Negative priming of unattended part primes ...

This article was downloaded by: [Bangor University] On: 23 November 2011, At: 08:28 Publisher: Psychology Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

The Quarterly Journal of Experimental Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/pqje20

Negative priming of unattended part primes: Implications for models of holistic and analytic processing in object recognition a

b

Lina I. Conlan , Julian C. Phillips & E. Charles Leek

a

a

Wolfson Centre for Clinical and Cognitive Neuroscience, School of Psychology, Bangor University, UK b

Department of Social and Psychological Sciences, Edge Hill University, Lancashire, UK Available online: 06 Nov 2009

To cite this article: Lina I. Conlan, Julian C. Phillips & E. Charles Leek (2009): Negative priming of unattended part primes: Implications for models of holistic and analytic processing in object recognition, The Quarterly Journal of Experimental Psychology, 62:12, 2289-2297 To link to this article: http://dx.doi.org/10.1080/17470210903104420

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY 2009, 62 (12), 2289 –2297

Short article Negative priming of unattended part primes: Implications for models of holistic and analytic processing in object recognition Lina I. Conlan

Downloaded by [Bangor University] at 08:28 23 November 2011

Wolfson Centre for Clinical and Cognitive Neuroscience, School of Psychology, Bangor University, UK

Julian C. Phillips Department of Social and Psychological Sciences, Edge Hill University, Lancashire, UK

E. Charles Leek Wolfson Centre for Clinical and Cognitive Neuroscience, School of Psychology, Bangor University, UK

The “hybrid” model of object recognition (Hummel, 2001) proposes that unattended objects are processed holistically, while attended objects are processed both holistically and analytically. Supporting evidence for this claim was reported by Thoma, Hummel, and Davidoff (2004) who showed that, unlike whole object primes, unattended split object parts (presumed to require analytic processing) do not elicit repetition priming. Here we tested the generality of this finding by contrasting priming for whole and part prime stimuli as a function of prime informativeness and by modifying the design so that both unattended whole and part prime displays contained a single perceptual object. Unlike Thoma et al. (2004) the results showed negative (rather than an absence of) priming for unattended half object primes. These findings place new constraints on theoretical models of the role of attention in object recognition. Keywords: Object recognition; Negative priming; Attention.

Object recognition is a fundamental cognitive ability that underlies a broad range of human activity. An important issue that is currently attracting considerable interest is the role of attention in object recognition (e.g., Dux & Harris,

2007; Harris, Dux, Benito, & Leek, 2008; Hummel, 2001; Hummel & Stankiewicz, 1996, 1998; Murray, 1995; Thoma, Davidoff, & Hummel, 2007; Thoma, Hummel, & Davidoff, 2004; Vecera, Behrmann, & Filapek, 2001).

Correspondence should be addressed to Charles Leek, Wolfson Centre for Clinical and Cognitive Neuroscience, School of Psychology, Bangor University, Bangor, Gwynedd, LL57 2AS, UK. E-mail: [email protected] The authors would like to thank John Hummel and two anonymous reviewers for their invaluable comments and suggestions on an earlier version of this manuscript. # 2009 The Experimental Psychology Society http://www.psypress.com/qjep

2289 DOI:10.1080/17470210903104420


CONLAN, PHILLIPS, LEEK

Hummel and colleagues (Hummel & Stankiewicz, 1996, 1998; Thoma et al., 2007; Thoma et al., 2004) have described one hypothesis, known as the “hybrid” model, which proposes that recognition is mediated by both “holistic” image-based representations and “analytic” structural descriptions. On this account, attention plays an important role in the “dynamic” binding of information about object shape and spatial configuration during access to analytic structural descriptions (Hummel, 2001; Hummel & Biederman, 1992; Hummel & Stankiewicz, 1996, 1998). In contrast, image-based holistic recognition does not require binding since the representations do not separately encode feature dimensions such as shape and configuration. Thus, the hybrid model predicts that attended images can be processed both analytically and holistically, while unattended images are only processed holistically.

One source of evidence in support of this proposal comes from a study by Thoma et al. (2004). They investigated the extent to which attended or unattended primes facilitate the subsequent identification of whole objects. Prime displays consisted of either whole objects or two split object parts (see Figure 1a). Participants named one of the two primes indicated via a cueing box. A second whole object or probe was presented at the end of the trial. This probe item could either be related to the attended or unattended prime or unrelated to either. The results showed faster responses in probe identification for related trials in both attended conditions. In contrast, while unattended whole images elicited small but significant positive priming effects, unattended split part images showed no priming at all. It was suggested that the absence of priming for the unattended split part primes shows that attention is required to support analytic processing.

Figure 1. Schematic outline of the trial sequences in the “unattended” condition for (a) the original experiment reported by Thoma et al. (2004, Experiment 1) and (b) the current study. Participants first named the cued prime image following presentation of the prime. The probe was named following probe onset. The key difference between the two procedures is whether the part prime stimulus consisted of a dual split part image (as in Thoma et al., 2004), or a single half image (current study). The prime mask was followed by a blank interstimulus interval (ISI) of 1,995 ms. The probe mask was followed by a blank ISI of 2,000 ms and a subsequent feedback display. The effective response deadline for the prime was 2,610 ms (prime to fixation stimulus onset asynchrony, SOA) and for probe naming was 2,150 ms (probe to feedback SOA).

2290

THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (12)


NEGATIVE PRIMING OF UNATTENDED PRIMES

However, there is one methodological difference between the whole and split part priming conditions that might provide an alternative explanation for the results. As shown in Figure 1a, in the split part priming condition two object halves were shown, while only one whole stimulus was shown in the intact condition. The reduced priming in the unattended split part condition may thus have been a consequence of a capacity limitation in processing multiple objects without attention such that observers may have only been able to reliably process one of the two split object primes in the unattended condition. It is also relevant that the two split parts varied in the extent to which they predicted target identity. This is apparent, for example, when one considers how informative the front (i.e., head) and rear (tail) are in predicting the identity of four-legged animals (generally speaking, heads are more informative than bottoms). It follows, on this account, that on any given trial processing of unattended split parts would sometimes have involved predictable parts and on other trials unpredictable part primes. Overall mean responses in this condition would therefore represent a summation of individual effects for predictive and nonpredictive parts, reducing the mean priming relative to the unattended whole object condition—consistent with the pattern that was reported. Given current interest in the role of attention in object recognition, and the specific theoretical claims of the hybrid model, the aim of the current study was to directly examine the contribution of these factors to the pattern of results reported by Thoma et al. (2004). The key issue is whether the same results would be obtained for unattended primes when the whole and part prime conditions are equated in terms of the number of distinct perceptual objects that are

shown (see Figure 1b) and whether the pattern of priming in the unattended condition is dependent on prime predictability.

Method Participants A total of 44 Bangor University undergraduates (13 males; mean age ¼ 22.1 years, SD ¼ 1.66) participated in exchange for printer credits. All participants were first language English speakers with normal or corrected to normal vision.1 Stimuli and apparatus The experiment used the same E-Prime programs and stimuli as those used by Thoma et al. (2004).2 The stimulus set consisted of 84 blackand-white line drawings of objects taken from Snodgrass and Vanderwart (1980). All images were standardized in size to subtend 48 of visual angle. For each intact image there was a corresponding split object prime identical to those used in the original study (see Figure 1a). An additional set of part primes was created, consisting of part primes showing only one of the two split parts. To determine half object prime predictability (i.e., the extent to which each half predicts whole object identity) a group of 20 observers were asked to name each prime and to indicate their naming confidence on a 5-point Likert scale. These data were used to create two lists of highpredictability and low-predictability primes. Mean naming accuracy scores for the high-predictability and low-predictability sets were .84 and .64 respectively, t(167) ¼ 3.75, p , .001. Reported confidence ratings were also greater for highpredictability (M ¼ 4.6, SD ¼ 0.48) than for low-predictability lists (M ¼ 3.8, SD ¼ 1.07); t(167) ¼ 6.76, p , .001.

1 A further 12 participants were excluded due to high error rates in naming probe items (.60%). This may have been due to the relatively high proportion of bilingual English–Welsh speakers in the native population. For this reason a strict criterion for accuracy in probe naming was adopted in order to ensure consistency of responses across subject groups. To avoid ambiguity, and to ensure consistency, all responses where the exact English language target name was not given as the first utterance within the deadline were treated as errors (including synonyms). 2 We are grateful to Volker Thoma (University of East London) for providing the original stimulus materials and E-Prime programs used in their original study for this purpose.


2291



Image displays were viewed on a high-resolution CRT monitor at a screen resolution of 1,024 768 pixels from a viewing distance of 90 cm as in the original study. Design A 2 (prime predictability: high, low) 2 (relatedness: primed, unprimed) 2 (prime type: attended, unattended) 2 (prime shape: whole, part—split or half image) mixed design was used. Prime predictability was a between-subjects factor. Prime shape was nested since half images were only shown in the unattended part prime condition. The priming conditions were: attended –whole, attended –split object, unattended –whole, unattended – half object. For each there was a corresponding unprimed baseline condition in which the same probe stimuli were presented, but in which they were preceded by an unrelated prime in either the corresponding attended or unattended trial condition. As in Thoma et al. (2004) the stimuli were divided into 14 clusters of 6 different images from the 84-image stimulus set. These clusters were used to distribute stimuli across conditions and participants such that no object was seen more than once by any participant. As in the original study, participants completed 36 experimental trials with 6 trials for each of the four primed conditions (N ¼ 24) and 3 trials for each unprimed baseline (N ¼ 12). In order to contrast priming effects as a function of part prime predictability for the same probe stimuli, participants were tested in two groups to prevent presentation of both the high- and low-predictable

pairs of the same items to the same participant. Trial order was randomized across participants. Procedure The procedure was identical to that described by Thoma et al. (2004, Experiment 1). The sequence of trial events and timing parameters are shown in Figure 1b. Participants first completed 18 practice trials (with stimuli not used in the experimental trials).

Results Preliminary analysis and error rates Voice key errors (M ¼ 2.8%, SD ¼ 3.66% of all trials) and timeouts (M ¼ 4.7%, SD ¼ 4.30% of all trials) for naming responses were removed from the dataset. A strict criterion for accuracy in prime and probe naming was adopted in order to ensure consistency in responses across participants. All cases where the exact target name was not given as the first utterance within the response deadline were treated as errors—including synonyms. Mean error rates per condition are shown in Table 1. The mean global error rate (trials in which either the attended, cued, prime or probe response were incorrect) was 20.4% (SD ¼ 17.08%). Mean global error rates for predictable and nonpredictable trials were 18.2% (SD ¼ 16.4%) and 22.6% (SD ¼ 17.70%), respectively. This difference was not significant ( p . .02). Error rates were analysed in a 2 (predictability: high, low) 2 (prime type: attended, unattended) 2 (prime shape: whole, part) mixed analysis of variance (ANOVA). This showed main effects of prime type,

Table 1. Mean error rates per participant for each condition. Attended Predictability High Low

Unattended

Unprimed

Whole

Part (split)

Whole

Part (half)

Whole

Part

8.33 (12.3) 14.39 (14.8)

18.18 (15.3) 25.00 (19.7)

21.21 (18.6) 18.94 (18.0)

21.21 (17.2) 23.4 (15.1)

15.5 (15.3) 21.2 (16.4)

25.0 (19.7) 32.5 (22.1)

Note: Data show the percentage of total trials in which either (or both) attended (cued) prime and probe items were incorrectly named. Standard deviations are shown in parentheses.

2292




F(1, 42) ¼ 4.25, p , .04, consistent with more errors on unattended than on attended prime trials, and prime shape, F(1, 42) ¼ 6.25, p , .01, with more errors following part primes than whole object primes. There were no other significant main effects or interactions. There was no significant correlation between mean reaction times (RTs) per condition and error rates (r 2 , .01) and no indication of a speed– accuracy trade-off. Analyses of naming latencies and priming effects Mean naming latencies (by participants across mean RTs per condition) for correct trials as a function of condition are shown in Table 2. Following the procedures described in Thoma et al. (2004), priming effects were calculated by subtracting mean RTs per condition from the corresponding unprimed baseline conditions. The pattern of priming effects across conditions is shown in Figure 2. Mean RTs were initially analysed in a global 2 (predictability: high, low) 2 (relatedness: primed, unprimed) 2 (prime type: attended, unattended) 2 (prime shape: whole, part) mixed ANOVA. This showed main effects of relatedness, F(1, 42) ¼ 36.28, p , .001; and prime type, F(1, 42) ¼ 65.49, p , .001. There were also significant interactions between relatedness and prime type, F(1, 42) ¼ 13.46, p , .001, and between relatedness and prime shape, F(1, 42) ¼ 7.06, p , .01. The three-way interaction of Relatedness Prime Type Prime Shape

was marginally significant, F(1, 42) ¼ 13.31, p , .07 (two-tailed). Further planned analyses were conducted on mean RTs for the highpredictability and low-predictability conditions. For the low-predictability primes a 2 (relatedness: primed, unprimed) 2 (prime type: attended, unattended) 2 (prime shape: whole, part) repeated measures ANOVA showed main effects of relatedness, F(1, 21) ¼ 23.59, p , .001, and prime type, F(1, 21) ¼ 35.66, p , .001. There was also a significant interaction between relatedness and prime type, F(1, 21) ¼ 41.84, p , .001. Decomposition of the interaction for the attended prime trials using a 2 (relatedness: primed/unprimed) 2 (prime shape: whole, split object) repeated measures ANOVA showed a main effect of relatedness, F(1, 21) ¼ 46.8, p , .001. Simple effects analyses showed that the relatedness effect was significant for both attended whole, t(21) ¼ 4.48, p , .001, and attended split parts, t(21) ¼ 6.06, p , .001. In contrast, the same analyses carried out on mean RTs for the low-predictability unattended prime trials showed no significant main effects or interactions. However, planned comparisons showed a significant simple effect of priming for the unattended whole primes, t(21) ¼ 3.5, p , .001. There was no significant priming effect for the unattended half primes. This pattern of results replicates the data reported by Thoma et al. (2004)—that is, larger priming effects for attended than unattended primes and weaker,

Table 2. Participant mean correct probe naming latencies for high-predictability and low-predictability primes as a function of priming condition. Attended Predictability

Unattended

Whole

Split part

Whole

Half

High

Primed Unprimed Priming effect

552.0 (86.9) 698.1 (119.2) 146.1 (151.8)

543.5 (97.2) 704.3 (209.1) 160.8 (236.5)

670.4 (105.7) 732.7 (203.6) 62.3 (162.5)

741.5 (120.5) 679.1 (83.9) –62.4 (117.2)

Low

Prime Unprimed Priming effect

557.2 (77.9) 707.8 (131.9) 150.6 (157.5)

566.2 (86.4) 717.9 (140.2) 151.7 (117.4)

643.0 (104.1) 719.2 (116.4) 76.2 (101.3)

693.9 (124.4) 693.6 (136.2) –0.3 (158.1)

Note: Participant mean correct probe naming latencies, in ms. Means of mean RTs across conditions are shown. Standard deviations are in parentheses. Priming effects are shown—see also Figure 2. THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (12)

2293



Figure 2. Mean priming effects for (a) low-predictability and (b) high-predictability half prime conditions. Bars show standard error of the mean. Whole conditions shown in black, part prime conditions in light grey. Att ¼ attended, Unatt ¼ unattended priming conditions.

but significant, priming for unattended whole object primes only. There was no evidence of priming for the low-predictability unattended single half primes. The same analyses were conducted on mean RTs for the high-predictability set. Here a 2 (relatedness: primed, unprimed) 2 (prime type: attended, unattended) 2 (prime shape: whole, part) repeated measures ANOVA showed main effects of relatedness, F(1, 21) ¼ 13.70, p , .001, and prime type, F(1, 21) ¼ 34.20, p , .001. There were also significant interactions between relatedness and prime type, F(1, 21) ¼ 51.81, p , .001, and between relatedness and prime shape, F(1, 21) ¼ 4.05, p , .05.

2294

As expected, separate analyses of the attended prime trials showed a pattern similar to that for the low-predictability group (recall that prime predictability was only manipulated in the unattended conditions). A 2 (relatedness: primed/unprimed) 2 (prime shape: whole, split part) repeated measures ANOVA showed a main effect of relatedness, F(1, 21) ¼ 32.07, p , .001, only. Simple effects analyses revealed that this effect was significant for both attended whole, t(21) ¼ 4.51, p , .001, and attended split parts, t(21) ¼ 3.18, p , .001. There was no interaction in the size of the priming effects for attended primes as a function of prime type. Again, this replicates the pattern for attended primes reported by Thoma et al. (2004). In contrast, the high-predictability unattended prime trials showed a strikingly different pattern. Here there was a significant interaction between relatedness (primed, unprimed) and prime type (whole, half object), F(1, 21) ¼ 6.92, p , .01. The reason for this interaction can be clearly seen in Figure 2b. Critically, the unattended whole object primes show a positive priming effect of þ 62.3 ms, t(21) ¼ 1.7, p , .04, while unattended half object primes show a negative priming effect of –62.4 ms, t(21) ¼ 2.49, p , .01. Unlike the data reported by Thoma et al. (2004), these results provide evidence that unattended part stimuli can indeed prime subsequent recognition of probe target objects providing that the parts are predictive of object identity.

GENERAL DISCUSSION Using the same design, programme, and stimulus materials, we replicated the patterns of priming observed for attended whole and split objects reported by Thoma et al. (2004). Unlike the original study, we used a modified version of the unattended condition in which either high- or low-predictable single half objects were presented rather than split object primes. The results showed two new findings: First, the magnitude of priming in the unattended condition for single half objects was modulated by predictability. Low-predictable unattended single half objects




did not elicit priming. In contrast, we found that predictive unattended single half objects, unlike the split object primes in the original study, elicited priming for subsequent object recognition. Second, we found that these predictive unattended single half object primes produced a negative, rather than a positive, priming effect. These data provide new evidence about the processing of unattended shape information. In terms of the hybrid model the implications of the data depend, in part, on whether the unattended single half objects are assumed to require holistic or analytic processing. If the single half objects are assumed to require analytic processing in order to activate subsets of shape information in a structural description, then the finding of priming for unattended half objects presents a challenge to the hypothesis that attention is required for analytic processing. The original findings of no priming for unattended split parts in the study by Thoma et al. (2004) might then, as suggested earlier, be potentially interpreted by a combination of two factors: a capacity limitation preventing processing of both unattended split parts and prime shape informativeness—that is, the extent to which the processed split part predicts (probe) object identity. Thus, the absence of priming in the split part condition in Thoma et al. (2004) may have resulted from the processing of uninformative part primes on some trials weakening any overall priming effects.3 Alternatively, one might argue that single object fragments (such as the half object primes) are processed as holistic perceptual units and that the priming effect derives from partial activation of a holistic representation in the unattended condition. This possibility is consistent with the

proposal by Hummel (2003) that holistic components may consist of any subset of image features (such as a part) “when the visual system fails to segment the object into independent parts. . .” (ibid. p. 224). Here, the observation of negative priming in the unattended condition for half object primes and positive priming for whole object primes is potentially relevant. If both types of prime are processed holistically using the same mechanism, it is unclear why they should elicit priming in opposite directions. One possible explanation for this stems from previous accounts of negative priming. Positive and negative priming have been assumed to reflect differential preactivation of internal mental representations by facilitatory and inhibitory processes (e.g., Houghton & Tipper, 1994; Park & Kanwisher, 1994; Tipper, 1985, 2001). In the context of the present study, the contrasting patterns of priming may be related to differences in the magnitude of activation elicited by unattended half and whole object primes. Weaker (i.e., subthreshold) preactivation of object representations from the partial shape information in unattended half primes may be insufficient to overcome subsequent inhibition of the representation, slowing its reactivation by a probe item. In contrast, stronger (i.e., suprathreshold) activation by unattended whole object primes (by virtue of greater feature overlap) may be sufficient to overcome such inhibitory activity resulting in a net facilitatory (positive priming) effect.4,5 In relation to the hybrid model, one possibility is that negative “holistic” priming could be accommodated through the incorporation of lateral inhibitory connections among units at the level of the “holistic surface map” (Hummel & Stankiewicz, 1996).

3 Against this “capacity limitation” account are findings of the absence of priming from single inverted whole unattended objects, rather than solely split parts (e.g., Thoma et al., 2007). However, other data have shown orientation invariant priming from unattended misoriented stimuli (e.g., Dux & Harris, 2007). 4 These contrasting priming effects in the unattended condition are also potentially relevant in the context of recent observations from other paradigms investigating object-based visual selection (e.g., Leek, Reppa, & Tipper, 2003; Vecera et al., 2001). For example, Leek et al. have shown how facilitation and inhibition are differentially constrained by occluding contour and internal part boundaries of objects in spatial cueing tasks. In the present task the observation of negative priming for half object primes is consistent with a differential distribution of inhibition and facilitation across object parts and wholes. 5 We are grateful to Steve Tipper and Paloma Mari-Beffa for their helpful discussions on this aspect of the data.


2295



More generally, the current findings add to the growing body of evidence concerning the way in which unattended objects are perceptually processed. This evidence remains controversial, in part, because of seemingly conflicting results. For example, in a series of studies Dux and Harris (2007) and Harris et al. (2008; see also, Murray, 1995) have shown that object shape identity can be computed from misoriented stimuli in the absence of attention—findings that imply analytic processing of unattended objects, but which also appear to be at odds with data from other studies suggesting a failure of recognition for unattended but inverted familiar objects (Thoma et al., 2007). A straightforward resolution of these broader findings is beyond the scope of the current study. However, the observation here of negative priming for unattended half object primes provides new evidence of direct relevance to understanding the role of attention in object identification. As currently formulated, the hybrid model cannot account for this effect (though it could potentially be modified to do so), and these data provide a new constraint on future theory development. Original manuscript received 11 February 2009 Accepted revision received 1 June 2009 First published online 17 July 2009

REFERENCES Dux, P. E., & Harris, I. M. (2007). Viewpoint costs occur during consolidation: Evidence from the attentional blink. Cognition, 104(1), 47 – 58. Harris, I. M., Dux, P. E., Benito, C. T., & Leek, E. C. (2008). Orientation sensitivity at different stages of object processing: Evidence from repetition priming and naming. PloS One, 3(5), e2256. Houghton, G., & Tipper, S. P. (1994). A model of inhibitory mechanisms in selective attention. In D. Dagenbach & T. H. Carr (Eds.), Inhibitory processes in attention, memory and language (pp. 53 – 112). San Diego, CA: Academic Press.

2296

Hummel, J. E. (2001). Complementary solutions to the binding problem in vision: Implications for shape perception and object recognition. Visual Cognition, 8, 489– 517. Hummel, J. E. (2003). The complementary properties of holistic and analytic representations of shape. In. M. A. Peterson & G. Rhodes (Eds.), Perception of faces, objects and scenes: Analytic and holistic processes. New York: Oxford University Press. Hummel, J. E., & Biederman, I. (1992). Dynamic binding in a neural network for shape recognition. Psychological Review, 99, 480– 517. Hummel, J. E., & Stankiewicz, B. J. (1996). An architecture for rapid, hierarchical structural description. In T. Inui & J. McCelland (Eds.), Attention and Performance. XVI: On information integration in perception and communication (pp. 93 –121). Cambridge, MA: MIT Press. Hummel, J. E., & Stankiewicz, B. J. (1998). Two roles for attention in shape perception: A structural description model of visual scrutiny. Visual Cognition, 5, 49 – 79. Leek, E. C., Reppa, I., & Tipper, S. P. (2003). Inhibition-of-return for objects and locations in static displays. Perception and Psychophysics, 65, 388– 395. Murray, J. E. (1995). Negative priming by rotated objects. Psychonomic Bulletin & Review, 2, 534– 537. Park, J., & Kanwisher, N. (1994). Negative priming for spatial locations: Identity mismatching, not distractor inhibition. Journal of Experimental Psychology: Human Perception and Performance, 20, 613 – 623. Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174 – 215. Thoma, V., Davidoff, J., & Hummel, J. E. (2007). Priming of plane-rotated objects depends on attention and view familiarity. Visual Cognition, 15, 179– 210. Thoma, V., Hummel, J. E., & Davidoff, J. (2004). Evidence for holistic representations of ignored images and analytic representation of attended images. Journal of Experimental Psychology: Human Perception and Performance, 30, 257– 267.



of conflicting views. Quarterly Journal of Experimental Psychology, 54A, 321– 343. Vecera, S. P., Behrmann, M., & Filapek, J. C. (2001). Attending to the parts of a single object: Partbased selection limitations. Perception & Psychophysics, 63, 308– 321.


Tipper, S. P. (1985). The negative priming effect: Inhibitory effects of ignored primes. Quarterly Journal of Experimental Psychology, 37A, 571 – 590. Tipper, S. P. (2001). Does negative priming reflect inhibitory mechanisms? A review and integration


2297