Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
How can failure sometimes be better than success? ss? Varying effects of emotion on lexical processing David 1 Vinson , Paraskevi 1 Argyriou , Sara Rodriguez 2 Cuadrado , Gabriella 1 Vigliocco 1Cognitive, Perceptual and Brain Sciences Research Department, rtment, University College London 2Department of Technology, Universitat Pompeu Fabra Experiment 1: Lexical decision sion, single words (n=34) 640 630 CAKE PRISON Compare: Positive words vs. negative words (Exp 1-4) Interaction between face and word emotion (Exp 3,4) Supported by the Economic and Social Research Council (Grant RES-062-23-2012) Contact: d.vinson@ucl.ac.uk 0.9 0.85 830 0.8 610 Correct RT (msec) 620 0.97 0.965 0.96 820 810 800 0.75 0.7 0.65 790 600 0.955 0.6 780 590 0.95 580 0.945 Negative words Neutral words Positive words Negative words Neutral words 770 0.55 760 0.5 Positive words Emotional faster and more accura urate than neutral Positive more accurate than nega egative (RT: no difference) Negative words Neutral words Positive words Negative words Neutral words Positive words Emotional less accurate than neutral (but bias can explain) Negative far more accurate than positive. (RT: no differences at all) ● Effects of emotion dissociate ate within subjects and within items. Differences for neutral stimuli can be discounted due to bias in valence decision, but negative tive advantage occurs only in valence decision, not in lexical decision. ● Replicates both Kousta et al (2009) and Nasrallah et al (2009), ensuring that differences are not due to item variation. Experiment 3: Lexical decision sion, words over faces (n=43) Experiment 4: Valence decision, words over faces (same n=43) Angry face Angry face Happy face Neutral face Angry face Happy face 1 840 820 0.98 Angry face Happy face Neutral face 0.96 Happy face Neutral face 650 0.91 800 640 0.86 620 610 600 780 0.96 0.94 Proportion Correct 630 Proportion Correct Correct RT (msec) Experiments 1 and 2: Single word displays, fully visible. 1: Lexical decision (is a string of letters a word or not?) 2: Valence decision (is a word emotionally loaded or not?) Experiments 3 and 4: Compound displays: word + emotional face 3. Lexical decision (ignore the face) 4. Valence decision (ignore the face) 850 0.975 660 Examine emotion effects in different lexical tasks, using the same highly controlled set of items, manipulating task within-subjects to minimise the possibility that individual differences may contribute to variation in emotion effects. 0.95 840 Aims Words: positive, negative, neutral (based on Kousta et al., 2009). Faces: (Exp 3,4), happy, neutral, angry 0.98 860 Correct RT (msec) Inconsistencies could arise for various reasons: ● Different sets of items used, some not sufficiently controlled ● Emotion effects may vary for visible vs. threshold/masked words ● Different task demands may modulate or mask effects of emotion ● Individual differences may modulate emotion effects 0.985 Proportion Correct ● Negative word advantage vs. positive Nasrallah et al. (2009); valence decision, masked words ● Negative word disadvantage vs. positive and neutral Estes & Adelman (2009); lexical decision, visible words Vinson et al. (2011); location decision, visually suppressed words ● Emotional word advantage vs. neutral Kousta et al. (2009); lexical decision, visible words. 650 Correct RT (msec) Emotional characteristics of words affect language processing Not just words referring to emotional experience (love, hate) but also those with emotional connotations (cake, prison) But there are many apparently incompatible effects across studies which leads to difficulties in developing comprehensive accounts. Experiment 2: Valence decision, single words (same 34 subjects as Exp1) Proportion Correct Background 760 740 720 0.81 0.76 0.71 0.92 700 0.66 590 680 0.9 580 0.61 660 570 0.88 Negative words Neutral words Positive words 640 Negative words Neutral words Emotional words faster and more re accurate than neutral No effect of face and no word-fac face emotion interaction Positive words 0.56 Negative words Neutral words Positive words Negative words Neutral words Positive words Negative words faster and more accurate than positive Interaction: Happy faces improve accuracy for positive words Discussion ● Task differences are real and reflect differing task demands and variation in involvement of emotion. ● Attentional focus accounts exp explain negative advantage only in early processing or in tasks directly related to emotion. Further supported by the effects of faces in Experiment 4. ● On the other hand, non-emot emotional tasks using fully visible stimuli benefit from emotional content regardless of polarity. Lack of face effects suggest st these effects are due to general motivational effects rather than specific emotion processes.