The authors investigated whether effects of the Implicit Association Test (IAT) are influenced by salience asymmetries, independent of associations. Two series of experiments analyzed unique effects of salience by using nonassociated, neutral categories that differed in salience. In a 3rd series, salience asymmetries were manipulated experimentally while holding associations between categories constant. In a 4th series, valent associations of the target categories were manipulated experimentally while holding salience asymmetries constant. Throughout, IAT effects were found to depend on salience asymmetries. Additionally, salience asymmetries between categories were assessed directly with a visual search task to provide an independent criterion of salience asymmetries. Salience asymmetries corresponded to IAT effects and also accounted for common variance in IAT effects and explicit measures of attitudes or the self-concept.

Priming effects of ignored distractor words were investigated in a task-switching situation that allowed an orthogonal variation of priming and response compatibility between prime and probe. Across 3 experiments, the authors obtained a disordinal interaction of priming and response relation. Responding was delayed in the ignored repetition condition if different responses were required for identical stimuli in the prime and probe (negative priming). Repeating the prime distractor in the probe facilitated responding if the same response was required in the prime and in the probe (positive priming). The same pattern of results was replicated in a letter-matching task without task switching (Experiment 4). Findings lend support to a new model that explains negative priming in terms of an automatic retrieval of incidental stimulus-response associations.

The automatic processing of information was investigated, varying valence (positive vs. negative) and relevance (other-relevant traits [ORT] vs. possessor-relevant traits [PRTj; G. Peeters, 1983) of stimuli. ORTs denote unconditionally positive or negative consequences for persons in the social environment of the holder of the trait (e.g., honest, brutal) whereas PRTs denote unconditionally positive or negative consequences for the trait holder (e.g., happy, depressive). In 2 experiments using the Stroop paradigm, larger interference effects were found for ORTs than PRTs. This is due to the behavior-relatedness of ORTs. In a go/no-go lexical decision task (Experiment 3), participants either had to withdraw their finger from a pressed key (i.e., "avoid") or had to press a key (i.e., "approach") if a word was presented. Responses to negative ORTs were relatively faster in the withdraw condition, whereas positive ORTs were relatively faster in the press condition.

Response retrieval theories assume that stimuli and responses become integrated into "event files" (Hommel, 1998) in memory so that a second encounter with a specific stimulus automatically retrieves the response that was previously associated with this stimulus. In this article, we tested a specific prediction of a recent variant of stimulus retrieval theories as introduced by Rothermund, Wentura, and De Houwer (2005): In selection tasks where target stimuli are accompanied by distractors, responses to target stimuli are automatically bound to distractor stimuli as well; repeating the distractor should retrieve the response to the target that formerly accompanied the distractor. In three experiments we confirmed this prediction: Distractor repetition facilitated responding in the probe in the case of response repetition whereas repeating the distractor delayed responding in the case of response change.

T h e a ffect ive co n gru en cy effect , wh er ein sh o r t er r esp o n se lat encies a re o b ser ved fo r affect ively co n gr u en t p rim e-t a r get p a irs in t h e eva lu a tio n ta sk (i.e. t h e ta r get is eva lu at ed a s p o sit ive o r n ega t ive) h a s o ften b een in t er p ret ed in t er m s o f a sp rea d o f a ct ivat io n fro m t h e p r im e t o a ffect ively co n gr u en t ta rget s. I n th e p resen t a r t icle, it is ar gu ed t h at t h e effect m igh t b e d u e to a co n¯ict b et ween t h e resp o n ses t h at a re a ct iva ted by t h e p r im e a n d t a r get , a ssu m in g t h at t h e p rim e ser ves a s a d istr a ct o r fo r p ro cessing th e t a rget. I f su ch a con¯ict o ccu r s, t h is wo u ld resu lt in a n egat ive p r im in g effect , th a t is, a n a ffect ive in co n gr u en cy o f p r im e (e.g. death) an d t ar get (e.g. wise) o n t r ia l n-1 (p r im e-t r ia l) will resu lt in a slo win g o f t h e resp o n se o n tr ia l n (p ro b e-t r ia l) if th e p rob e t a rget (e.g. lonely) is a ffect ively co n gr u en t t o t h e d istr a ct o r o f t h e p rim e-t r ial. T h is hyp o t h esis wa s co n ® r m ed in E xp er im en t 1 (N = 35) wit h a seq u en t ia l p resen t a tio n o f d ist r a ct o r a n d t a rget (i.e. SO A = 300m sec) a n d fu r t h er co rro b o r at ed by E xp er im en t 2 (N = 72) w ith a fact o r ia l m a n ip u lat io n o f seq u ent ia l a nd sim u lta neo u s (SO A = 0m sec) p resen t at io n . T h ese r esu lts favo u r a view t h at a ffect ive co n gru en cy effect s in t h e eva lu at io n t a sk a r e d u e to r esp o n se p at h in t er fer en ce p ro cesses wh ich a r e reso lved b y an in h ib itio n o f th e t en d en cy t o r esp o n d t o t he d ist r a cto r. C O G N I T I O N A N D E M O T I O N , 1999, 13 (1), 65±91R eq uest s fo r r epr int s sh o u ld b e sen t to D r D irk Wen tu ra , P sych o logisch es I n st itut I V, Westfa È lisch e W ilh elm s-U n iversita È t, F liedn erst r. 21, D -48149 M u È nst er, G er m an y; e-m a il: wentu r a @ psy.un i-m u enst er.d e.Po r tio n s o f th e r esult s wer e rep o rt ed at t h e 38th E xp erim ent a l P sych olo gy M eet ing in A pr il 19 96 a t E ich st a È tt , G erm a ny, an d at t he 40th C o ngr ess o f t h e D eu tsch e G esellscha ft fu È r P sycho logie, M u n ich, G erm a ny, Sep tem ber 1996.T h is r esea r ch wa s su p p or ted b y a gr an t fro m t he F o un d atio n fo r P ro m ot io n o f t he U n iver sity o f M u È n st er (Gesellschaft zur Fo È rderung der Westfa È lischen W ilhelms-Universita È t), G er m a ny.I t ha n k A nja Sch n eid er (E xp erim en t 1) a nd M a r tina L u za r (E xp erim en t 2) fo r t heir help in co n du ctin g th e experim en t s. I a m gr atefu l t o T im D algleish , Ja n D e H o uwer, C o lin M . M a cL eo d, a n d Tra m m ell N eill fo r th eir com m en ts a n d su ggestio n s o n a n ea rlier ver sion o f th is m a nu scr ipt . I th a n k D avid Bu r m edi fo r his help in im p ro vin g th e E n glish o f th is m a nu script . q 19...