Pqje118805 1224.1236

THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY2006, 59 (7), 1224 – 1236 Heinrich-Heine-Universita¨t, Du¨sseldorf, Germany Performance of 28 schizophrenic patients and 28 matched controls was compared in an auditorypriming task. A large auditory negative priming effect was obtained for the patients as well as forthe control group, and the size of the negative priming effect was approximately the same for bothgroups. Under the same conditions, positive or repetition priming for the patients was enhancedcompared to that of the control group. The present findings from an auditory priming task are con-sistent with a growing body of evidence from the visual domain showing normal rather than reducedor eliminated negative priming in schizophrenic patients.
The negative priming phenomenon manifests 1977), negative priming reflects the operation of itself in slowed-down or more error-prone an inhibitory attentional selection mechanism reactions to recently ignored stimuli compared to that prevents access of recently ignored objects to those for control stimuli that are unrelated to the overt responses by suppressing competing distrac- previous stimuli (for reviews, see Fox, 1995; tor input. This inhibitory mechanism enables May, Kane, & Hasher, 1995; Neill, Valdes, & more efficient responding to the current target Terry, 1995; Tipper, 2001). Several models are under normal circumstances, but causes a delay currently available to explain the negative in responding when, as in a negative priming labo- priming phenomenon. Of these models, the so- ratory task, the previously ignored (and, hence, called distractor inhibition model has a special supposedly suppressed) distractor becomes the status not only because it is historically the oldest model that explains this phenomenon but also This attentional explanation of the negative because it suggests that the negative priming priming phenomenon is interesting for schizo- paradigm may be an appropriate task for testing phrenia research because attentional impairment predictions of loss-of-inhibition theories of the is one of the fundamental cognitive deficits associ- changes in cognitive functioning induced by ated with schizophrenia. In particular, increased schizophrenia, ageing, and other conditions.
distractibility in the presence of irrelevant infor- According to the version of this model proposed mation has been widely attributed to disrupted by Tipper (1985; see also Dalrymple-Alford & mechanisms responsible for the direction and Budayr, 1966; Houghton & Tipper, 1994; Neill, control of attention—that is, to the impaired Correspondence should be addressed to Axel Buchner, Institut fu¨r Experimentelle Psychologie, Heinrich-Heine-University, D-40225 Du¨sseldorf, Germany. Email: [email protected] The research reported in this article was supported by a grant from the Deutsche Forschungsgemeinschaft (Bu 945/2–1).
# 2006 The Experimental Psychology Society functioning of inhibitory attentional mechanisms priming, but only in subgroups of schizophrenic (e.g., Frith, 1979; McGhie, 1977; McGhie & patients. For instance, normal negative priming Chapman, 1961). With respect to the negative has been reported for chronic outpatients but not priming paradigm, the assumption of impaired for acutely psychotic schizophrenic inpatients inhibitory attentional mechanisms in schizo- (S. Park, Lenzenweger, Pu¨schel, & Holzman, phrenia predicts less or no slow-down of the 1996; S. Park, Pu¨schel, Sauter, Rentsch, & Hell, reactions of schizophrenic patients to previously 2002; see also Salo, Henik, Nordahl, & Robertson, ignored distractors compared to those of normal 2002), and for medicated but not for unmedicated controls, provided that the distractor inhibition schizophrenic patients (Salo, Robertson, & model is valid. The negative priming task itself Nordahl, 1996; Salo, Robertson, Nordahl, & appears particularly useful for testing the loss-of- Kraft, 1997), as well as conversely for unmedicated inhibition assumption in schizophrenia because a but not for medicated schizophrenic patients cognitive deficit is expressed in improved perform- (David, 1995). In contrast, Moritz et al. (2001) ance (no reaction slow-down), rather than impo- reported that symptomatology and neuroleptic verished performance, which is usually explained medication did not moderate the size of the nega- by a generalized cognitive deficit rather than a tive priming effect. Interestingly, they noted that specific inhibitory attentional impairment.
procedural details may determine whether or not negative priming is observed in schizophrenic Claridge (1989) started off such a research pro- patients. In their study, control participants and gramme and reported reduced negative priming schizophrenic patients did not differ in terms of in a group of schizophrenic patients compared to negative priming when the prime presentation dur- that in a control group. This finding was inter- ation was 250 ms, but schizophrenic patients failed preted as being consistent with the assumption to show any priming when primes were presented that schizophrenia implies a reduction in the ability for 100 ms and were followed by a pattern mask, of the cognitive system to inhibit and suppress as in Beech et al. (1989). Moritz et al. argued that previous findings of no or reduced negative However, subsequent research has produced priming in schizophrenic patients may have been different and somewhat inconsistent results. For due to a procedural artifact—that is, the very instance, Laplante, Everett, and Thomas (1992) short presentation durations combined with did not find reduced negative priming in schizo- pattern masking, which simply may have impaired phrenic patients relative to depressive patients the patients’ processing of the prime distractor.
and healthy control participants.1 Parallel results have recently been reported in several studies Wagner (2002) reported normal levels of identity (Baving, Wagner, Cohen, & Rockstroh, 2001; negative priming in schizophrenic patients in a Moritz, Jacobsen, Mersmann, Kloss, & Andresen, task that required selection by stimulus identity, 2000; Roesch-Ely, Spitzer, & Weisbrod, 2003; but the patients showed reduced spatial negative Wagner, Loeper, Cohen, & Rockstroh, in press).
priming when spatial position was incidental and Other researchers also found preserved negative not task relevant. However, MacQueen, Galway, 1 The reported negative priming effects were descriptively large for schizophrenic patients classified as negative (255 ms, N ¼ 10) or positive (100 ms, N ¼ 8), for depressive patients (228 ms, N ¼ 21), and for control participants (69 ms, N ¼ 35). An analysis ofvariance (ANOVA) showed that the size of the effect did not differ significantly between groups. However, when the authors testedwhether the effect was significant within a particular group, statistically significant results emerged only for the depressive and controlparticipants. The problem with these group-specific tests and the conclusions drawn from them is that the sample size and, hence,the statistical power were so much smaller in the schizophrenic than in the other groups that results favouring the statistical nullhypothesis (“no negative priming”) are highly expected on purely methodological grounds and must not be interpreted as indicatingno negative priming in the schizophrenic groups.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) Goldberg, and Tipper (2003) reported reduced to classify, by an appropriate key press, the spatial negative priming of schizophrenic patients attended tone as originating from an instrument compared to control participants when spatial or an animal. Each trial consisted of a prime pair position was relevant and not incidental.
and a probe pair of stimuli. Trials were separated purpose of the present research was to add A preliminary version of this task using experi- further evidence to the body of available findings mental parameters that had previously been found on the relation between negative priming and appropriate for use with elderly participants schizophrenia. In particular, we wanted to extend (Buchner & Mayr, 2004) was implemented in a the range of available findings from the visual to pilot study to check its adequacy for a clinical the auditory domain. While this empirical exten- sample with diagnoses of schizophrenia. In this sion into a new sensory modality may be regarded pilot experiment, the participants’ task was to clas- as interesting in its own right, another purpose for sify the attended tones as belonging to either the moving into the auditory modality was to provide “wind instrument” (flute, trumpet, and saxophone) for a potentially more sensitive test of the hypoth- or the “string instrument” (piano, balalaika, and esis of reduced negative priming in schizophrenia.
pizzicato violin) category by responding with an Specifically, Banks, Roberts, and Ciranni (1995) appropriate key press. Each trial consisted of a argued that attending to a tone while avoiding prime pair and a probe pair of stimuli, with a cue – auditory distraction must operate almost entirely target interval of 500 ms and an interval of 500 ms by internal processing mechanisms. Visual selec- between reaction to prime and the following probe tion, in contrast, may be supported by peripheral mechanisms such as eye or head movements.
The main result of the pilot study was that the Auditory selection by internal distractor inhibition clinical sample had great difficulties performing a may thus be regarded as much more demanding, task that had been shown to be clearly feasible leaving little or no room for compensating for for elderly participants. In general, the patients any inhibitory deficits. As a consequence, reduc- produced huge numbers of errors. What is more, tions in the efficiency of inhibitory mechanisms of the 46 clinical participants, 12 prematurely may be more clearly apparent in the auditory terminated the experiment because they felt over- than in the visual domain, so that group differ- taxed. For the experiment reported here, therefore, ences in negative priming may be expected to be stimuli were selected from categories that were stronger and more reliable in the auditory than easier to distinguish (instruments and animal in the visual domain. Furthermore, given the sounds), and the temporal parameters between widespread occurrence of auditory hallucinations experimental events were modified to ensure that in schizophrenia, the auditory domain seemed patients could comply with the task without especially relevant for the investigation of being overtaxed, so as to avoid serious and attentional mechanisms in this patient group.
unwanted selection bias. It also seemed importantto take these steps in order to avoid possible per-ceptual problems that might prevent the patientsfrom fully processing the prime distractors, since this is thought to be a variable that could explainthe observation of no or reduced negative priming in schizophrenic patients (cf. Moritz In the task used here, participants heard pairs oftones displayed via headphones. One tone was presented to each ear. A click indicated the ear Participants for the clinical group were 28 psychia- that had to be attended. Participants were asked tric patients (10 female) who were recruited from THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) three different psychiatric hospitals and met the 1,949), and the distribution of dosages was posi- diagnostic criteria for schizophrenia according to tively skewed. A total of 10 patients also received ICD-10 (Dilling, Mombour, & Schmidt, 1993).
other medication (tranquillizers, sedatives, or anti- Six of the patients were outpatients.2 Diagnoses depressants), and 6 were also taking anticholinergic were established by experienced clinicians at the medication. Controls were 28 healthy participants hospitals. Furthermore, one of the authors and the (10 female) recruited mainly from the university responsible ward psychiatrists and psychologists staff and without a history of psychiatric illness.
carefully screened all available clinical records in Controls and schizophrenia participants were order to exclude patients with previous or concur- matched for age, gender, and years of education rent diagnoses that were incompatible with an as a measure of overall intellectual level, premorbid unambiguous diagnosis of schizophrenia. For an in the case of the patients (see Table 1). None of unambiguous diagnosis of schizophrenia, only the participants used hearing aids, and controls patients without double psychiatric diagnoses, and patients did not differ with respect to their schizoaffective disorders, or potentially drug- self-reported hearing ability (using a 3-point scale induced psychoses were accepted. On average the with “above average”, “normal”, and “below first known diagnosis of schizophrenia had average”), z ¼ 1.38, p . .17. None of the patients occurred 5.9 years (minimum, 0.5 years; maximum, and control persons was older than 45 years so 26 years; SD ¼ 5.7) before the time at which the that possible confusion with age-related effects on experiment was conducted. Thus, the sample of the size of the negative priming phenomenon patients had the advantage of being relatively hete- (cf. Verhaeghen & De Meersman, 1998) could be rogeneous with respect to the duration of their excluded, although in retrospect this additional illness, rendering it less likely that the findings of control may have been unnecessary given that a this experiment could be due to the specific charac- more recent meta-analysis suggests that there is teristics of a highly selective group of patients. The no age-related difference in negative priming responsible clinicians were requested to rate the patients on the Brief Psychiatric Rating Scale’s All participants were paid for their participation (BPRS, Overall & Gorham, 1962) German and gave informed consent previous to their version (Collegium Internationale Psychiatriae participation in the experiment. They were also Scalarum, 1996) based on their daily interaction explicitly informed that they could terminate the with patients and previous clinical interviews.
experiment at any time if they so wished.
The total sum of the 18 individual ratings can beinterpreted as an approximate total pathology score. For our clinical sample, the BPRS total The stimuli were six digitized tones, which could score mean was 48 (SD ¼ 9.9). All patients be identified and categorized easily and unambigu- received neuroleptic medication. Chlorpromazine ously as “musical instruments” (piano, guitar, and equivalents were calculated for 27 patients accord- clarinet) or “animal sounds” (duck, lamb, and ing to Jahn and Mussgay (1989) for conventional frog). Each tone was 300 ms long, complete with attack and decay. The participants heard the Woods (2003) for newer atypical antipsychotic tones over earphones that were fitted with high- medication. The mean daily dosage in chlorproma- isolation hearing protection covers and plugged zine units was 449.6 (minimum, 41; maximum, directly into an Apple PowerBook computer.
2 Note that removing the outpatients from the patient group did not change any of the statistical conclusions reported in this article, except for one secondary result pertaining to the supplementary analysis of the error data: The overall difference in errorsbetween ignored repetition (IR) and ignored repetition control (IRC) trials for the entire sample of schizophrenic patients andhealthy control participants was just barely statistically significant when the outpatients were included in the sample, and it narrowlymissed the preset level of statistical significance when the six outpatients were removed (see the Results section).
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) Table 1. Data characterizing the matched patient and control groups Trials consisted of a prime and a probe pair of the IR and its corresponding IRC trial. The stimuli. There were four basic types of trial: same holds for the AR and their corresponding ignored repetition trials (IR), ignored repetition control trials (IRC), attended repetition trials For each prime and each probe pair of stimuli, (AR), and attended repetition control trials (ARC).
only combinations of tones from the two cat- Each IR trial corresponded to one IRC trial in egories (animal sound or musical instrument) terms of the tone configuration except for the were used in order to necessitate real selection ignored prime, which differed but was taken for each category. In other words, combinations from the same category. Similarly, each AR trial of stimuli from one category were never presented corresponded to one ARC trial in terms of the together. Combining all six different tones in the tone configuration except for the attended prime, way illustrated in Table 2 yields 72 unique trials which differed but was taken from the same cate- of each type: IR, IRC, AR, and ARC, resulting gory. This is illustrated in Table 2. If negative, or in a total of 288 different trials. Of this original positive, priming occurs with this arrangement of pool of stimuli, a subset of 144 combinations (36 corresponding trials, then it must be due to pro- per trial type) was selected in such a way that cesses operating at the level of the stimulus iden- the absolute frequencies of the different tones tity of the tone and cannot be due to processes were equal overall as well as within each trial related to the response category. To illustrate, type. Furthermore, the frequencies of the combi- the guitar tone is ignored in the IR prime pair nations of attended and ignored tones were identi- example given in Table 2, and the piano tone is ignored in the corresponding IRC prime pair.
An important difference between IR and IRC Both belong to the same category of instruments.
trials on the one side and AR and ARC trials on If the reaction to the guitar tone was slowed the other is that the required response always in the IR relative to the IRC probe pair, then changed in the former and remained the same in the processes causing the slow-down must operate the latter types of trial. For the experimental pro- on the stimulus identity of the ignored prime, cedure, this has the desired consequence that the because that is the only difference between required reaction to the attended probe could not Table 2. Examples of stimulus configurations THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) be predicted from the prime. For the data analysis, to-be-attended probe tone was presented either the consequence is that IR trials can only be to the same ear as the prime or to the other ear.
compared to IRC trials, and AR trials can only The relation of the to-be-attended prime and probe location was varied in blocks. The block-wise presentation was a consequence of the pilot study that had shown that schizophrenic patients The task was introduced as a tone categorization find it difficult to cope with attended primes and task. Extensive practice was provided in order to probes appearing unpredictably at the same or at familiarize the participants with the task. They different ears. One half of the participants in were first introduced to the tones and the reaction each group first received the block of randomly keys by hearing each tone individually, and by ordered trials in which the attended primes reacting to single tones for each category presented and probes were presented to the same ear, fol- first to one ear, and then to the other. A metro- lowed by the remaining trials in which the nome click indicated the ear at which the tone attended primes and probes were presented to would be presented. Following this, participants different ears. For the other half or the participants were asked to react to 24 randomly presented in each group this sequence of blocks was reversed.
single tones. The last part of the practice session The ordering of blocks was counterbalanced over introduced the actual experimental task. Here subjects. Within each block, the ear at which the two different tones were presented to the right to-be-attended prime was presented was thus and left ear, and the preceding 20-ms metronome randomly selected but predicted the ear of the click now indicated which ear was to be attended to-be-attended probe. Within each trial type, the to (determined at random). Participants reacted attended primes and probes were presented to to consecutive pairs of stimuli—that is, a prime the same ear on 18 trials, whereas attended pair and a probe pair. After a 1,000-ms cue – primes and probes were presented to different target interval, a pair of tones was presented, one ears for the other 18 trials. This implies that to the left and one to the right ear. The interval on IR trials, the ignored prime changed location in the former case, but did not change location 500 ms. Participants reacted to the attended tone by pressing, as quickly as possible, the “instru- Prime or probe reactions faster than 100 ms and ment” or the “animal” key, depending on the slower than 4,000 ms were counted as invalid, and category to which the tone belonged.
the entire trial was repeated. After each trial, The experiment itself consisted of 144 trials, participants were given visual and acoustic feed- each of which was composed of a prime and a back on the correctness of their reactions to the probe pair of tones as in the final phase of the prac- prime and probe (visual feedback indicated tice session. The participants initiated each trial whether the reactions to prime and probe had at their own pace. A trial began with a brief been correct or false, and acoustic feedback was countdown followed by the click indicating the “very good” for two correct responses, “almost” ear on which the to-be-attended tone would be for one correct response, and “too bad” for two presented. The instructions emphasized correct- incorrect responses). A summary feedback about ness, but reactions were also to be made as error percentages and average reaction times was quickly as possible. The interval between partici- provided after every 12th trial, and to further pants’ reactions and the click preceding the probe motivate the participants they were informed of pair of tones was 500 ms. The probe tone pair how they had performed compared to the previous was presented with the same temporal parameters block of 12 trials. Once again, after the summary feedback, participants continued the experiment The to-be-attended ear for the prime pair of at their own discretion. On completion of stimuli was always selected at random. The the experiment, participants were given the THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) opportunity to be informed about the purpose of Probe reaction times were evaluated only for trialsin which both the probe and the prime reactions were correct. The means of participants’ average The design consisted of two 2 Â 2 Â 2 subdesigns.
reaction times and the corresponding error rates The ignored repetition subdesign comprised IR are presented in the upper and lower panels of versus IRC trials as the levels of the within- subject priming factor and same versus different Patients reacted more slowly than controls.
presentation sides of the attended prime and probe Reaction times on IR trials were longer than as the levels of the within-subject presentation side reaction times on IRC trials, and they were factor. The attended repetition subdesign differed longer when the attended prime was presented to by having AR versus ARC trials as the levels of the a different ear from that of the attended probe.
priming factor. Both subdesigns also comprised Corresponding to these descriptive data, a 2  2 the two levels, patients versus matched control  2 analysis of variance (ANOVA) with partici- persons, of the quasi-experimental between- pant group (schizophrenic patients vs. controls) subjects variable. The primary dependent variables as between-subjects variable, and priming (IR vs.
were participants’ reaction times, but error proba- IRC) as well as presentation side (same vs. differ- ent) as within-subject variables showed significant A power analysis with respect to the negative main effects of patient group, F(1, 54) ¼ 11.22, (or positive) priming effect showed that given p , .001, of priming, F(1, 54) ¼ 64.71, p , .001, a sample size of n ¼ 28 in each of the two and of presentation side, F(1, 54) ¼ 68.17, groups (i.e., N ¼ 56), and a ¼ .05, effects of size p , .001. The critical interaction of participant group and priming was not significant, F(1, 54) ¼ z ¼ .45 (cf. Cohen, 1977) could be detected with a probability of (1 – b) ¼ .95 (the sample 0.22, p . .64, as were all other interactions, negative priming effect actually turned out to be Fs(1, 54) , 1.90, p . .17. Follow-up t tests much larger than this assumed population value, showed that negative priming was significant for both the schizophrenic patients, t(27) ¼ 5.68, 28, and a ¼ .05, effects of size d ¼ .80 (“large” p , .001, and the control participants, t(27) ¼ effects in terms of the conventions suggested by 5.73, p , .001. In terms of the reaction times, Cohen, 1977) could be detected with a probability the mean and standard deviation of the negative of (1 – b) ¼ .91 for the one-tailed test of whether priming effect was somewhat larger for the negative priming is smaller in patients than in patients (M ¼ 88 ms, SD ¼ 82 ms) than for the controls. Assuming a population effect size of controls (M ¼ 78 ms, SD ¼ 72 ms). In terms of d ¼ .80 for the difference in negative priming standardized effect sizes, the sample effects for between patients and controls seemed conservative patients and control participants were quite large and reasonable given that the sample effect sizes in studies that reported an overall difference between patients and control participants, respectively.
these groups were typically much larger and This very close match across groups can be con- sidered strong evidence in favour of the hypothesis that the negative priming effect does not differ The level of a was set to .05 for all analyses with the reaction time data although only the 3 The power calculations were conducted using the G.Power program (Buchner, Faul, & Erdfelder, 1996; Erdfelder, Faul, & THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) Figure 1. Reaction times (upper panel) and error rates (lower panel) as a function of trial type and presentation side for schizophrenic patientsand control participants. The error bars depict the standard errors of the means.
difference in errors between IR and IRC trials was reactions on ARC trials. Again patients showed statistically significant, F(1, 54) ¼ 4.45, p ¼ .04.
slower reaction times than the control persons, Turning to the positive priming subdesign, we and reaction times were longer when the attended find that reactions on AR trials were faster than primes and probes were presented to different ears.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) A 2 Â 2 Â 2 ANOVA with participant group important conceptual replications and extensions (patients vs. control participants) as between- of previous findings of normal visual negative subjects variable, and priming (AR vs. ARC) as priming in schizophrenic patients (e.g., Baving well as presentation side (same vs. different) as et al., 2001; Moritz et al., 2001; Roesch-Ely within-subject variables showed significant main et al., 2003; Wagner et al., in press). If we also con- effects of participant group, F(1, 54) ¼ 14.27, sider that the original demonstration of reduced p , .001, of priming, F(1, 54) ¼ 39.55, p , .001, and of presentation side, F(1, 54) ¼ (Beech et al., 1989) seems to have been con- 90.35, p , .001. There was also an interaction founded with atypically impoverished presentation between the participant group and priming, conditions (cf. Moritz et al., 2001), then the con- F(1, 54) ¼ 4.66, p , .04, reflecting the fact that clusion suggests itself that attentional deficits positive priming was larger for schizophrenic associated with schizophrenia may not after all be reflected in a paradoxically “faster” performance than for control participants (M ¼ 55 ms, SD ¼ on ignored repetition trials. Instead, schizophrenic patients seem to show a slow-down in reactions scores were significantly different from zero, to previously ignored distractors in the negative t(27)s . 3.06, p , .01. The interaction between priming paradigm, which is comparable to that priming and presentation side was also significant, found with healthy participants. Does this neces- F(1, 54) ¼ 26.71, p , .001, indicating that positive sarily have to be counted as evidence against the priming was larger when attended primes and assumption of an impaired functioning of inhibi- probes were presented to the same ear (M ¼ tory attentional mechanisms in schizophrenia? 133 ms, SD ¼ 105 ms) than when they were pre- sented to different ears (M ¼ 35 ms, SD ¼ First, according to the inhibition model of 142 ms). No other interactions were significant, negative priming (Dalrymple-Alford & Budayr, 1966; Houghton & Tipper, 1994; Neill, 1977) With respect to the error data, there were the characteristic slow-down when reacting to a significant main effects of priming, F(1, 54) ¼ probe that was previously ignored is caused by 16.14, p , .001, and of presentation side, F(1, the need to overcome the inhibition imposed 54) ¼ 21.55, p , .001. An interaction between during prime selection. Therefore, one possible participant group and presentation side, F(1, conclusion is that the kinds of inhibitory processes 54) ¼ 4.30, p , .05, reflected that the presentation thought to be impaired in schizophrenia (“a defect side effect was larger for schizophrenic patients in the mechanism that controls and limits the (M ¼ .06, SD ¼ .20) than for control participants contents of consciousness”, cf. Frith, 1979, p. 225) are actually different from those involved in thesuppression of perceptual distractors in negativepriming tasks. Indeed, it may be an oversimplifica- tion to conceive of only one homogeneous type of The main result of the present experiment is that Second, one could assume that negative priming showed the same amount of negative priming in tasks do not really measure inhibition at all, and this auditory priming task, as is most obvious indeed there are alternative explanations of the from the very close match of these groups in basic negative priming phenomenon of which terms of the standardized negative priming feature mismatch, temporal discrimination, and effect. Considering that reductions in the effi- episodic retrieval appear to be most relevant (see ciency of inhibitory mechanisms may be more Fox, 1995; May et al., 1995; Milliken, Joordens, clearly apparent in the auditory than in the visual Merikle, & Seiffert, 1998; Neill et al., 1995). For domain (Banks et al., 1995), these results represent instance, J. Park and Kanwisher (1994) assumed THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) that negative priming was caused by a feature mis- remembered nonresponse information conflicts match between prime and probe—that is, by a with the response requirements implied by the change in the bindings of symbol identities to probe target and in that way slows down the locations between the prime and probe. In the probe responses. If such low-capacity, short-term present experiments, bindings of tone identities retrieval as specified by this model were intact in (e.g., the piano tone) to locations (e.g., the left schizophrenia, then we would expect the pattern ear) changed when, on ignored repetition trials, of findings reported in the present experiment.4 the prime distractor presented to one ear was also Thus, the only conclusion we can come to from the probe target subsequently presented to the the present findings is that negative priming as other ear (mismatch condition). In contrast, the an empirical phenomenon is not reduced in bindings stayed the same on those ignored rep- schizophrenia. We cannot reject, based on these etition trials on which prime and probe distractors findings, the assumption that attentional inhibi- were presented to the same ear (no mismatch con- tory functions are impaired in schizophrenia, dition). Obviously, the size of the negative priming either because there may be more than one type effect was completely unaffected by whether or not of inhibitory process, or because the characteristic a mismatch occurred, adding the present study to slow-down to previously ignored stimuli in nega- the growing body of evidence suggesting that tive priming tasks actually may not measure inhi- feature mismatching does not play an important bition. We have reason to believe that the latter role in negative priming (Baylis, Tipper, & alternative is quite unlikely, because existing evi- Houghton, 1997; Buchner & Mayr, 2004; dence suggests that inhibitory processes are at Buchner & Steffens, 2001; Buchner, Steffens, & least to some extent involved in the generation of Berry, 2000; Fuentes, Humphreys, Agis, Carmona, & Catena, 1998; Milliken, Tipper, & Weaver, present experimental task (Buchner & Steffens, 1994; Tipper, Weaver, & Houghton, 1994).
2001). However, it could well be that negative In contrast, the episodic retrieval model priming tasks do not yield the best measure of suggested by Neill and colleagues (Neill & inhibition, or at least not an exclusive one.
Valdes, 1992; Neill et al., 1995; Neill, Valdes, Another result worth mentioning is that the Terry, & Gorfein, 1992) could explain the positive priming effects were significantly larger current data pattern. According to this model the in schizophrenic patients than in healthy controls.
probe target cues the retrieval of the perceptually This replicates a data pattern first reported by similar prime display in which the distractor Baving et al. (2001; see also Wagner et al., in representation contains the information that no press) for a situation with relatively long stimulus response was (to be) made to that stimulus. This presentations in which both the prime and the 4 The so-called temporal discrimination account (Milliken, Joordens, Merikle, & Seiffert, 1998) may, in principle, also explain the present data, but past experiments suggest that the discrimination account does not seem adequate to explain negative priming inthe present paradigm, and it also does not add to the understanding of negative priming phenomena beyond what can already beexplained by the distractor inhibition and episodic retrieval models (Buchner & Steffens, 2001; Buchner, Zabal, & Mayr, 2003)which is why it is not mentioned in the main text. Within this framework, two classes of process are assumed to occur when aprobe response is generated. First, if the probe can be categorized as old, then automatic processes are likely to determine the responsein that the prior action is simply retrieved and executed. This explains fast responses on attended repetition trials. Second, if the probetarget is categorized as new so that prior learning is an inappropriate basis for action, complete perceptual analysis of the stimulus isnecessary in order to arrive at a response. This takes more time than simply retrieving a recent response and corresponds to controltrials. A third situation is given on ignored repetition trials. The target has been presented as part of the prime display, but it has notbeen fully attended. Thus, the probe target is somewhat familiar so that it cannot be quickly categorized as new, but it is also not quitefamiliar enough to be immediately categorized as old. According to Milliken et al. (1998), “this ambiguity in the temporal discrimi-nation process for ignored repetition trials is presumed to underlie negative priming” (p. 210). There is no reason to assume that theambiguity should be eliminated in schizophrenia, so that, in principle, this account could explain intact negative priming inschizophrenia.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) probe stimuli required a response. As Baving et al.
Baving, L., Wagner, M., Cohen, R., & Rockstroh, B.
also noted, positive priming is normal or reduced (2001). Increased semantic and repetition priming with short stimulus presentations and without in schizophrenic patients. Journal of Abnormal the requirement to respond to the primes. Thus, just as with negative priming, the relative size of Baylis, G. C., Tipper, S. P., & Houghton, G. (1997).
Externally cued and internally generated selection: the positive priming effect in schizophrenic Differences in distractor analysis and inhibition.
patients seems to depend, among other things, on whether the conditions for perceiving the Perception and Performance, 23, 1617 – 1630.
Beech, A. R., Powell, T., McWilliam, J., & Claridge, G.
The investigation of the causes of such positive (1989). Evidence of reduced “cognitive inhibition” in priming or repetition effect has a long history.
schizophrenia. British Journal of Clinical Psychology, It dates back to the work of Bertelson (e.g., Bertelson, 1961, 1965) who already argued that Bertelson, P. (1961). Sequential redundancy and speed the effect was primarily caused by processes in a serial two-choice responding task. Quarterly located at the level of response mechanisms and Journal of Experimental Psychology, 13, 90 – 102.
not by a speedup of signal processing. More Bertelson, P. (1965). Serial choice reaction-time as a function of response versus signal-and-response recently, Pashler and Baylis (1991) reaffirmed Bertelson’s earlier reasoning and refined it, con- Buchner, A., Faul, F., & Erdfelder, E. (1996). G.Power: cluding that, first, the repetition effect was very A priori, post-hoc, and compromise power analyses stimulus specific and, second, that it was due to for the Macintosh (Version 2.1.2) [Computer transient short-cuts in response selection for program]. Du¨sseldorf, Germany: Heinrich-Heine- immediately repeated stimuli (presumably a Universita¨t. Available from http://www.psycho.
direct translation from early stimulus represen- uni-duesseldorf.de/aap/projects/gpower/.
tations all the way to fairly specific responses) Buchner, A., & Mayr, S. (2004). Auditory negative and not due to faster perceptual processing of the priming in younger and older adults. Quarterly repeated stimuli. Thus, schizophrenic patients Journal of Experimental Psychology, 57A, 769 – 787.
seem to benefit more from such “short-cuts” than Buchner, A., & Steffens, M. C. (2001). Auditory negative priming in speeded reactions and temporal order judgements. Quarterly Journal of Experimental To summarize, then, the present findings from the auditory domain are consistent with a growing Buchner, A., Steffens, M. C., & Berry, D. C. (2000).
body of evidence from the visual domain showing normal rather than reduced or eliminated negative Extending and reversing the gender bias in fame priming in schizophrenic patients when stimulus judgments. Journal of Experimental Psychology: presentation parameters allow proper perceptual Learning, Memory, and Cognition, 26, 1215 – 1227.
identification by the patients, while at the same Buchner, A., Zabal, A., & Mayr, S. (2003) Auditory, time positive or repetition priming is enhanced.
Psychonomic Bulletin & Review, 10, 917 – 923.
Cohen, J. (1977). Statistical power analysis for the beha- Original manuscript received 20 December 2004 vioral sciences (Rev. ed.). Hillsdale, NJ: Lawrence PrEview proof published online 8 August 2005 Collegium Internationale Psychiatriae Scalarum (Ed.).
(4th ed.). Weinheim, Germany: Beltz.
Dalrymple-Alford, E. C., & Budayr, B. (1966).
Banks, W. P., Roberts, D., & Ciranni, M. (1995).
Examination of some aspects of the stroop color- Negative priming in auditory attention. Journal of word test. Perceptual and Motor Skills, 23, 1211–1214.
Experimental Psychology: Human Perception and David, A. S. (1995). Negative priming (cognitive inhibition) in psychiatric patients: Effects of neuro- THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) leptics. Journal of Nervous and Mental Disease, 183, MacQueen, G. M., Galway, T., Goldberg, J. O., & Tipper, S. P. (2003). Impaired distractor inhi- Dilling, H., Mombour, W., & Schmidt, M. H. (Eds.).
bition in patients with schizophrenia on a nega- tive priming task. Psychological Medicine, 33, diagnostische Leitlinien [German translation of May, C. P., Kane, M. J., & Hasher, L. (1995).
Chapter V (F) of the 10th revision of the Determinants of negative priming. Psychological International Classification of Diseases (ICD-10) of the WHO] (2nd ed.). Bern, Switzerland: Huber.
McGhie, A. (1977). Attention and perception in schizo- Erdfelder, E., Faul, F., & Buchner, A. (1996).
phrenia. In B. A. Maher (Ed.), Contributions to the GPOWER: A general power analysis program.
McGhie, A., & Chapman, J. (1961). Disorders of atten- Fox, E. (1995). Negative priming from ignored distrac- tion and perception in early schizophrenia. British tors in visual selection: A review. Psychonomic Bulletin Journal of Medical Psychology, 34, 103 – 115.
Milliken, B., Joordens, S., Merikle, P. M., & Frith, C. D. (1979). Consciousness, information proces- Seiffert, A. E. (1998). Selective attention: A reevalua- sing and schizophrenia. British Journal of Psychiatry, tion of the implications of negative priming.
Psychological Review, 105, 203–229.
Fuentes, L. J., Humphreys, G. W., Agis, I. F., Milliken, B., Tipper, S. P., & Weaver, B. (1994).
Carmona, E., & Catena, A. (1998). Object-based Negative priming in a spatial localization task: Feature mismatching and distractor inhibition.
Journal of Experimental Psychology: Human Learning Perception and Performance, 24, 664 – 672.
Gamboz, N., Russo, R., & Fox, E. (2002). Age differ- Moritz, S., Jacobsen, D., Mersmann, K., Kloss, M., & ences and the identity negative priming effect: An Andresen, B. (2000). Negative priming in schizo- updated meta-analysis. Psychology and Aging, 17, phrenia: No evidence for reduced cognitive inhi- bition. Journal of Nervous and Mental Disease, 188, Hoenig, K., Hochrein, A., Mu¨ller, D. J., & Wagner, M.
(2002). Different negative priming impairments Moritz, S., Ruff, C., Wilke, U., Andresen, B., Krausz, M., in schizophrenia and subgroups of obsessive- & Naber, D. (2001). Negative priming in schizo- compulsive disorder. Psychological Medicine, 32, phrenia: Effects of masking and prime presentation time. Schizophrenia Research, 48, 291–299.
Houghton, G., & Tipper, S. P. (1994). A model of Neill, W. T. (1977). Inhibitory and facilitatory processes inhibitory mechanisms in selective attention. In in selective attention. Journal of Experimental D. Dagenbach & T. H. Carr (Eds.), Inhibitory Psychology: Human Perception and Performance, 3, mechanisms of attention, memory, and language (pp. 53 – 112). San Diego, CA: Academic Press.
Neill, W. T., & Valdes, L. A. (1992). Persistence of Jahn, T., & Mussgay, L. (1989). Die statistische negative priming: Steady state or decay? Journal of Kontrolle mo¨glicher Medikamenteneinflu¨sse in Experimental Psychology: Learning, Memory, and experimentalpsychologischen Schizophreniestudien: Ein Vorschlag zur Berechnung von Chlorpromazi- Neill, W. T., Valdes, L. A., & Terry, K. M. (1995).
naquivalenten [Statistical control of possible drug Selective attention and the inhibitory control of effects in psychological research on schizophrenia: cognition. In F. N. Dempster & C. J. Brainerd A proposal for calculating chlorpromazine equiva- (Eds.), Interference and inhibition in cognition lents]. Zeitschrift fur Klinische Psychologie. Forschung (pp. 207 – 261). San Diego, CA: Academic Press.
Neill, W. T., Valdes, L. A., Terry, K. M., & Laplante, L., Everett, J., & Thomas, J. (1992).
Gorfein, D. S. (1992). Persistence of negative Inhibition through negative priming with Stroop priming: II. Evidence for episodic trace retrieval.
stimuli in schizophrenia. British Journal of Clinical Memory, and Cognition, 18, 993 – 1000.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7) Overall, J. E., & Gorham, D. R. (1962). The Brief Salo, R., Robertson, L. C., & Nordahl, T. E. (1996).
Psychiatric Rating Scale. Psychological Reports, 10, Normal sustained effects of selective attention are absent in schizophrenic patients withdrawn from Park, J., & Kanwisher, N. (1994). Negative priming medication. Psychiatry Research, 62, 121–130.
for spatial locations: Identity mismatching, not Salo, R., Robertson, L. C., Nordahl, T. E., & Kraft, L. W. (1997). The effects of antipsychotic Psychology: Human Perception and Performance, 20, medication on sequential inhibitory processes.
Journal of Abnormal Psychology, 106, 639 – 643.
Park, S., Lenzenweger, M. F., Pu¨schel, J., & Tipper, S. P. (1985). The negative priming effect: Holzman, P. S. (1996). Attentional inhibition Inhibitory priming by ignored objects. Quarterly in schizophrenia and schizotypy: A spatial nega- Journal of Experimental Psychology, 37A, 571 – 590.
tive priming study. Cognitive Neuropsychiatry, 1, Tipper, S. P. (2001). Does negative priming reflect inhibitory mechanisms? A review and integration of Park, S., Pu¨schel, J., Sauter, B. H., Rentsch, M., & conflicting views. Quarterly Journal of Experimental Hell, D. (2002). Spatial selective attention and inhi- bition in schizophrenia patients during acute psycho- Tipper, S. P., Weaver, B., & Houghton, G. (1994).
sis and at 4-month follow-up. Biological Psychiatry, Behavioural goals determine inhibitory mechanisms Pashler, H., & Baylis, G. C. (1991). Procedural Experimental Psychology, 47A, 809 – 840.
Verhaeghen, P., & De Meersman, L. (1998). Aging and the negative priming effect: A meta-analysis.
Psychology: Learning, Memory, and Cognition, 17, Psychology and Aging, 13, 435 – 444.
Wagner, M., Loeper, U., Cohen, R., & Rockstroh, B.
Roesch-Ely, D., Spitzer, M., & Weisbrod, M. (2003).
(in press). Visuo-spatial negative priming in schizo- Cognitive inhibition and thought disorder in schizo- phrenic patients and healthy controls: A behavioral phrenia. Psychopathology, 36, 23–32.
Salo, R., Henik, A., Nordahl, T. E., & Robertson, L. C.
(2002). Immediate versus sustained processing in Woods, S. W. (2003). Chlorpromazine equivalent doses for the newer atypical antipsychotics. Journal of Neuropsychological Society, 8, 794 – 803.
Clinical Psychiatry, 64, 663 – 667.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2006, 59 (7)

Source: http://www.psychologie.hhu.de/fileadmin/redaktion/Oeffentliche_Medien/Fakultaeten/Mathematisch-Naturwissenschaftliche_Fakultaet/Psychologie/AAP/Publikationen/2006/Zabal_Buchner__2006_.pdf