly slowerthan for the unrelated list, but only in thegroup that had performed first the unrelatedlist test, as can be seen in Figure 2. However,the time taken to complete the related listpresented first was equivalent to that ob-served when the related list was completedafter the unrelated one. Thus, the order of listpresentation seemed to affect just the unre-lated condition, with higher marking timeswhen the unrelated list test was performedafter the related list.

ANOVA conducted on the marking timesof the 8 participants that completed theunrelated - related - additional unrelated lists,revealed that list condition was significant,F(2, 14) = 9.74, P = 0.002. A reliable slowingfor the related condition was observed rela-tive to both first and additional unrelated lists,with no significant difference between them.The profile in marking times (in seconds)was related (M = 45.15, SD = 8.56) >unrelated (M = 40.58, SD = 8.06) = addi-

tional unrelated (M = 42.61, SD = 9.09), asindicated by post hoc multiple comparisons.

Analyses of errors did not yield significantdifferences across list conditions or groups.The error means, 0.09 (SD = 0.38) for theunrelated list and 0.18 (SD = 0.46) for therelated list, were not significantly different, T= 3.00, Z = 1.21, P = 0.225. Groups did notdiffer from one another in errors for theunrelated list, U = 133, Z = -0.07, P = 0.942,or related list, U = 129, Z = -0.20, P = 0.842.

Finally, a two-way ANOVA (2 sex x 2 listcondition) revealed that men were overtlyfaster than women, F(1, 31) = 8.06, P <0.008; both groups were slower for therelated condition, F(1, 31) = 14.77, P =0.0006, with no interaction between sex andlist condition, F(1, 31) = 0.01, P = 0.943. NPwas reliable for both the groups of men, F(1,15) = 7.14, P = 0.017, and women, F(1, 16)= 7.58, P = 0.014. Since women showed anoverall slowing in response times, ANOVAwas conducted on the proportional NP meas-ure, i.e., [(marking time for related list -minus - marking time for unrelated list)/marking time for unrelated list] which did notyield significant difference in proportionalNP between sex groups, F(1, 31) = 0.09, P= 0.767.

The digit-comparison task seemed to re-flect NP, as indicated by slower times tocomplete the related list, with recently ig-nored digits, compared to the time taken tocomplete the unrelated list with differentdigits in successive digit-pairs. The order ofthe list conditions affected the NP index, i.e.,the difference between the unrelated andrelated conditions was absent when the re-lated list was completed first. However, thiseffect may not be explained by an improve-ment in time related to repeated performanceof lists, since the reliable slowing for therelated list occurred just when this wascompleted after practice on the unrelated list.Furthermore, differences in marking timesfor the related condition were not detectedbetween groups with inverted order of the

Mar

king

tim

e (s

)

48

46

44

42

40

3836

34U list R list

U list firstR list first

Figure 2. Effect of the order ofadministration of the lists on thedifference in marking time be-tween the unrelated (U list) andrelated (R list) conditions, in Ex-periment 1. Data are reportedas the mean ± SEM. Eighteenparticipants performed the U listfirst, and 15 participants per-formed the R list first. Partici-pants were significantly slowerin completing the related listwhen they had performed theunrelated list before, ANOVA (2list x 2 order), F(1, 31) = 6.16, P< 0.019, followed by Duncan’smultiple range test *(P < 0.01).

*

1139



lists (see Figure 1). This result addressed theconcern of whether practice and possibleattentional strategies that could develop acrosstrials, also including motor practice (and/orfatigue), could have influenced performanceindices and consequently affected the NPmeasure. For the group of 8 participants thatcompleted the final unrelated list there was atrend towards an increase in marking timefor the unrelated condition after a prior expo-sure to the related list, although the differ-ence did not reach significance, and a reliableslowing for the related condition was ob-served relative to both first and additionalunrelated lists.

Therefore, Experiment 2 was carried outto reassess the NP finding by means of thepaper-and-pencil digit-comparison task andextended the question of whether practicewas affecting differences in marking timesacross list conditions by requiring all partici-pants to complete an additional unrelated listafter the related condition. Additionally, Ex-periment 2 also included the asterisk list withnon-conflicting distractors, which served asa baseline measure of speed of performanceand to estimate the additional cost in compar-ing and selecting between two digits. So, theconditions, in the order in which they wereadministered to the participants, were theasterisk list, the unrelated list, the related list,and finally, the additional unrelated list.

Experiment 2

Method

Participants. Twenty-three university stu-dents, 8 women and 15 men (range = 19-38years, mean = 23.48) with normal or cor-rected-to-normal near visual acuity, partici-pated in this study. None had participated inExperiment 1. All participants gave informedconsent. The research was approved by theEthics Committee as cited in Experiment 1.

Material and procedure. Stimuli and pro-cedure were the same as described in detail

in Experiment 1. The procedure was other-wise identical to that described above, butincluded the asterisk condition which re-quired participants to circle the digit of thedigit-asterisk pairs. For the unrelated andrelated lists the asterisk list also contained 54stimulus pairs distributed across three col-umns laid out on A-4 size sheets (See Figure1). For the digit stimuli, asterisks were alsobold font characters, but subtending 2 x 2mm. Both asterisk and unrelated lists werepreceded by a sample sheet containing eightstimulus pairs to ensure that the subject hadreally understood the instruction. Partici-pants were told to circle down the digit of thedigit-asterisk pairs for the asterisk list, andthe greater digit of the digit pairs for thefollowing lists, as fast as possible. The unre-lated and related lists were identical to thosedescribed in Experiment 1. The asterisk listwas the first one and contained digit-asteriskpairs. The second list was the unrelatedcondition, in which target and distractordigits, in successive digit pairs, were neverthe same. The third list was the relatedcondition, which contained related digit-pairs,i.e., the distractor in one digit pair was thetarget in the next digit pair. Finally, all partici-pants completed an additional unrelated list.

As described for Experiment 1, the basicdependent measure was each subject’s mark-ing time on lists in each condition. Primarily,NP was evaluated by comparing the markingtime in the unrelated condition to the markingtime in the related condition. Furthermore,the additional time related to conflicting com-parison between digits was also estimated bycomparing the marking time in the unrelatedcondition to the marking time in the asteriskcondition.


Table 1 summarizes the data, responsetimes and errors. Significant differences inmarking times across asterisk, unrelated,and related conditions were observed, with

1140


F.M. Rosin

reliable NP. ANOVA was used for compar-ison of the marking times in the appropriatelist conditions (asterisk vs unrelated vs re-lated), which revealed that the effect ofcondition was significant, F(2, 44) = 52.42,P < 0.0001. As expected, post hoc analysisrevealed that participants took more time incompleting the unrelated list than the asterisklist. NP was reliable, with slower markingtimes for the related list than for the unrelatedlist.

To assess whether practice affected thedifferences in time taken to complete the listconditions, the marking times were sub-jected to ANOVA with repeated measures oncondition (unrelated vs related vs additionalunrelated). The effect of condition was sig-nificant, F(2, 44) = 15.09, P < 0.0001, withlonger marking times for the related listindicating reliable NP. Again, post hoc com-parisons showed that subjects were slowerin completing the related list than in complet-ing each of the two, first and additional,unrelated lists. Furthermore, the markingtimes did not differ significantly between thetwo unrelated lists. Planned comparisonsalso showed a non-significant differencebetween the unrelated lists, F(1, 22) = 2.33,P < 0.141. Exam of errors showed that theydid not differ across conditions. FriedmanANOVA indicated that errors were neithersignificantly different across asterisk vs un-related vs related lists, χ2(2, 23) = 1.00, P <

0.607, nor different across unrelated vs re-lated vs additional unrelated lists, χ2(2, 23) =0.50, P < 0.779.

The results replicated those obtained inExperiment 1, showing reliable NP for thepaper-and-pencil digit-comparison task, ex-pressed as a significant delay in completingthe related digit-pair list relative to the unre-lated list. The results of Experiment 2 do notsupport the notion that NP should be attrib-uted to practice biases or fatigue effects. Thesuggestion that such delay may have reliedupon expectancy biases for successivestimuli, which may have developed afterrepeated exposure to successive ignored-repeated stimuli of the related list, was ad-dressed in this study by completing twounrelated lists, one before and the other afterperforming in the related condition. In bothunrelated lists the marking times were fasterthan the time taken to complete the relatedlist, with no significant differences betweenthem. These results disagree with the claimthat slowing in a related digit-pair list may beexplained by expectancy biases developedafter repeated exposure to recently ignoreddigit stimuli.

Additionally, this study included an aster-isk list containing non-conflicting distractors,a less demanding list than the digit-distractorlist. The asterisk list serves to obtain a base-line marking time, also permitting between-subject comparisons of proportional addi-tional cost in comparing and selecting be-tween two digits, as done in Experiment 3.

Experiment 3

This study included younger and olderadults in order to determine whether NPmeasurement, as assessed by means of thepaper-and-pencil digit-comparison task, wasaffected by aging. Several findings providedevidence that older adults do not show con-sistent slowing for recently ignored itemsacross different NP tasks and under the sameexperimental circumstances as do younger

Table 1. Marking time and errors for the paper-and-pencil digit-comparison task inExperiment 2.

Digit-asterisk Digit-pairs

Unrelated Related Additionalunrelated

Marking time 34.61 ± 8.45 42.64 ± 10.01 46.26 ± 9.95 43.68 ± 10.94Errors 0.00 ± 0.00 0.04 ± 0.21 0.04 ± 0.21 0.22 ± 0.85Negative 9.22 ± 8.33priming (%)

Data are reported as means ± SD, N = 23, in seconds for marking time, and numberof errors.

1141



adults (6,13). Although contradictory exper-imental outcomes also exist (10,31), olderadults seemed not to demonstrate reliable NPprimarily in identity NP tasks, which requirea response to the identity of a target (13,14).In contrast, NP seemed to be preserved inaging when the task required a response tothe location of the target (12). Strongerevidence of a dissociation between identityand spatial NP was obtained in the presentexperiment, in which younger, middle-agedand older adults were tested on both identity-and location-based NP tasks.

Although the present version of the digit-comparison task could combine some as-pects of the spatial component with identifi-cation of the target, since the target appearssometimes at the same site as the disregardedsmaller digit in the previous digit-pair, thedesign of the task was thought to focus onsemantic aspects of the target for both selec-tion and response rather than on its localiza-tion. The instruction explicitly oriented par-ticipants to mark the larger digit, and did notinclude any mention to site-related features.Furthermore, the site alternatives were veryreduced - just two -, and there was nofixation point or a spatial distribution thatwould allow to organize the visual field intohemifields. Actually, the spatial type of NPcould involve some difficulties to be detectedby means of typical list procedures; how-ever, it would be very practical for clinicalassessment in order to develop a paper-and-pencil version of a site-based NP task. Pre-liminary results of a pilot test in which aversion of the site-based ‘OX’ task wasimplemented in a paper-and-pencil designwith a massed-list procedure did not demon-strate a particular slowing for targets im-printed at the previously ignored sites (RosinFM, unpublished results). In any case, thepotential site-based effect could be mixedwith identity-based processes in the responsesto the digit-comparison task. Thus, besidesthe paper-and-pencil digit-comparison task,middle-aged and older participants performed

on a computerized version of the digit-com-parison task in which the identity was theattribute overtly reported. In the computer-implemented digit-comparison task, they wereasked to name aloud the digit target thatappeared in digit-pairs on a screen display.Then, participants performed a modified ver-sion of the ‘OX’ spatial NP task such that thetarget location was the reported attribute(24).

Since older participants exhibited slow-ing in response times for all tests and propor-tional (ratio) scoring has been proposed inprevious studies of NP as a more conserva-tive method for comparisons between groupswith different baseline in response times (8),in this study we included proportional meas-ures of NP in data analysis to compare groupeffects. Analysis of the proportional effectson response times has been proposed be-cause slowing usually tends to magnify ab-solute differences between conditions, andANOVA performed on the difference in scoresbetween groups with different baseline re-sponse times - or across tasks that differ intheir respective baseline measures - will of-ten show significant group interaction ef-fects despite exhibiting similar proportionaldifferences in response times. Using propor-tions rather than absolute differences couldcontinue to be debated because convertingdifferences to proportions of baseline foreach group would yield diminished NP forsubjects who respond in an overall slowermanner than for their corresponding controlgroups. As pointed out, a method whichassumes a proportional relationship in theresponse times accounts for the generalslowing problem that is closely involved inmeasuring specific attentional effects notonly in aging subjects, but also in patientswith brain damage (8,9), and proportionalscoring would provide a more accurate de-piction of performance than the alternative ofexamining only absolute difference scores.While proportional scoring may fail to beaccurate enough to capture the complexity

1142


F.M. Rosin

of an effect if there is differential slowing ofspecific processes underlying task perfor-mance, it may be appropriate if the assump-tion is made that all processes underlyingtask performance are slowed uniformly.Developmental studies specifically targetingthis problem have shown that a global, notlocal, slowing may affect information pro-cesses underlying performance in memoryand selective attention tasks in older people(33). This is a reason to emphasize theimportance of first including a measure thatdifferentiates between individuals with po-tentially different baselines in their responselatencies when more specific attentionalmeasures should be assessed. Then, it ispossible to compare gross performance on amore basic level of processing and to testwhether there exist group effects on in-creases in latency, and to contrast overallslowing effects of increasing complexity ofthe task with process-specific effects. Fur-thermore, meta-analyses conducted fromseveral developmental studies provided evi-dence to support the view that NP is propor-tional rather than additive, with diminishedNP in older adults as compared with youngeradults (34).

In the present study the possible dispro-portionate effects on response times wereassessed when age group performance dif-fered in the simplest condition, i.e., when thetarget was paired with a non-conflictingstimulus (asterisk) or when it was presentedwithout distractors in one of the four pos-sible locations on the site-based NP task.Then, the cost for the condition of increasingcomplexity of the task was estimated besidesthe experimental effect of NP. Moreover,central conclusions were derived in a moreconservative way, based on patterns of per-formance that are qualitatively similar whenestimated as absolute or proportional differ-ences.

Other question that the present experi-ment addresses is the age-group differencein years of education that could affect NP

measures from the paper-and-pencil digit-comparison task. Although this task wasspecially designed to not be affected byeducation, the Experiment 3 reassessed NPfrom the paper-and-pencil task with youngerparticipants that were university students,and with middle aged and older adults withequivalent years of education.

Method

Participants. Eight younger adults, threewomen and five men (range = 18-24 years,mean = 21.38), eight middle-aged adults, fivewomen and three men (range = 31-54 years,mean = 41.00), and eight older adults, threewomen and five men (range = 68-77 years,mean = 70.88), participated in this study.The younger adults were undergraduate stu-dents and had a mean of 13.38 (SD = 1.51)years of education. Mini-Mental State Ex-amination (35) was given for the middle-aged and older adults to test their generalcognitive abilities. The middle-aged adultswere recruited from University Campus per-sonnel and had a mean of 8.25 (SD = 3.65)years of education. The older adults wererecruited from several recreation centerswhere they were actively involved; theywere selected from a larger sample in orderto be matched to the middle-aged individualsin terms of years of education and Mini-Mental Score, and they had a mean of 6.75(SD = 3.20) years of formal education. Noneof the participants reported histories or signsof neurological or psychiatric diseases and allhad normal or corrected-to-normal near vi-sual acuity required to perform the tasks. Allsubjects gave informed consent to partici-pate in the investigation and none of themparticipated in more than one study. Theresearch was approved by the Ethics Com-mittee as cited in Experiment 1.

Since performance of tasks supposed todemand attentional inhibitory processes seemsto be influenced by circadian rhythms inaging (36), the best circadian peak of each

1143



older participant was considered when plan-ning the schedule of the experimental ses-sions by applying the Morningness-Evening-ness Questionnaire (37), which was appliedto the participants by telephone. Three olderparticipants were definitely morning typesand five older participants were moderatelymorning types.

Procedure. Participants were assessedindividually. Younger participants completedthe paper-and-pencil digit-comparison task.The Mini-Mental State Examination (versiontraduced and revised by Bertolucci PHF,Universidade Federal de São Paulo, SP, Bra-zil) was given to the middle-aged and olderadults to test their general cognitive abilities.The experimental tasks, in the order in whichthey were administered to the middle-agedand older participants, were the paper-and-pencil digit-comparison task, the computer-ized version of the digit-comparison task,and the computerized ‘OX’ local-based task.For the computerized tasks, testing tookplace in a quiet and dimly lit room andparticipants sat in a chair with adjustableheight, with their heads standing approxi-mately 60 cm from the center of the screen.

Paper-and-pencil digit-comparison task

Material and procedure were identical tothose described in Experiment 2, except thatthe subject did not complete the additionalfinal unrelated sheet. Thus, there were threelist conditions (asterisk vs unrelated vs re-lated), and the marking times for each listwere the critical measures to indicate thebaseline of speed of performance and toestimate NP.

Computerized digit-comparison task

An additional measure of NP was ob-tained from a digit-comparison task that wasprogrammed using the MEL ProfessionalV2.01 software (38). In all trials, subjectsnamed aloud the digit of a digit-asterisk pair

or the greater digit of a digit-pair. Stimuluspairs were presented binocularly on a colordual scan display (16 cm x 21 cm) of an IBM-compatible notebook computer elevated 20cm from the desk and equipped with a serial-response box (Psychological Software Tools,PA, USA). A voice key interface was used tomeasure naming latencies. Response timingwas measured at 1-ms resolution.

Digits (from 0 to 9) measured 4 (width)x 8 (height) mm, subtending approximately0.39º x 0.76º of the visual angle; and theasterisk measured 5 x 6 mm, subtendingapproximately an angle of 0.48º horizontallyand of 0.57º vertically. The fixation pointwas a plus sign, subtending approximately0.57º x 0.57º vertically (6 x 6 mm) and wascentered on the screen. The font type was ofthe ‘system56.fnt’ graphics of ProfessionalMEL2. Pairs of stimuli were centered on thescreen with approximately 1.7 cm separatingtheir nearest edges. The color of the stimuliwas white and the stimuli were shown againsta dark background.

There were three trial conditions corre-sponding to the three levels of the within-subjects factor: asterisk trial, which pre-sented a digit-asterisk pair; unrelated trial,which presented a digit-pair with digits in theprevious trial different from digits in thecurrent trial, and related trials with a digittarget identical to the distractor in the previ-ous trial. Trial conditions were randomlydistributed. There were 50 asterisk trials, 54unrelated trials, and 59 related trials. Targetand distractor locations, i.e., to the left or tothe right of the fixation, were evenly bal-anced and varied randomly from trial to trial.Subjects responded to the target in a fixedongoing sequence of trials. Thus, the stimulithat acted as a probe in a particular trial alsoacted as a prime in the next trial. Each trialproceeded as follows: a fixation cross con-sisting of a 6 x 6 mm plus sign (+) centeredon the computer screen for 250 ms, a blankscreen for 250 ms, and the display containinga digit-asterisk pair or a pair of digits which

1144


F.M. Rosin

remained visible until the participant made anoral response. Then, the cross fixation reap-peared for 250 ms for the next trial, and soforth. There were two first practice blocks,each consisting of five trials, and three ex-perimental blocks consisting of 67, 73 and 76trials, respectively. Practice trials, the firsttrial for each experimental block, and trialsfollowing asterisk trials were not included indata analysis.

Subjects were encouraged to rest be-tween blocks. Each block was initiated bythe experimenter by pressing the space bar.The naming latencies were measured as theinterval between the stimulus onset and theparticipant’s vocal response into a micro-phone. Accuracy was recorded by the ex-perimenter on a response list. Both speed andaccuracy were emphasized, but subjectswere told that if they made a mistake theyshould continue with the digit naming taskrather than trying to correct it.

Error trials, spoiled trials when theparticipant’s response failed to trigger themicrophone or when it was triggered prior to100 ms, and trial immediately following anerror were excluded from the naming-la-tency analysis of the data. The dependentvariables were median naming latencies anderror percentages by trial condition. Thenaming latencies for asterisk trials provideda baseline measure of performance. NP wasassessed by comparing the median naminglatencies for the unrelated trials to the naminglatencies for the related trials.

Location-based priming task

One spatial measure of NP was obtainedfrom a computerized task designed on thebasis of the ‘OX’ task for location-based NP(24) using the MEL Professional V2.01 soft-ware to record reaction times, with accuracyof up to 1 ms (38). In all trials, subjectsindicated the location of a target (the letter‘O’) appearing, alone or with a distractor(the letter ‘X’), at one of four locations by

pressing the key corresponding to the loca-tion. Stimuli were presented on an IBM-compatible notebook computer, which waselevated 20 cm from the desk and equippedwith a serial-response box (PsychologicalSoftware Tools) with five buttons and lamps(rightmost button and lamps were covered)aligned with the screen.

The four locations were four horizontalbars (4 x 1.5 mm) centered on the screen,aligned in a row subtending approximately6.4º of the visual angle (67 mm) between thetwo outside positions, and remained on thescreen throughout all displays marking fourpositions. The fixation point was a plus sign,subtending approximately 0.48º horizontallyand 0.39º vertically (5 x 4 mm), and wasdisplayed 7 mm above the row of locationsand between the second and third location.The target ‘O’ and distractor ‘X’ (upper caseletters of ‘system56.fnt’ graphics MEL2font type) subtended approximately 0.86ºvertically and 0.43º horizontally (9 x 4.5mm). The color of the stimulus was whiteand the stimulus shown against a dark back-ground.

Participants responded to the target, whileignoring the distractor, if present, in a fixedongoing sequence of trials. Thus, the posi-tion of the ‘O’ that was the probe on aparticular probe trial also acted as the primein the next trial. There were three types oftrial conditions: target-alone, unrelated tar-get-plus-distractor and related target-plus-distractor, which were randomly distrib-uted. In both target-alone and unrelated tri-als, the stimuli reappeared at locations thatwere not occupied in the preceding primetrial. In related trials, the target occupied thesame location the distractor had occupied inthe preceding prime trial. The dependentvariables were median response times andpercent errors for each trial condition. Twotypes of effects were observed: the interfer-ence of the distractor and NP. The measureof interference was derived by comparingthe response times for target-alone trials with

1145



those for unrelated trials. The difference inresponse time between these two types oftrials was a measure of the extent to whichthe presence of a distractor interfered withresponses to the location of the target. NPwas assessed by comparing the time torespond to unrelated trials with the time torespond to related trials. The difference inresponse time between these two types oftrials was a measure of spatial NP.

Trials were divided into five blocks: twofirst practice blocks consisting of six andfive trials, respectively, and three experimen-tal blocks consisting of 44, 49, and 47 trials,respectively. Locations of target and distractorwere evenly balanced. There were 37 relatedtrials, 37 unrelated trials, and 32 target-alonetrials. The first trial of each block, two buffertrials, 29 trials following target-alone trials,error trials and the trial immediately follow-ing an error trial, and response made prior to150 ms following target-display presentationwere not included in the response time anal-ysis of the data.

Participants were shown the stimulusdisplay and the four keys of the serial box andwere asked to fixate on the fixation pointwhile it was activated and to press the re-sponse key corresponding to the ‘O’ loca-tion. All participants were instructed to usethe middle and index fingers of the left handto indicate the left locations, and the middleand index fingers of the right hand to indicatethe right locations. Participants were askedto ignore the ‘X’ as it was irrelevant to thetask and to try to be as quick and accurate aspossible in responding to the location of the‘O’. The experimenter pressed the space barwhen the participant was ready to resume.Each trial began with the onset of the fourlocation markers. After 500 ms, a fixationcross was presented and left on the screenfor 800 ms. The target, alone or presentedwith the distractor, followed the offset of thefixation cross and remained on the screenuntil a response was made. After the subjectresponded to the display, another sequence

delay-fixation-display was reinitiated, and soforth. There was no feedback for correctresponses. A 0.5-s tone (400 Hz) signaled theincorrect responses.


Subject comparisons

As previously stated, the group of youngerstudent participants had more years of edu-cation than did the older and middle-agedgroups, U = 3.00, Z = -3.05, P = 0.002. Olderand middle-aged participants did not differ intheir years of education, U = 26.0, Z = -0.63,P = 0.650, or their Mini-Mental Scores, U =19.0, Z = -1.37, P = 0.172, with a mean score(Max score = 30) of 28.38 (SD = 1.06) forthe older group and of 29.13 (SD = 0.83) forthe middle-aged group.

The alpha level for all analyses was P <0.05. The data were first subjected to ANOVAwith age as between-subject factor and con-dition as within-subject factor, with repeatedmeasures of conditions. This analysis wasfollowed by Duncan’s multiple range test forpost hoc comparisons. Since older partici-pants exhibited a different baseline in re-sponse times, with overall slowing for alltasks, and since a proportional (ratio) scor-ing has been proposed as a more conserva-tive way for comparisons between groupswith different baseline in response times (8),proportional measures were also used tocompare group effects. Thus, proportionalscoring relative to the appropriate controlcondition was computed for group compari-sons and across-task comparisons.

Paper-and-pencil digit-comparison task

Table 2 summarizes the mean markingtimes and errors for each list condition by agegroup. Errors were infrequent (14 in thewhole experiment) and did not differ signifi-cantly across age groups or types of list.Percent errors made in each list condition for

1146


F.M. Rosin

each subject were subjected to separate Fried-man ANOVA, with repeated measures oncondition, and Kruskal-Wallis ANOVA withage as between-subject factor, both of whichyielded non-significant effects. The differ-ence in errors across list conditions was notstatistically significant for younger partici-pants, χ2(8, 2) = 4.00, P = 0.135, or middle-aged participants, χ2(8, 2) = 4.67, P = 0.097,or older adults, χ2(8, 2) = 3.50, P < 0.174.Age groups did not differ in their errors forasterisk, related, or unrelated lists, H (2, 24)= 0.00, P = 1.00. Related and unrelated listswere the critical conditions for measuringNP. The data were first subjected to 3 (age:younger vs middle-aged vs older) x 3 (condi-tion: asterisk vs unrelated vs related) ANOVA.This analysis revealed that older adults wereslower than younger and middle-aged adults,F(2, 21) = 4.67, P = 0.021. List conditionwas significant, F(2, 42) = 104.05, P <0.0001, with the following profile in markingtimes: asterisk < unrelated < related, indicat-

ing reliable NP. The age x condition interac-tion was not significant, F(4, 42) = 0.93, P <0.454.

Further ANOVA tests were first appliedto the absolute differences in performanceacross lists. Two effects were observed: thecost in speed of performance due to thedifference in complexity of the required cog-nitive operations between the digit-pairs andasterisk-digit conditions (unrelated markingtime minus asterisk marking time), and NP(related marking time minus unrelated mark-ing time). Groups did not differ in their coststo complete the digit-pair condition relative tothe easier asterisk-digit condition, F(2, 21) =0.15, P = 0.863. Differences in the cost tocomplete the related condition as expressedby absolute differences were not statisticallysignificant across groups, F(2, 21) = 2.77, P= 0.086. However, planned comparison anal-ysis indicated that the differences betweenrelated and unrelated lists were significantlysmaller for the older group than for the

Table 2. Response time and errors for conditions across tasks by age groups in Experiment 3.

Younger adults Middle-aged adults Older adults

Paper-and-pencil digit-comparison task(response time in seconds)

Asterisk 31.47 ± 5.48 (0.00 ± 0.00) 37.63 ± 8.13 (0.00 ± 0.00) 49.20 ± 9.60 (0.00 ± 0.00)Unrelated 46.33 ± 11.44 (0.00 ± 0.00) 51.17 ± 9.15 (0.46 ± 1.31) 62.41 ± 13.22 (0.23 ± 0.65)Related 52.00 ± 13.17 (0.25 ± 0.46) 56.73 ± 9.91 (0.50 ± 0.76) 64.09 ± 14.36 (0.63 ± 1.06)Unrelated-asterisk 14.85 ± 7.41 13.53 ± 5.21 13.20 ± 6.38Related-unrelated 5.68 ± 3.67 5.57 ± 3.06 1.68 ± 4.70

Computerized digit-comparison task(naming latencies in milliseconds)

Asterisk - 593.69 ± 65.74 (0.00 ± 0.00) 657.75 ± 92.83 (0.00 ± 0.00)Unrelated - 620.13 ± 62.10 (0.95 ± 1.76) 752.19 ± 102.03 (2.29 ± 0.62)Related - 635.69 ± 71.65 (2.52 ± 1.99) 721.13 ± 103.49 (3.93 ± 0.85)Unrelated-asterisk - 26.44 ± 19.50 94.44 ± 34.89Related-unrelated - 15.56 ± 18.38 -31.06 ± 22.52

Computerized local-based task(response time in milliseconds)

Target-alone - 591.56 ± 77.01 (0.00 ± 0.00) 710.81 ± 127.15 (3.52 ± 6.53)Unrelated - 616.69 ± 67.43 (1.50 ± 1.60) 742.50 ± 123.07 (5.09 ± 3.77)Related - 632.56 ± 83.85 (3.25 ± 4.39) 765.38 ± 122.92 (4.73 ± 3.64)Unrelated-target alone - 25.13 ± 22.02 31.69 ± 30.47Related-unrelated - 15.88 ± 21.43 22.88 ± 48.39

Response time and its units are given for each task. Errors reported as percent are given in parentheses. Data are reported as means ± SD forthe paper-and-pencil task and mean of medians ± SD for the other two tasks. N = 8 participants in each group.

1147



younger groups, F(1, 21) = 5.53, P < 0.029.Since the groups exhibited different

baselines in response times, statistical analy-ses were also conducted on proportionalscorings to compare group effects. Thus,the proportional (ratio) increase in markingtimes for the unrelated list relative to theasterisk condition was estimated as follows:digit-comparison cost = [(unrelated) - (as-terisk)]/(asterisk); and the proportional in-crease in marking times for the related listrelative to the unrelated condition was com-puted as follows: proportional NP = [(re-lated) - (unrelated)]/(unrelated). These esti-mates were submitted to separate one-wayANOVA, with age as random between-sub-ject factor. The proportional scores demon-strated effects that shared the same tenden-cies identified with absolute difference scores.The analyses revealed that proportional dif-ferences in the cost to compare betweendigits relative to the baseline measure ofperformance in the asterisk condition werenot statistically significant among groups,F(2, 21) = 3.12, P = 0.065. In contrast, theeffect of age on the specific proportionalmeasure of NP was statistically significant,F(2, 21) = 4.76, P < 0.02, with diminishedrelative slowing on the related list for olderadults relative to both groups of middle-agedand younger participants (see Figure 3).

Because specific predictions were madeand a significant difference in NP for olderparticipants was observed, planned com-parisons were used to explore the differ-ences among list condition means for thedifferent age groups. NP from the paper-and-pencil digit-comparison task, reportedas increased time to complete the related listcompared to the unrelated list, was evident inthe middle-aged group, F(1, 21) = 16.56, P< 0.0006, and younger group, F(1, 21) =17.21, P < 0.0005, with no significant differ-ences between them, F(1, 21) = 0. 03, P =0.956. NP was not observed in older adults,F(1, 21) = 1.51, P = 0.233, in agreement withprevious studies that suggested a failure in

identity suppression with aging. This studyalso shows that the observable NP in youngpeople seems to be reliable during middle-ageadulthood. In addition, this result makes itextremely unlikely that the NP measured bythe paper-and-pencil digit-comparison taskis compromised by years of education.

Computerized tasks. Mean percentage oferrors and mean of median naming latenciesand response times calculated after exclud-ing error and spoiled trials, for the three trialconditions by age group and the absolutedifferences between trial conditions for eachtask are reported in Table 2. Spoiled trials inthe digit-comparison task averaged 3.41%(SD = 1.65) for the middle-aged adults and18.80% (SD = 6.13) for the older adults. Toassess performance in the computerized tasks,the basic ANOVA used throughout was a 2(age) x 3 (trial condition), with repeatedmeasures of within-subject conditions.

For the digit-comparison task, middle-aged adults were faster than older adults,F(1, 14) = 5.04, P < 0.042; the naminglatencies differed across trial conditions, F(2,28) = 64.97, P < 0.0001, and age interactedwith trial condition, F(2, 28) = 18.16, P <0.0001.

To interpret the interaction and to com-pare the cost derived from comparisonsbetween digits relative to selecting in asterisktrials and the magnitude of NP betweengroups, ANOVA was applied to the absolute

1234512345

123412341234

Pro

port

iona

l N

P (

%)

20

15

10

5

-5

0

-10

Paper-and-pencildigit-comparison Computerized

digit-comparisonLocal-based ’OX’

Younger

Middle-aged

Older

123451234512345

1212

Figure 3. Proportional negative priming (NP) scores for each task in Experiment 3. Data arereported as the mean ± SEM. NP did not differ across younger and middle-agedparticipants in the paper-and-pencil digit-comparison task, F(1, 21) = 0.03, P = 0.956. Thegroup of older participants showed reduced NP for digit-comparison tasks but preservedlocal-based NP relative to the NP scores derived from the performance of middle-agedparticipants; MANOVA (2 age x 3 task), Rao R (3, 12) = 16.02, P < 0.0002. Based onTukey’s HSD *(P < 0.05) and **(P < 0.001).

***

1148


F.M. Rosin

differences in performance across lists. Simi-lar analyses of the proportional scorings:cost in the comparison-task = (unrelated -asterisk)/unrelated, and NP = (related - unre-lated)/unrelated yielded a similar pattern ofsignificant results. Older adults showed moreimpairment in their performance in the con-flicting digit-pair condition than middle-agedadults, F(1, 14) = 23.15, P < 0.0003, butmiddle-aged adults showed more NP thanolder adults, F(1, 14) = 20.58, P < 0.0005.Note in Table 2 that the difference betweenthe related and unrelated conditions in olderadults had a positive sign, indicating a re-versed NP effect rather than slowing forrecently ignored targets.

Error data were analyzed by the Wil-coxon matched pairs test applied to each ofthe effects of interest, i.e., the differencebetween the unrelated and asterisk condi-tions and NP, and age groups were com-pared by the Mann-Whitney U-test. For thedigit-comparison task, errors were not dif-ferent across the two trials in which NP wasassessed (related vs unrelated) for the middle-aged participants, T = 3.00, Z = 1.57, P =0.116, or older participants, T = 8.50, Z =0.93, P = 0.353. This result makes it unlikelythat slowing for the related condition wascompromised by speed/accuracy trade-off.Older adults made significantly more errorswhen the target was in a digit-pair than whenthe target was presented with a non-conflict-ing stimulus (unrelated vs asterisk), T = 0.00,Z = 2.20, P < 0.028. This finding followedthe same pattern as that observed for thenaming latencies, with slowing when a com-parison between conflicting stimuli was re-quired. Overall errors were not significantlydifferent between groups, i.e., for relatedtrials, U = 17.50, Z = -1.52, P = 0.127; forasterisk trials, U = 32.00, Z = 0.00, P = 1.00;and for unrelated trials, U = 17.50, Z = -1.68,P = 0.093, although older adults tended tomake more errors than middle-aged adults intrials with unrelated digit-pairs. Thus, no agedifferences existed for the error rates in the

related condition. Therefore, the speed/ac-curacy trade-off differences between agegroups did not limit the strength of thefinding that NP was absent for the group ofolder adults.

For the location-based task, the criticalconditions were related and unrelated target-plus-distractor trials for measuring NP, andtarget-alone and unrelated trials for measur-ing the interference by the distractor. Again,older adults were significantly slower thanmiddle-aged adults, F(1, 14) = 6.21, P <0.026. The main effect of conditions wasalso significant, F(2, 28) = 13.98, P < 0.0001,with significantly slower responses in relatedtrials, which presented the target at a recentlyignored location, and faster responses intarget-alone trials. The interaction betweenage and condition was not significant, F(2,28) = 0.28, P = 0.759.

The absolute differences in median re-sponse time for distractor interference (un-related minus target alone) and location-based NP (related minus unrelated) weresubmitted to separate ANOVA, with age as arandom between-subject factor. There wasno difference between the distractor inter-ference scores for the older participants’performance and those observed for themiddle-aged group, F(1, 14) = 0.24, P =0.629. As expected from previous findings inthe literature, the magnitude of spatial NP didnot differ significantly between the older andthe middle-aged groups, F(1, 14) = 0.14, P =0.714. Similar analyses conducted on pro-portional effects of distractor (interference =(unrelated - target alone)/unrelated and NP =(related - unrelated)/unrelated) yielded a simi-lar pattern of significant results.

Error data analyses showed that neitherdifferences between related and unrelatedtrials for the middle-aged, T = 1.50, Z = 1.27,P = 0.201, and older participants, T = 16.00,Z = 0.28, P = 0.779, nor differences betweenunrelated and target-alone trials for the middle-aged, T = 0.00, Z = 1.83, P = 0.068, and olderparticipants, T = 8.00, Z = 0.73, P = 0.463,

1149



were significant. Older adults made moreerrors under interference from the distractorin unrelated trials, U = 12.00, Z = -2.10, P =0.036, but there were no different error ratesacross age groups in related, U = 24.00, Z =-0.84, P = 0.401, and target-alone trials, U =24.00, Z = -0.84, P = 0.401.

Comparisons of negative priming measuresacross tasks

Finally, proportional NP scores (displayedin Figure 3) for each task, i.e., the paper-and-pencil digit-comparison task, the computer-ized digit-comparison task, and the local-based task, were assessed by 2 (age: middle-age vs older) x 3 (task) MANOVA, with ageas a random effect. The age effect wassignificant, Rao R (3, 12) = 16.02, P <0.0002. Post hoc comparisons revealed thatthe performance of older adults showeddiminished NP for both paper-and-pencil (P< 0.027) and computerized digit-comparisontasks (P < 0.0007) when compared to themagnitude of NP for the middle-aged partici-pants, whereas both older and middle-agedgroups showed spatial NP to an equivalentextent (P = 0.728). MANOVA conducted onthe absolute differences between the criticalconditions, related vs unrelated, yielded asignificant effect of age, Rao R (3, 12) =14.90, P = 0.0002, with a significant reduc-tion of NP in the older group performance inthe computer-implemented digit-comparisontask (P < 0.0006), approached significancein the paper-and-pencil digit-comparison task(P = 0.070), and absence of significantdifference in spatial NP (P = 0.714).

The primary results from this experimentshow not only that NP is reliable for thepaper-and-pencil digit-comparison task butalso a dissociation between the kind of NPderived from digit-comparison tasks and theNP from the spatial task which appears to beevident from older adults’ performance. Thisfinding is consistent with earlier reports show-ing that older adults, unlike younger adults,

appear to be unable to suppress the identity ofirrelevant stimuli while they show reliablesuppression of location (12). For older adults,the magnitude of proportional NP from thelocation-based task, but not from the digit-comparison tasks, was equivalent to the NPfor middle-aged adults, thus providing evi-dence that the paper-and-pencil digit-com-parison task may measure another type ofNP that is different from the spatial one.

Discussion

The central issue examined in this inves-tigation was whether reliable slowing in re-sponses associated with ignored repeatedtargets, i.e., NP, can be detected and meas-ured by means of a new and practical paper-and-pencil version of a digit-comparison task.A secondary aim of this research was toexplore the magnitude of this effect of NP inolder adults, for whom failures to demon-strate NP may be particularly evident in someselective attention tasks that require identifi-cation of a stimulus which had been previ-ously disregarded as being a distractor. Themagnitude of NP from the paper-and-penciltask was assessed in younger, middle-agedand older adults, and age-related differencesin types of NP were studied across digit-comparison and location-based tasks.

The paper-and-pencil digit-comparisontask appears to be simple, it can be easily andrapidly administered, and detects NP, asindicated by slowing to complete the condi-tion with related items (with the distractor ina digit-pair appearing as a target in the nextdigit-pair) relative to the condition with unre-lated digit pairs. Three separate markingtimes are recorded, one for each list condi-tion (i.e., one list containing non-conflictingpairs, a second list with unrelated digit-pairs,and a third list with related digit-pairs), whichpermit a baseline measure of performance toestimate the additional time needed to makethe choice of the target by comparisonsbetween digits and provide an index of NP of

1150


F.M. Rosin

slowing for recently unselected targets. Us-ing proportional scoring of differences acrosslists may be pertinent here because overallolder adults perform more slowly in thetasks, as it became evident in Experiment 3,and because of previous models of perfor-mance in selective attention tasks that sug-gest a uniform increment in latency in normalaging rather than a phenomenon requiringprocess-specific explanations (33). Whilethe age groups exhibited a different baselinein marking times, manipulating the complex-ity of target selection had no demonstrabledifferential influence on proportional (or ab-solute differences) measures of performanceacross the age groups. On the basis of theassumption that the variation of older indi-viduals’ performance obeyed a general slow-ing instead of depending on the type ofinformation processes involved, proportionalindices can provide an appropriate way tointerpret possible age differences in perfor-mance in the related condition.

Data analyses clearly showed a reliableeffect of NP; i.e., subjects were slower incompleting the form containing the relateddigit-pairs than in completing the list withunrelated digit-pairs. In the last version of thetask, performance in the unrelated conditionproved to be faster, before or after the relatedcondition. This finding is inconsistent withthe claim that slowing is produced by expect-ancy biases developed across the lists. NPappears to be present to the same extent foryounger and middle-aged adults, in contrastto the performance of older adults for whomNP was reliably diminished. That is, olderadults did not seem to complete the relatedlist more slowly, suggesting that the paper-and-pencil digit-comparison task may tap akind of NP that appears to be impaired byaging, possibly reflecting the existence ofage-related deficits in distractor suppressionsuch as those proposed in previous develop-mental studies (6,15).

According to this theory, a reduction inthe efficiency of inhibitory mechanisms may

underlie age differences in a variety of cog-nitive tasks so that the capacity of controllingaccess to and sustaining the non-relevantinformation in working memory may beimpaired in aging under some circumstances,explaining the failure to gate out of workingmemory stimuli and thoughts that are notdirectly relevant to current goals. However,multiple inhibitory systems have been pro-posed, not all of which are reduced in effec-tiveness with age. While in identity suppres-sion tasks (i.e., where the distractor in a pairreappears as the target on the next pair) olderadults typically show reduced or no NPcompared to younger adults, studies em-ploying specific location-based tasks (i.e.,where target in a current display appears inthe same location as the distractor in thepreceding display) have demonstrated thatolder adults show as much NP as youngeradults (12,14). Indeed, in testing a group bypriming interaction, it is necessary to pro-ceed with caution to affirm the overall ab-sence of NP in normal aging, since there existseveral reports of reliable NP in older adultscompared to younger adults not only inspatial tasks but also in identity tasks (10,27,31,34). Besides the type of NP involved,mnemonic strategies which could be elicitedto some extent by experimental contextualvariables (e.g., difficulty in perceptual selec-tion, reduction of the probe display expo-sure) seem to affect the age group by priminginteraction (13,39), with evident NP forolder adults only demonstrable under condi-tions designed to encourage episodic re-trieval. However, what processes moderatenegative priming in aging is still under debate(27,40).

Although this issue is not resolved here, theresults of the present study (Experiment 3) areconsistent with the previous statement that NPderived from tasks that require suppressingthe identity of distractors would be lessrobust in older adults. Sparing of spatial NPin the older group is consistent with thehypothesis that inhibitory mechanisms that

1151



operate to prevent processing of irrelevantlocations seem to be preserved in aging.

The paper-and-pencil digit-comparisontask uses the referent-size selection proce-dure, a variant of the NP procedure thatensures that both targets and distractors areattended through comparisons between theirsemantic referents, instead of using a typicalNP procedure requiring participants to at-tend a feature of the target while ignoring thedistractor. The referent-size selection proce-dure is not exclusive to the present study. Ithas been explored recently (5,27) by usingcomputerized tasks and words as stimuli(e.g., participants were asked to read thename of the larger animal, or the heaviernoun), with the finding that NP not onlyoccurred when the to-be-ignored distractorwas overtly attended but was also actuallyenhanced in this procedure. This report hastheoretical and practical implications. It goesagainst the view that NP is presumably ameasure of pure attentional inhibition arisingfrom ignoring prime distractors or not di-recting attention to them, by which deepprocessing of distractors as that involved indecision-type tasks that afford post-lexicalprocessing should prevent NP. On this basis,participants that were able to manifest aware-ness upon NP manipulation were excludedfrom further data analysis in several studies(e.g., 3,12). The finding of reliable NP forattended distractors is not consistent withthe notion that a mechanism that limits phe-nomenal awareness to relevant items onlyunderlies NP (5). This issue is also importantfor appropriate interpretations when NP tasksare used in clinical groups. A practical impli-cation relates to the increased magnitude ofNP when the procedure involves overtlyattending distractors relative to the typicalNP procedures, by which the referent-sizeselection procedure seems to be more suit-able for the clinical-experimental setting.

Diminished NP in older adults in taskswith an overtly attended distractor may sup-port a memory model, because this type of

task favors greater depth of processing inwhich episodic retrieval would be advanta-geous. On the basis of the assumption thatthe semantic representation of both attendedtarget and distractor is positively primed,response times to these items in subsequenttrials may be facilitated. Further faster re-sponse time to the attended distractor at itsoccurrence, however, adds to slowing of theresponse derived from a “do not respond to”tag that became associatively linked with itssemantic memory representation, which isautomatically retrieved when a similar objectis subsequently encountered. These encod-ing and retrieval mechanisms that mediateNP have been proposed to make distractinginformation less available to flow automati-cally into actions and enable coherent goal-directed behavior, and they may vary withage. NP seems to reflect cognitive processesthat are functionally important for effectiveattentional processing and executive func-tioning and that could be partially compro-mised by age-related factors.

Intense efforts are currently being madein an attempt to obtain a clearer picture ofexecutive functioning by carrying out exper-imental tests of cognitive models. Executiveprocesses are thought of in terms of a generalattentional resource involved in several day-to-day abilities such as reasoning, decisionmaking, comprehension, and memory (1).As explained above, NP has been broadlyinvestigated because of its relevance to theo-ries of selective attention. That is, NP istypically said to measure either attentionalinhibition or episodic memory retrieval. Thereexist many kinds of NP tasks which havebeen developed to understand NP, an effectthat constitutes a challenge for the mostcurrent theories of attention. More sophisti-cated neuropsychological tests could be valu-able for advances in specifying more pre-cisely how different NP tasks and conditionscan moderate the processes involved in NP,which vary in both direction and magnitude.On the other hand, a promising application to

1152


F.M. Rosin

the diagnosis of cognitive dysfunction re-quires the development of simple tasks thatare suitable for neuropsychological clinicalroutines. Computerized NP tasks broadlyused in the studies reviewed often presentseveral problems for day-to-day clinical prac-tice. They require expensive and specializedequipment and a more controlled environ-ment (e.g., technical parameters of equip-ment and software, luminance of the displayand testing room) such that standardizationbecomes a difficult goal. Reading-based NPtasks show constraints under some circum-stances, such as when the population pre-sents heterogeneity in the level of readingability.

Since the paper-and-pencil digit-compar-ison task can be practically administered, itcan be very useful for neuropsychologistsseeking practical measures of NP that do notrequire cumbersome technical equipmentwithin the clinical-experimental setting, in-volving within- and between-subject designs,since it has proved to be a sensitive andpractical screening tool to index NP and todetect individual differences. In the presentstudy, the number of errors and omissionswere relatively low, and the method of ad-ministration and time recording were rapidand easy. The magnitude of NP from thedigit-comparison task (with a mean differ-

ence between the related and unrelated listsof approximately 6 s and a proportional NP ofapproximately 12% for younger and middle-age adults, as can be seen in Table 2 andFigure 3) was considerably enhanced ascompared with other non-computerized NPtasks with ignored distractors, e.g., with amean difference of less than 1 s (proportionalNP = 4.57%) in a previous study using the listprocedure (30). In addition, sex or years ofeducation did not appear to affect the NPscore for the paper-and-pencil digit-compar-ison task. These characteristics of the taskare favorable for the clinical setting andsupport further studies to determine andimprove its effective utility as a neuropsy-chological tool.

Acknowledgments

I am indebted to Cesar Galera for guid-ance during all my Ph.D. studies and forproviding the facilities for this research.Special thanks are due to Rolando P. Sylwanfor advice and assistance during all stages ofthis research. I am also grateful to the anony-mous reviewers of an earlier version of thisarticle, and to Paul Stephaneck, J. Lino O.Bueno, Luiz Gawryszewski, and Luis Ribeirodo Valle for valuable comments that havebenefited this work.

References

1. Baddeley AD, Della Sala S, Gray C, Papagno C & Spinnler H (1997).Testing central executive functioning with a pencil-and-paper test.In: Rabbitt P (Editor), Methodology of Frontal and Executive Func-tions. Psychology Press, Hove, East Sussex, UK.

2. Dalrymple-Alford EC & Budayr B (1966). Examination of someaspects of the Stroop colour-word test. Perceptual and Motor Skills,23: 1211-1214.

3. Tipper SP (1985). The negative priming effect: inhibitory priming byignored objects. Quarterly Journal of Experimental Psychology,37A: 571-590.

4. Fox E (1995). Negative priming from ignored distractors in visualselection: a review. Psychonomic Bulletin and Review, 2: 145-173.

5. MacDonald PA & Joordens S (1999). Negative priming effects thatare bigger than a breadbox: attention to distractors does noteliminate negative priming, it enhances it. Memory and Cognition,27: 197-207.

6. Tipper SP (1991). Less attentional selectivity as a result of declining

inhibition in older adults. Bulletin of the Psychonomic Society, 29:45-47.

7. Beech A, Powell T, McWilliam J & Claridge G (1989). Evidence ofreduced “cognitive inhibition” in schizophrenia. British Journal ofClinical Psychology, 28: 109-116.

8. Stuss DT, Toth JP, Franchi D, Alexander MP, Tipper S & Craik FIM(1999). Dissociation of attentional processes in patients with focalfrontal and posterior lesions. Neuropsychologia, 37: 1005-1027.

9. Metzler C & Parkin AJ (2000). Reversed negative priming followingfrontal lobe lesions. Neuropsychologia, 38: 363-379.

10. Sullivan MP, Faust ME & Balota D (1995). Identity negative primingin older adults and individuals with dementia of the Alzheimer type.Neuropsychologia, 9: 537-555.

11. Stout JC, Wylie SA, Simone PM & Siemers ER (2001). Influence ofcompeting distractors on response in Huntington’s disease andParkinson’s disease. Cognitive Neuropsychology, 18: 643-653.

12. Connelly SL & Hasher L (1993). Aging and the inhibition of spatial

1153



location. Journal of Experimental Psychology: Human Perceptionand Performance, 19: 1238-1250.

13. Kane MJ, May CP, Hasher L, Rahhal T & Stoltzfus ER (1997). Dualmechanisms of negative priming. Journal of Experimental Psychol-ogy: Human Perception and Performance, 23: 632-650.

14. Simone PM & McCormick EB (1999). Effect of a defining feature onnegative priming across the life span. Visual Cognition, 6: 587-606.

15. McDowd JM & Oseas Kreger DM (1991). Aging, inhibitory pro-cesses, and negative priming. Journal of Gerontology, 46: 340-345.

16. Houghton G & Tipper SP (1994). A model of inhibitory mechanismsin selective attention. In: Dagenbach D & Carr TH (Editors), Inhibito-ry Processes in Attention, Memory, and Language. AcademicPress, New York.

17. Milliken B, Joordens S, Merikle PM & Seiffert AE (1998). Selectiveattention: a reevaluation of the implications of negative priming.Psychological Review, 105: 203-229.

18. Neill WT, Valdes LA, Terry KM & Gorfein DS (1992). Persistence ofnegative priming: II. Evidence for episodic trace retrieval. Journal ofExperimental Psychology: Learning, Memory, and Cognition, 18:993-1000.

19. May CP, Kane MJ & Hasher L (1995). Determinants of negativepriming. Psychological Bulletin, 118: 35-54.

20. Tipper SP (2001). Does negative priming reflect inhibitory mechan-isms? A review and integration of conflicting views. QuarterlyJournal of Experimental Psychology. Section A, 54: 321-343.

21. Tipper SP, MacQueen GM & Brehaut JC (1988). Negative primingbetween response modalities: evidence for the central locus ofinhibition in selective attention. Perception and Psychophysics, 43:45-52.

22. Tipper SP, Eissenberg T & Weaver B (1992). The effects of practiceon mechanisms of attention. Bulletin of the Psychonomic Society,30: 77-80.

23. Neill WT (1977). Inhibitory and facilitatory processes in selectiveattention. Journal of Experimental Psychology: Human Perceptionand Performance, 3: 444-450.

24. Tipper SP, Weaver B, Cameron S, Brehaut JC & Bastedo J (1991).Inhibitory mechanisms of attention in identification and localizationtasks: time course and disruption. Journal of Experimental Psychol-ogy: Learning, Memory, and Cognition, 17: 681-692.

25. Neill WT & Valdes LA (1992). Persistence of negative priming:steady state or decay? Journal of Experimental Psychology: Learn-ing, Memory, and Cognition, 18: 565-576.

26. Tipper SP, Weaver B & Milliken B (1995). Spatial negative primingwithout mismatching: comment on Park and Kanwisher (1994).Journal of Experimental Psychology: Human Perception and Per-

formance, 21: 1220-1229.27. Pesta BJ & Sanders RE (2000). Aging and negative priming: is

ignored information inhibited or remembered? Experimental AgingResearch, 26: 37-56.

28. Goldman Rakic PS (1995). Architecture of the prefrontal cortex andthe central executive. Annals of the New York Academy of Sci-ences, 769: 71-83.

29. Neill WT & Westberry RL (1987). Selective attention and thesuppression of cognitive noise. Journal of Experimental Psychol-ogy: Learning, Memory, and Cognition, 13: 327-334.

30. Tipper SP & Cranston M (1985). Selective attention and priming:inhibitory and facilitatory effects of ignored primes. Quarterly Jour-nal of Experimental Psychology, 37A: 591-611.

31. Langley LK, Overmier JB, Knopman DS & Prodhomme MM (1998).Inhibition and habituation: Preserved mechanisms of attentionalselection in aging and Alzheimer’s disease. Neuropsychology, 12:353-366.

32. Bestgen Y & Dupont V (2000). Is negative priming a reliablemeasure for studying individual differences in inhibition? CurrentPsychology of Cognition, Cahiers de Psychologie Cognitive, 19:287-305.

33. Cerella J (1991). Age effects may be global, not local: Comment onFisk and Rogers. Journal of Experimental Psychology: General,120: 215-233.

34. Verhaegen P & De Meersman L (1998). Aging and the negativepriming effect: a meta-analysis. Psychology and Aging, 13: 435-444.

35. Folstein MF, Folstein SE & McHugh PR (1975). “Mini-mentalState”. A practical method for grading the cognitive state ofpatients for the clinician. Journal of Psychiatric Research, 12: 189-198.

36. May CP & Hasher L (1998). Synchrony effects in inhibitory controlover thought and action. Journal of Experimental Psychology:Human Perception and Performance, 24: 363-379.

37. Horne JA & Ostberg O (1976). A self-assessment questionnaire todetermine morningness-eveningness in human circadian rhythms.International Journal of Chronobiology, 4: 297-310.

38. Schneider W (1988). Micro experimental laboratory: an integratedsystem for IBM-compatibles. Behavior Research Methods, Instru-ments, and Computers, 20: 206-217.

39. Gamboz N, Russo R & Fox E (2000). Target selection difficulty,negative priming, and aging. Psychology and Aging, 15: 542-550.

40. Grant JD & Dagenbach D (2000). Further considerations regardinginhibitory processes, working memory, and cognitive aging. Ameri-can Journal of Psychology, 113: 69-94.

Assessing negative priming by attended distractors in a ... · 1131 Braz J Med Biol Res 37(8) 2004...

Documents

Transcript of Assessing negative priming by attended distractors in a ... · 1131 Braz J Med Biol Res 37(8) 2004...