subliminal em priming

I. Skandrani-Marzouki & Y. Marzouki: Emotion and Decision MakingSwissJ. Psychol. 69 (4) 2010 Verlag Hans Huber, Hogrefe AG, Bern

Original Communication

Subliminal Emotional Priming andDecision Making in a Simulated

Hiring SituationIns Skandrani-Marzouki1 and Yousri Marzouki2

1Universit dAix-Marseille I, Laboratoire de Psychologie Sociale, France2Universit dAix-Marseille I & Laboratoire de Psychologie Cognitive, CNRS, France

Swiss Journal of Psychology, 69 (4), 2010, 213219

DOI 10.1024/1421-0185/a000025

Abstract. The present study examines the unconscious influence of emotional information on decision making in a simulated hiringsituation. We used a subliminal masked priming paradigm with varying faces as primes, which were presented for a duration of 50 msand had two levels of emotion: positive emotion (happiness) and negative emotion (anger). These primes were followed by emotionallyneutral target faces. Primes were congruent (same faces) or incongruent (different faces). Prime Emotion (positive vs. negative) wascrossed with Prime Repetition (repeat vs. unrelated) in a 2 2 factorial design. Each participant was tested in all four of the experimentalconditions, each of which had 5 different trials. The participants were asked to indicate as rapidly as possible whether they were favor-able or unfavorable toward the selection of the candidate (target face). Two dependent measures were analyzed: number of target faceschosen (i.e., number of favorable responses to target faces) and reaction time (RT). Results revealed a strong effect of emotional priming.Participants tended to choose more target faces preceded by positive prime faces than by negative prime faces. Moreover, they reactedfaster when presented with target faces preceded by negative primes. Despite its exploratory nature, this study provides further evidencefor the role of emotional processing in modulating decision processes and extends the experimental manipulation of subliminal emotionto the case of the masked repetition priming technique.

Keywords: subliminal priming, emotion, decision making

Introduction

Since Murphy and Zajoncs (1993) seminal study it hasbeen known that affective primes have a very fast effect onour behavior. Many subsequent studies have shown suchinfluences in various ways, including having participantsjudge the affect of an ambiguous expression of target faces(Li, Zinbarg, Boehm, & Paller, 2008), judge the validity ofadvertising messages (Channouf, 2000), and inducingemotional priming with word stimuli of varying emotionalvalence (e.g., Bargh & Pietromonaco, 1982; Fazio, 2001;Gibbons, 2009). In the present study, we further demon-strate that subliminal emotional information can be a use-ful, indirect source of information for optimizing the deci-sion process.

A perpetually thorny problem in the social and psycho-logical sciences is the fuzziness in defining what emotionactually is. It is generally assumed that emotion is basicallya component process (Scherer, 2005). This recent approachprovides a direct link between each emotion componentand its associated function as follows: the cognitive com-ponent is responsible for the evaluation of objects andevents; the neurophysiological component determines the

system regulation; the motivational component affordspreparation for action and execution of actions; the motorexpression component defines behavioral intentions andcommunication; and the subjective feeling component rep-resents the monitoring of the interactions between internaland external states of the organism.

Moreover, three dimensions frequently emerge to ex-plain this concept (see Smith & Lazarus, 1990): Emotion as a state: This dimension is very close to Da-

masios (1994) definition of feelings. It entails a signif-icant change in a bodys state (i.e., organism) caused bythe real presence of external agents responsible for thephysiological reactions of the body and also for internalvariations where the causes can be simply representa-tions and mental states.

Emotion as a process: This dimension is rooted in thenotion of action associated with the source cause ofthe emotional reaction, where the organism builds upstrategies by taking into consideration contextual con-straints (e.g., Russell, 2003).

Emotion as a source of knowledge: It is obvious thatwhen we picture or imagine fictional events, we experi-ence deep emotional reactions that allow the body to

Swiss J. Psychol. 69 (4) 2010 Verlag Hans Huber, Hogrefe AG, Bern

acquire expectations about the way these reactions varywhen facing unknown events. Ultimately, the body es-tablishes an elaborate knowledge structure about theemotional quality of our behavior (see also Russell,2003).

Cognitive approaches consider emotion, for the most part,to be an indirect source of knowledge. The use of affect1 asinformation involves both identification of the affective re-actions and the determination of the relevance of these re-actions for a given judgment. The affect-as-informationmodel (see Clore, Gasper, & Garvin, 2001; Schwarz & Clo-re, 1983, 1988, 1996) is a good candidate for providing aframework for a cognitive approach to understanding theunconscious influence of emotion on behavior. Indeed, thismodel is based upon the assumption that emotional pro-cesses occur without conscious awareness. Hence, affec-tive feelings allow us to explicitly learn about our own im-plicit judgments and decisions.

Three levels of behavioral analysis of emotional infor-mation can be discussed when we refer to the cognitivemodels of emotion:1) level of cognitive structures, which refers to different

forms of organization of emotional information;2) level of cognitive strategies, which is mostly involved in

problem-solving situations (e.g., Forgas, 1991; Schwarz& Clore, 1983, 1988);

3) level of cognitive propositions, which refers to seman-tic contents (e.g., Bower, 1981).

In the absence of conscious awareness, the content of emo-tional stimuli can influence human decision making (Bargh& Pietromonaco, 1982; Greenwald, Klinger, & Liu, 1989).One way to test this assumption is to use the masked repe-tition priming paradigm, which has become a dominanttool in the investigation of early cognitive and emotionalprocesses (Bargh, 1998; Dixon, 1971, 1981; Forster & Da-vis, 1984; Marcel, 1983). In this paradigm, the prime stim-ulus is preceded by a pattern mask and followed by a targetstimulus that requires a response. Because the prime is pre-sented for a very short duration, participants are typicallynot aware of the prime stimulus. In order to assess the sizeof priming effects, a comparison is required of the congru-ent/related condition (with physical similarity between theprime and the target) with the incongruent/unrelated con-dition (no relationship between prime and target). Thestandard result corresponds to an advantage in favor of therelated condition with faster reaction times (RTs) and ahigher rate of accuracy as compared to the unrelated con-dition.

Many studies have used this experimental procedurewith various manipulations and typically show that partic-ipants are faster to recognize a target word when it is

primed by a semantically related as opposed to a semanti-cally unrelated word (e.g., Marcel, 1983; Pillon, 1993). Asimilar result was observed when a subliminal image primeand a target word shared the same semantic category (e.g.,Channouf, Canac, & Gosset, 1999). The first demonstra-tion of an implicit affective effect was shown by Clore andStorbeck (2006) who manipulated participants mood, in-ducing positive or negative mood by presenting partici-pants with sad or happy music prior to the categorization,judgment, or lexical decision task. Their results revealedthat positive mood significantly facilitated priming effectsrelative to negative mood. The same pattern of results wasobserved over the three priming tasks and also for differentprime presentations.

Most relevant to our study is subsequent research thatmanipulated facial expressions by using the priming tech-nique. Channouf (2000) conducted an experiment in whichparticipants evaluated advertising messages by pressingbuttons corresponding to true or false answers. Thesemessages were primed by faces displaying either positiveor negative emotion. In the condition in which the presen-tation of facial expression was subliminal, participants re-sponded with more true answers to the advertising targetsentences preceded by positive as compared to negativeprime faces. These results were observed only with thenumber of true/false answers as a dependent variable,whereas no significant effect was reported for reaction time(RT). More recently, Winkielman, Berridge, and Wilbarger(2005) found that participants exposed to subliminal pic-tures of happy faces drank more and were willing to paymore for their beverage than participants exposed to angryfaces.

Pittman and Bornstein (1998) conducted a subliminalpriming experiment aimed at understanding the effect ofgender and social discrimination in a simulated hiring sit-uation. Prior to the experiment, the authors subliminallypresented the participants with a series of pictures of can-didates who might aspire to the position. Their resultsshowed that, in 70% of the cases, the participants chose thecandidates seen beforehand.

Since a very brief presentation of stimuli produced thismere exposure effect (see Zajonc, 1968), one may ask howemotions can be a useful indirect source of information.Because the Pittman and Bornstein (1998) study did notuse an emotion manipulation, in the present experiment wecombined the standard masked priming procedure with asubliminal manipulation of emotion. We also opted for atwo-alternative forced choice task (2-AFC) to capture thedecision processes in the RT measure. In addition, the pre-sent study aimed to provide further support for the sublim-inal influence of emotion on decision-making processes.Unlike previous studies, we used the hiring situation as ascenario and a context in order to enhance the relevance ofthe decision made by the participant and to provide a more

214 I. Skandrani-Marzouki & Y. Marzouki: Emotion and Decision Making


1 The use of affect as a synonym of emotion is very common. We consider affect, as did Schwarz and Clore (1983, 1988, 1996), to bean instant expression of emotion.

straightforward evaluation of emotional influence. Finally,by excluding participants from the final analysis who wereaware of prime presentation, we obtained a more conser-vative measure of the behavioral effects due to a subliminalemotional influence at the prime level.

Method

Participants

A total of 65 students (11 males, mean age = 22, SD = 2.4years) from the Universit de Provence volunteered to takepart in the experiment. In the end, only 60 participants wereused in the analyses because of their verbal self-reports ina brief postexperimental interview (e.g., Chalmers, 1996;Critchfield, 1993; Geissler, 1990; Merikle, 1992). These 60participants reported that they were unable to detect oridentify any stimulus prior to the target presentation duringthe entire experiment.

Stimuli and Design

In order to obtain the different faces used as stimuli, weasked another group of 35 student volunteers (M = 24years, SD = 2.3) to express the required facial expressions(positive, neutral, and negative) for the experiment. Thesestudents consented to be photographed. A total of 128 pic-tures of faces were collected, but only 40 were used in theexperiment. The pictures selected illustrated the emotionalexpression most saliently. All faces selected were plain-looking, without any particular clue (e.g., eyeglasses, hat,excessive make-up, etc.) making their perception confus-ing, difficult, or cued. The facial expressions chosen asprime faces (positive emotion vs. negative emotion) andtarget faces (neutral) were selected based on Kendallsmethod of evaluating the degree of consensus between fiveraters. For the ratings, all pictures were presented in se-quence on a computer screen and each rater used a Likert-type scale from 2 obviously negative expression to +2:obviously positive expression to decide which levelmatched the given facial expression best. The resultsshowed highly significant interjudge agreement, W = 0.89,p = .0001.

In the present study, all target faces were neutral, where-as the prime faces were either positive or negative. A trialin which the target neutral face was preceded by the sameprime face, whether expressing positive or negative emo-tion, was considered a related (repeat) trial. The opposite an unrelated trial consisted of a neutral target face primedby a different face, regardless of facial emotion.

Two factors were manipulated in this experiment: PrimeEmotion (positive vs. negative) and Prime Repetition (re-peat vs. unrelated) crossed in a 2 2 factorial design. Eachparticipant was tested in all four experimental conditions,

each of which contained five different trials. The entire setof 20 trials was presented randomly during the experiment.Two dependent measures were recorded: number of targetfaces chosen by participants to continue on in the hiringprocess and reaction time (RT).

Procedure

Situation Scenario

The situation scenario consisted in showing the participanta document summarizing the main steps in a professionalhiring protocol. A report was enclosed in the documentstating that the remaining candidates involved in this im-portant selection have a remarkable education and careerpathway (see Appendix). It also stated that because of thelimited number of positions required for the job, the par-ticipant was to choose the candidates most likely (accord-ing to the participant) to proceed to the final hiring step.The experimenter walked the participant through each stepof the document in order to make sure that he or she under-stood what the task required before moving on to the ex-periment.

The Experiment

The experiment was run in a sound attenuated and dimlylit room. The participants were seated in front of a comput-er screen on which stimuli were displayed on a monitorwith a resolution of 1024 768 pixels and a 60-Hz refreshrate. The experiment was conducted using SuperLab soft-ware. Before the experiment started, participants carefullyread the instructions and performed a practice session of 5trials that did not appear in the main experiment.

Each trial began with a central fixation point lasting for300 ms. The fixation point was then replaced by a forwardmask (premask) consisting of a black and white spottedsquare, which appeared for 20 ms. The prime face imme-diately followed the forward mask and lasted for 50 ms,which, in turn, was replaced by a backward mask (post-mask) similar to the premask for 20 ms. The target facefollowed the backward mask and remained on the screenuntil the participant had made his or her response. The pro-cedure is described in Figure 1.

The participants were asked to indicate as rapidly as pos-sible whether they were favorable or unfavorable to-ward the selection of each candidate by pressing a greenresponse button with their right index finger for a favor-able answer and a red button with their left index fingerfor an unfavorable answer. The response hand was coun-terbalanced across participants. At the end of the experi-ment, a verbal response to the question, Just before thedisplay of faces on the screen, were you able to detect thevery brief appearance of other faces? was recorded forevery participant.

I. Skandrani-Marzouki & Y. Marzouki: Emotion and Decision Making 215


Results

Of the 65 participants, 5 were excluded (7.7%) from the mainanalysis because they had reported being able to perceiveblurry images, flashes, or brief patterns prior to thetarget presentation. Table 1 presents the average RTs andnumber of chosen target faces for all 60 participants in thefour experimental conditions. Overall, for both number oftarget faces chosen and RTs, repeated measures ANOVAsand paired t-tests techniques showed that the repetition effectwas not significant. Regarding the number of target faceschosen, the net repetition priming effect the difference be-tween Meanunrelated and Meanrepeat gives only 0.06 units(1 unit is the equivalent of an average of one target face cho-sen per participant) in the presence of positive emotion andonly 0.07 units in the presence of negative emotion. For RTs,this difference was not null, but also not significant, with anet repetition priming effect of 56 ms for positive emotionrelative to 53 ms for negative emotion. The trends of emo-tional priming for RTs and number of target faces chosen arein opposite directions; in fact, emotional priming is in favorof negative emotion when measuring RTs (faster RTs) and infavor of positive emotion when measuring the number oftarget faces chosen (higher rate).

ANOVA Results

Number of Choices

A repeated measures ANOVA was performed on the numberof target faces chosen data with Prime Emotion and PrimeRepetition as factors. Results revealed a significant main ef-fect of Prime Emotion, F(1, 59) = 7.06, p = .01 (see Figure2), but no significant main effect of Prime Repetition, F < 1,and no significant interaction between these two factors, F < 1.

Reaction Times

We also conducted a repeated measures ANOVA on the RTdata with Prime Emotion and Prime Repetition as factors.There was a significant main effect of Prime Emotion,F(1, 59) = 5.10, p = .02 (see Figure 3), but no significantmain effect of Prime Repetition, F(1, 59) = 2.77, p > 1, andno significant interaction, F < 1.

In summary, the results regarding number of target faceschosen showed significantly more positive primes and RTswere faster for negative primes.

Table 1Mean reaction times (ms) and number of target faces chosen plus or minus standard errors for all experimental conditions

Positive emotion Negative emotion

Prime repeat Prime unrelated Prime repeat Prime unrelated

Mean number of target faces chosen 3.13 3.19 2.86 2.93

Standard error 0.40 0.41 0.37 0.38

Mean RT 1417.52 1473.73 1326.48 1379.61

Standard error 183.0 190.26 171.25 178.11

Figure 1. Structure of typical trials. (a)Example of a trial in the positive emo-tional priming condition with repeti-tion. (b) Example of a trial in the neg-ative emotional priming conditionwithout repetition.



Discussion

The present study investigated the effects of emotional rep-etition priming on decision making by manipulating therelatedness of the prime and target faces (same face vs.different face) and also the emotional facial expression ofprimes (positive vs. negative). The results revealed a strongeffect of emotional priming. With respect to the number ofcandidates chosen for final selection, participants tended tochoose more target faces preceded by positive primes. Thisresult is consistent with previous studies that have showna facilitating role of positive emotion on human perfor-mance. This facilitation is reflected in favorable feedbackthat has been observed in various tasks such as evaluation,judgment, decision, etc. (e.g., Channouf, 2000; Channoufet al., 1999; Clore & Storbeck, 2006; Isen & Patrick, 1983;Petty, Schumann, Richman, & Strathman, 1993).

Unlike Channoufs research (2000), the present studyshowed a significant effect of priming with RT data. RT re-sponses were significantly faster when participants respond-ed to target faces preceded by primes with negative emotionrelative to positive emotion. This result is consistent with theevolutionist interpretation provided by Hansen and Hansen(1994). According to their interpretation, negative emotion

has a greater impact than positive emotion on our behaviorboth as individuals and also as a species. This evolutionistapproach highlights the adaptive aspect of a person trying toavoid potential sources of danger, for instance, negative emo-tional information within the environment. Thus, on the onehand, our participants are faster at avoiding negative emo-tional information by pressing the response button quickly.On the other hand, they have a preference for positive emo-tional information and therefore respond less quickly butchoose more target faces preceded by positive primes. Thespeed of such processing is consistent with the idea that neg-ative affect may foster a processing style that is characterizedby bottom-up processing, whereas positive affect may fostera top-down processing style that relies more on generalknowledge structures (Bless & Schwarz, 1999; Clore, Gas-per, & Garvin, 2001).

Our results do not reveal a reliable masked repetition effectfor faces, either for the number of target faces chosen or forRT. In other words, whether the target face was preceded bythe same candidates face or a different one did not changethe pattern of the priming effect. In the so-called sandwichpriming experiment, which is similar to the present experi-ment, Henson, Mouchlianitis, Matthews, and Kouider (2008)manipulated face familiarity (famous vs. nonfamous) andprime repetition condition (same view vs. different view vs.unprimed). Their results revealed a significant face primingrepetition effect with both behavioral and ERP data for famil-iar faces, but no priming effect in behavioral data for unfa-miliar or nonfamous faces. This latter result is consistent withour study as all of our stimuli are unfamiliar faces and theabsence of repetition priming in the present study could beexplained by the familiarity effect. In the Henson et al. study(2008), the repetition variation between the prime face andthe target face preserves the emotional expression even withviews from different perspective. Therefore, the absence ofrepetition effects in this study may be attributable to the ab-sence of emotion repetition regardless of face repetition (i.e.,repetition of the same persons face). These results are con-sistent with the idea of an emotional congruency effect be-tween prime and target (Hansen & Shantz, 1995) where eachis processed as two emotively incongruent behaviors.

It is conceivable that the participants affective statecould also play an important role in modulating the amountof priming effects observed under such experimental con-ditions. The effect of this variable has been demonstratedby many authors in a variety of tasks including social judg-ment and the judgment of facial expressions (e.g., Baron,1987; Bower, 1981; Forgas, 1991; Forgas & Moylan,1987). Bowers network model of emotions (Bower, 1981;Bower & Cohen, 1982) acknowledges the role of mood-dependent processes in modulating decision making. Thekey idea of this model is that moods can activate memories,concepts, and categories that are congruent with the va-lence (positive vs. negative) of a given mood. This influ-ence can operate at a very low level (attentional and per-ceptual level) as well as at the level of memory retrievaland judgment. It is possible that the subliminal facial ex-

Figure 2. Comparison of the number of target faces chosenby participants when primed by positive versus negativeemotion. Significant priming effect is indicated by asterisks(**p < .01).

Figure 3. Comparison of the decision RT means (ms) in thepositive and negative emotion priming condition. Signifi-cant priming effect is indicated by an asterisk (*p < .05).



pressions of our prime faces triggered a mood inductionthat consequently affected the judgment or the decisionalprocesses. Our results support Scherers proposal (2005)that emotion has a cognitive component responsible for theevaluation of objects and events.

From an applied perspective, the present findings sug-gest that it is advisable that recruiters heed emotional in-formation to optimize the likelihood of making an unbiaseddecision. Such a strategy could be crucial in a real environ-ment of an enterprise with high stakes. Alternatively, jobseekers advisors have tried to help candidates optimizetheir chances of being hired by taking into account factorssuch as their social background, gender, physical appear-ance, stance, and attire. However, more and more, currentresearch takes into account the influence of emotion andshows that mood, facial expression, and emotional infor-mation are of the same importance as all other factors.

Despite its exploratory dimension, this study providedfurther evidence for the role of emotion in modulating cog-nitive processes. Moreover, it extends the experimental ma-nipulation of subliminal emotion to the masked repetitionpriming technique. In light of the results from this simulat-ed hiring experiment, one possible interesting external val-idation of the study would be an action research projectcarried out within an organization and conducted with pro-fessionals in a real hiring situation.

Acknowledgments

The authors would like to thank Kaitlyn Litcofsky, SarahCleary, Fred Mast, and two anonymous reviewers for theirvery helpful comments.

References

Bargh, J. A. (1998). Quel est lintrt du subliminal pour la psy-chologie sociale? Pour une distinction entre conscience dustimulus et conscience de son influence [What is the purposeof the subliminal for social psychology? Toward a distinctionbetween the stimulus awareness and the awareness of its influ-ence]. In A. Channouf & M.-F. Pichevin (Eds.), Le pouvoirsubliminal: Influences non conscientes sur le comportement.Lausanne, Switzerland: Delachaux & Niestl.

Bargh, J. A., & Pietromonaco, P. (1982). Automatic informationprocessing and social perception: The influence of trait infor-mation presented outside of conscious awareness on impres-sion formation. Journal of Personality and Social Psychology,43, 437449.

Baron, R. A. (1987). Interviewers mood and reaction to job ap-plicants. Journal of Applied Social Psychology, 17, 911926.

Bless, H., & Schwarz, N. (1999). Sufficient and necessary condi-tions in dual process models: The case of mood and processingstyle. In S. Chaiken & Y. Trope (Eds.), Dual process modelsin social psychology (pp. 423440). New York: Guilford.

Bower, G. H. (1981). Mood and memory. American Psychologist,36, 129148.

Bower, G. H., & Cohen, P. R. (1982). Emotional influence on mem-ory and thinking: Data and theory. In S. Fiske & M. Clark (Eds.),Affect and cognition (pp. 291331). Hillsdale, NJ: Erlbaum.

Chalmers, D. J. (1996). The conscious mind. New York: OxfordUniversity Press.

Channouf, A. (2000). Subliminal exposure to facial expressionsof emotion and evaluative judgments of advertising messages.European Review of Applied Psychology, 50, 1923.

Channouf, A., Canac, D., & Gosset, O. (1999). Nonspecific ef-fects of subliminal advertising. European Review of AppliedPsychology, 49, 1319.

Clore, G. L., & Storbeck, J. (2006). Affect as information aboutliking, efficacy, and importance. In J. Forgas (Ed.), Hearts andminds: Affective influences on social cognition and behavior(pp. 123142). New York: Psychology Press.

Clore, G. L., Gasper, K., & Garvin, E. (2001). Affect as informa-tion. In J. P. Forgas (Ed.), Handbook of affect and social cog-nition (pp. 121144). Mahwah, NJ: Erlbaum.

Critchfield, T. S. (1993). Signal detection theory properties of ver-bal self-reports. Journal of the Experimental Analysis of Be-havior, 60, 495514.

Damasio, A. R. (1994). Descartes error: Emotion, reason, andthe human brain. New York: Penguin Putnam.

Dixon, N. F. (1971). Subliminal perception: The nature of a con-troversy. London, UK: McGraw-Hill.

Dixon, N. F. (1981). Preconscious processing. New York: Wiley.Fazio, R. H. (2001). On the automatic activation of associated

evaluations. Cognition and Emotion, 54, 127.Forgas, J-.P. (1991). Emotion and social judgment. Oxford, UK:

Pergamon.Forgas, J-.P., & Moylan, S. (1987). After the movies: The effect

of mood on social judgments. Personality and Social Psychol-ogy Bulletin, 13, 465477.

Forster, K. I., & Davis, C. (1984). Repetition priming and frequen-cy attenuation in lexical access. Journal of Experimental Psy-chology: Learning, Memory and Cognition, 10, 680698.

Geissler, H. G. (1990). Psychophysical maps for rectangles. InH.-G. Geissler (Ed.), Psychophysical explorations of mentalstructures (pp. 149164). Toronto, Canada: Hogrefe & Huber.

Gibbons, H. (2009). Evaluative priming from subliminal emo-tional words: Insights from event-related potentials and indi-vidual differences related to anxiety. Consciousness and Cog-nition, 18, 383400.

Greenwald, A. G., Klinger, M. R., & Liu, T. J. (1989). Uncon-scious processing of dichoptically masked words. Memory andCognition, 17, 3547.

Hansen, H. C., & Hansen, R. D. (1994). Automatic emotion: At-tention and facial efference. In P. M. Niedenthal & S. Kita-yama (Eds.), The hearts eye: Emotional influences in percep-tion and attention (pp. 217243). San Diego, CA: AcademicPress..

Hansen, H. C., & Shantz, C. A. (1995). Emotion-specific priming:Congruence effects on affect and recognition across negativeemotions. Personality and Social Psychology Bulletin, 21,548557.

Henson, R. N., Mouchlianitis, E., Matthews, W. J., & Kouider, S.(2008). Electrophysiological correlates of masked face prim-ing. Neuroimage, 40, 884895.

Isen, A. M., & Patrick, R. (1983). The effect of positive feelingson risk-taking: When the chips are down. Organizational Be-havior and Human Performance, 31, 194202.



Li, W., Zinbarg, R., Boehm, S., & Paller, K. (2008). Neural andbehavioral evidence for affective priming from unconsciouslyperceived emotional facial expressions and the influence oftrait anxiety. Journal of Cognitive Neuroscience, 20, 95107.

Marcel, A. J. (1983). Conscious and unconscious perception: Ex-periments on visual masking and word recognition. CognitivePsychology, 15, 197237.

Merikle, P. M. (1992). Perception without awareness: Critical is-sues. American Psychologist, 47, 792795.

Murphy, S. T., & Zajonc, R. B. (1993). Affect, cognition, andawareness: Affective priming with optimal and suboptimalstimulus exposures. Journal of Personality and Social Psy-chology, 64, 723739.

Petty, R. E., Schumann, D. W., Richman, S. A., & Strathman A. J.(1993). Positive mood and persuasion: Different roles for af-fect under high- and low-elaboration conditions. Journal ofPersonality and Social Psychology, 64, 520.

Pillon, A. (1993). La mmoire des mots [The memory of words].Lige, Belgium: Mardaga.

Pittman, T. S., & Bornstein, R. F. (1998). Nonconscious biasingeffects in hiring decisions. Paper presented at the 60th annualmeeting of the Eastern Psychological association, Boston, MA.

Russell, A. J. (2003). Core affect and the psychological construc-tion of emotion. Psychological Review, 110, 145172.

Scherer, K. (2005). What are emotions and how can they be mea-sured? Social Science Information, 44, 695729.

Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, andjudgment of well-being: Informative and directive functions

of addictive states. Journal of Personality and Social Psychol-ogy, 45, 513523.

Schwarz, N., & Clore, G. L. (1988). How do I feel about it? Theinformative function of affective states. In K. Fiedler & J. P.Forgas (Eds.), Affect, cognition and social behavior(pp. 4462). Toronto, Ontario, Canada: Hogrefe & Huber.

Schwarz, N., & Clore, G. L. (1996). Feelings and phenomenal expe-riences. In E. T. Higgins & A. Kruglanski (Eds.), Social psychol-ogy of basic principles (pp. 433465). New York: Guilford.

Smith, C. A., & Lazarus, R. S. (1990). Emotion and adaptation.In L. A. Pervin (Ed.), Handbook of personality theory and re-search (pp. 609637). New York: Guilford.

Winkielman, P., Berridge, K. C., & Wilbarger, J. L. (2005). Un-conscious affective reactions to masked happy versus angryfaces influence consumption behavior and judgments of value.Personality and Social Psychology Bulletin, 1, 121135.

Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Jour-nal of Personality and Social Psychology Monograph, 9, 127.

Yousri Marzouki

UFR de PsychologieUniversit de Provence29 Av. Robert SchumanFR - 13621 [email protected]

Appendix

Information About the Hiring Task

In order to help you select the appropriate candidates forthe position, we have prepared an overview of standardsteps in the hiring process. These steps start with:

1. Job description: Where the employer defines the criteriaabout what the position involves exactly.

2. Advertising the position: The job description is postedon the internet and in newspapers.

3. Screening applicants: This step is based on a review ofall applicants resumes, skills, and whether applicantsmeet the education requirements for the job.

4. Interviewing: In this step, the interviewer needs to havea clear idea of correct responses to a list of questionsinvolving the applicants skills, education, and experi-ence, ask and anticipate general questions based on thecandidates responses.

5. Comparing applicants: Applicants are rated accordingto how well they performed during the interview.

6. Decision: This step entails two possible outcomes: Joboffer or rejection of application.

The remaining candidates involved in this important selec-tion step showed remarkable education and career back-grounds. Because of the limited number of positions re-quired for the job, during this session, you are invited toselect the candidates who are most likely to proceed to thefinal hiring step. Below is a summary of the average re-sume that all remaining candidates share.

Job or career expectations: Reach high-level manage-rial position.

Education: Minimum is masters degree in managementwith high honors.

References: With respect to schools as well as compa-nies, candidates come from leading institutions.

Experience: Despite their young age (24 years average),all candidates have an impressive amount of experiencein various organizations and with different responsibili-ties.

Special skills: These candidates have excellent comput-er skills and programming expertise, including knowl-edge of specific software and operating systems. More-over, their talents include good social skills and the abil-ity to work efficiently in a team.



subliminal em priming

Documents