How Mood States Affect Information Processing During Facial Emotion Recognition: An Eye Tracking Study

(1)

Article

Reference

How Mood States Affect Information Processing During Facial Emotion Recognition: An Eye Tracking Study

SCHMID, Petra C., et al.

Abstract

Existing research shows that a sad mood hinders emotion recognition. More generally, it has been shown that mood affects information processing. A happy mood facilitates global processing and a sad mood boosts local processing. Global processing has been described as the Gestalt-like integration of details; local processing is understood as the detailed processing of the parts. The present study investigated how mood affects the use of information processing styles in an emotion recognition task. Thirty-three participants were primed with happy or sad moods in a within-subjects design. They performed an emotion recognition task during which eye movements were registered. Eye movements served to provide information about participants' global or local information processing style. Our results suggest that when participants were in a happy mood, they processed information more globally compared to when they were in a sad mood. However, global processing was only positively and local processing only negatively related to emotion recognition when participants were in a sad mood. When they were in a happy mood, processing style [...]

SCHMID, Petra C., et al . How Mood States Affect Information Processing During Facial

Emotion Recognition: An Eye Tracking Study. Swiss Journal of Psychology , 2011, vol. 70, no.

4, p. 223-231

DOI : 10.1024/1421-0185/a000060

Available at:

http://archive-ouverte.unige.ch/unige:101519

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/230800622

How mood states affect information processing during facial emotion recognition: An eye

tracking study

Article in Swiss Journal of Psychology · December 2011

DOI: 10.1024/1421-0185/a000060

CITATIONS

25

READS

2,268

5 authors, including:

Some of the authors of this publication are also working on these related projects:

Analog vs. Digital - The Emotional Impact of Film Recording Processes on the AudienceView project

Power and goal pursuitView project Marianne Schmid Mast

University of Lausanne

134PUBLICATIONS 2,388CITATIONS

SEE PROFILE

Dario Bombari

University of Lausanne 12PUBLICATIONS 163CITATIONS

SEE PROFILE

Fred W Mast Universität Bern

163PUBLICATIONS 2,422CITATIONS

SEE PROFILE

Janek S Lobmaier Universität Bern

56PUBLICATIONS 793CITATIONS

SEE PROFILE

All content following this page was uploaded by Marianne Schmid Mast on 02 June 2014.

The user has requested enhancement of the downloaded file.

(3)

Original Communication

How Mood States Affect Information Processing During Facial Emotion Recognition: An Eye Tracking Study

Petra C. Schmid

¹

, Marianne Schmid Mast

¹

, Dario Bombari

²

, Fred W. Mast

²

, and Janek S. Lobmaier

²

1Department of Work and Organizational Psychology, University of Neuchatel, Switzerland

2Department of Psychology, University of Bern, Switzerland Swiss Journal of Psychology, 70 (4), 2011, 223–231

DOI 10.1024/1421-0185/a000060

Abstract.Existing research shows that a sad mood hinders emotion recognition. More generally, it has been shown that mood affects information processing. A happy mood facilitates global processing and a sad mood boosts local processing. Global processing has been described as the Gestalt-like integration of details; local processing is understood as the detailed processing of the parts. The present study investigated how mood affects the use of information processing styles in an emotion recognition task. Thirty-three participants were primed with happy or sad moods in a within-subjects design. They performed an emotion recognition task during which eye movements were registered. Eye movements served to provide information about participants’ global or local information processing style. Our results suggest that when participants were in a happy mood, they processed information more globally compared to when they were in a sad mood. However, global processing was only positively and local processing only negatively related to emotion recognition when participants were in a sad mood. When they were in a happy mood, processing style was not related to emotion recognition performance. Our findings clarify the mechanism that underlies accurate emotion recognition, which is important when one is aiming to improve this ability (i.e., via training).

Keywords:emotion recognition, information processing, mood, gender

Introduction

Facial emotional expressions are important cues in social communication. Facial expressions coordinate social interactions by providing information about the sender’s emotional and intentional states and/or the nature of the relationship between the two interaction partners. Although the ability to correctly recognize other people’s emotions is crucial to experience positive social interactions (Keltner

& Kring, 1998), the mechanism underlying accurate emotion recognition is still unclear. We know that, in general, people reliably recognize emotions from faces, yet this ability seems to vary between individuals (Herba & Phil- lips, 2004; Moore, 2001).

A large body of research reports that people suffering from affective disorders such as depression or mania show decreased emotion recognition accuracy (e.g., Bouhuys, Geerts, & Gordijn, 1999; Heimberg, Gur, Erwin, Shtasel,

& Gur, 1992; Lembke & Ketter, 2002; Surguladze et al., 2004; Zuroff & Colussy, 1986). Not only affective disorders, but also normal variations in mood states in healthy people have consequences for the perception of emotional

expressions. Sad, compared to happy, people tend to per- ceive more sadness and less happiness in faces (Bouhuys, Bloem, & Groothuis, 1995; Niedenthal, Halberstadt, Mar- golin, & Innes-Ker, 2000). Such mood-congruity effects have also been reported for emotion recognition accuracy:

Sad people showed a decrease in recognition accuracy for happy emotions, whereas happy people showed a decrease in recognition accuracy for sad emotions (Schmid &

Schmid Mast, 2010). Schmid and Schmid Mast’s stimulus material only included happy and sad facial expressions.

Chepenik and colleagues (Chepenik, Cornew, & Farah, 2007) used stimulus material consisting of neutral, happy, sad, angry, and fearful facial expressions. They found that a sad mood compared to a neutral mood had a detrimental effect on overall emotion recognition accuracy (no mood- congruity effects). Because there was no happy mood condition in the Chepenik et al. (2007) study, we do not know how a happy mood influences emotion recognition accuracy when more than two emotional expressions are included. Moreover, the mechanism underlying how mood affects emotion recognition accuracy remains poorly understood. Our goal was to obtain a more refined under-

(4)

standing of how correct emotion recognition works. In the present paper, we test the assumption that emotion recognition accuracy depends on the use of global versus local information processing styles and that people in a happy or a sad mood use different information processing styles.

Previous studies have reported that women usually outperform men in emotion recognition tasks (McClure, 2000, for a meta-analysis). Here, we test the assumption that this gender difference may be explained by different processing styles. These predictions will be outlined in the next sections.

Information Processing During Emotion Recognition

Previous research suggests that when people are recognizing emotions, they process faces globally rather than locally, meaning that the face is more likely to be processed as a whole and not in a piecemeal fashion. Calder, Young, Keane, and Dean (2000) cut faces expressing different emotions in two halves. The cut was done horizontally be- low the eyes, dividing the face into an upper and lower part.

The upper part of one face was then combined with a lower part of another face in such a manner that the upper and the lower face part contained different emotional expressions.

The upper and lower parts were either aligned so that the whole represented a new, intact face (composite), or the two face parts were presented shifted apart on the horizon- tal axis (noncomposite). Participants were instructed to either focus on the upper or the lower part of the composite or noncomposite faces and to name the emotion expressed in the respective face part (the other face part was to be ignored). Participants took longer to recognize the emotions when the faces were composite than when they were noncomposite. The authors concluded that participants were slower in the composite condition because they processed the relevant and irrelevant parts as a whole, and the irrelevant part, thus, interfered with the recognition of the emotion in the relevant half. This finding can be seen as evidence of a global information processing style when recognizing emotions.

Further support for the assumption that a global information processing style is important for emotion recognition comes from studies on the face inversion effect.

Prkachin (2003) showed that emotional expressions were more difficult to recognize when the faces were inverted than when the faces were upright. Given that several studies have shown that inverting a face hampers configural processing (e.g., Freire, Lee, & Symons, 2000; Searcy &

Bartlett, 1996; Yin, 1969), inverted facial expression of emotions might have been recognized less easily because configural – or global – processing was not possible. Sear- cy and Bartlett (1996) argued that configural information is comparable to global information, as configural information refers to spatiorelational information in faces, such

as the intereye distance or the distance between the nose and the mouth. Similar to global processing, perceiving configural information allows one to recognize large co- herent units or global forms (Brosnan, Scott, Fox, & Pye, 2004). The face inversion effect found in emotion recognition, therefore, supports the idea that a global processing style aids emotion recognition.

It is possible that one processing style may be more ap- propriate than the other for one emotion, but not for another. McKelvie (1995) found that the recognition of anger, disgust, surprise, sadness, and fear was hindered when the faces were inverted. Only recognition of happiness was not affected by face inversion. Prkachin (2003) showed that the recognition of anger, disgust, and fear was more affected by face inversion than the recognition of happiness, sadness, and surprise. Configural information might, therefore, play an important role in the recognition of most emotions. Happiness seems to be an exception, as it was still recognized accurately after configural or global information had been reduced. An explanation for this might be that happy faces are generally easier to recognize than other emotions and that very little information is needed for recognition (happy face advantage; e.g., Leppänen & Hieta- nen, 2004). Indeed, detecting a smiling mouth may suffice to recognize happiness.

Mood Effects on Information Processing

The affect-as-information theory (Clore et al., 2001;

Schwarz, 1990) suggests that people in a sad mood process information more deliberately and search for specific information before making a judgment. People in a happy mood use a more automatic or heuristic information processing style and judgments are made on the basis of an overall impression. This theory is supported by empirical findings (De Vries, Holland, & Witteman, 2008; Forgas, 1992). Gasper and Clore (2002) assumed that a differential use of information processing styles has consequences for global versus local processing of visual stimuli. The authors asked happy and sad participants to perform the global-local focus test introduced by Kimchi and Palmer (1982). This test contains hierarchical figures consisting of small triangles or squares (local attributes) forming a larger triangle or square (global attribute). A standard figure was presented simultaneously with two other figures. In one of these figures, the global attribute was the same as the standard figure and, in the other figure, the local attribute was the same as the standard figure. The task was to indicate which of the two figures was more similar to the standard figure. Results showed that sad people were more likely to rely on the local attributes (and, consequently, less likely to rely on the global attributes) than happy people. This finding is consistent with the idea that a happy mood triggers global information processing whereas a sad mood triggers local information processing.

224 P. C. Schmid et al.: How Moods Affect Information Processing

(5)

Gender Effects

Women typically outperform men in emotion recognition accuracy (McClure, 2000; for a meta-analysis). Moreover, there is evidence that men and women process information differently when recognizing emotions. Hall, Witelson, Szechtman, and Nahmias (2003) investigated the brain activation of men and women during emotion recognition.

They found less limbic activity and greater left hemispheric frontal cortical activity in men than in women. According to Damasio (1994) and Ledoux (2000), the limbic system is associated with reactions to primary emotions (fast and innate reaction to threatening stimuli) while prefrontal activation is associated with reactions to secondary emotions (learned and controlled reactions). Referring to these findings, Hall et al. argued that men use a less automatic and more analytic information processing style than women when recognizing emotions. Because automatic processing is triggered by global processing and analytic processing by local processing (Förster & Dannenberg, 2010), one might conclude that men generally tend to engage more in item-specific (local) processing and women more in relational (global) processing (e.g., Gilligan, 1982; Putrevu, 2001). Hence, women might outperform men in emotion recognition because they use a (global) processing style that is beneficial for correct emotion recognition whereas men use a less advantageous (local) processing style.

The Present Study

In the present paper, we examine whether the use of global versus local information processing styles influences the accuracy of emotion recognition and whether the relationship between processing style and accurate emotion recognition depends on the perceiver’s mood and gender. Eye tracking measures were used to assess participants’ information processing styles. We analyzed whether a global eye scan pattern (enabling an integrative, holistic impression of the face) or a local eye scan pattern (allowing a deeper analysis of the detailed facial information, such as the eyes, the mouth, and the nose) is more beneficial for accurate emotion recognition. The following predictions were tested: Consistent with the literature stressing the role of global processing in emotion recognition (Calder et al., 2000; Prkachin, 2003), we predicted that the more global the processing style, the better the emotion recognition performance (Hypothesis 1a). Conversely, the more local the processing style, the worse the performance on the emotion recognition task would be (Hypothesis 1b). Further, we expected participants in a happy mood to use a more global processing style (Hypothesis 2a) and a less local processing style (Hypothesis 2b) than participants in a sad mood (e.g., De Vries et al., 2008; Gasper & Clore, 2002; Schwarz, 1990). Because we hypothesized that participants in a happy mood use the more beneficial, global processing style

(e.g., Gasper & Clore, 2002), we predicted that participants in a happy mood would be more accurate than those in a sad mood (Hypothesis 3). Based on the meta-analysis by McClure (2000), we hypothesized that women would recognize emotions more accurately than men (Hypothesis 4).

If the mechanism explaining the gender difference in emotion recognition indeed lies in different information processing styles of women and men (Hall et al., 2003), we would expect women to process information more globally (Hypothesis 5a) and less locally (Hypothesis 5b) than men.

Finally, we tested whether there were differences between the different emotional expressions without putting for- ward hypotheses, because of the lack of consistency in empirical findings (McKelvie, 1995; Prkachin, 2003).

Method

Participants

Participants were 34 university students (15 males, 19 fe- males). One male participant did not come back for the second testing session and was excluded from the analyses.

Because of calibration problems we had to remove four female participants from the analyses on the eye tracking measures, resulting inN= 29 (14 male, 15 female) participants. All participants were native French speakers re- cruited at the University of Lausanne (Mage= 23.18,SD= 3.43). They were remunerated with 30 Swiss francs. All participants reported normal or corrected-to-normal vision and were right-handed.

Stimuli and Materials

Mood Priming

Happy and sad moods were primed with short film scenes.

A meta-analysis showed that film scenes are the most effective and ethically sound way of priming mood (Wester- mann, Spies, Stahl, & Hesse, 1996). Happy mood was primed with a scene from the movie “When Harry Met Sally” in which the actress simulates an orgasm. Sad mood was primed with a scene from the movie “The Champ” in which a little boy cries at his father’s deathbed. As the participants were French speaking, we presented the French versions of these film scenes, which have also been shown to effectively prime mood states (Schaefer, Nils, Sanchez,

& Phillippot, 2010). The film scenes were chosen based on the validation and recommendation of Rottenberg, Ray, and Gross (2007) and both scenes were 2.46 min long.

Emotional Faces

We used stimuli from the DANVA 2-AF (Nowicki & Duke, 1994) and the Ekman and Friesen (1976) series. The P. C. Schmid et al.: How Moods Affect Information Processing 225

(6)

DANVA consists of 24 pictures of emotional facial expressions: happiness, fear, anger, and sadness. These 24 stimuli were combined with 40 stimuli from the Ekman and Frie- sen series of basic emotions expressing the same four emotions. The stimulus sets were divided into two subsets (one for each of the two testing sessions). Each subset consisted of 12 stimuli from the DANVA and 20 stimuli from the Ekman and Friesen series containing equal numbers of pictures of each emotion. The pictures were 14 × 21 cm in size, so the visual angle subtended approximately 13 ° × 20 °.

Because we were only interested in participants’ eye scan patterns while they were looking at a face, we removed irrelevant features such as hair and clothing from the stimulus faces. The modified pictures consisted of a standard- ized oval shape containing only the face (see Figure 1).

Eye Tracking

To register eye movements we used a Hi-Speed 1250 tracking system, SMI-SensoMotoric Instruments (Teltow, Ger- many). Only the left eye was recorded, with a sampling rate of 1250 Hz and a spatial resolution of 0.01. We only analyzed eye movements that were performed before the re- sponse to the facial expression was given and only eye data of correctly recognized emotions were analyzed. We de- fined regions of interest for four face features: the left eye, the right eye, the nose, and the mouth. We used two com- plementary eye tracking measures – interfeatural saccade ratio and feature gaze duration – that have previously been used to identify global and local scan patterns when processing faces (Bombari, Mast, & Lobmaier, 2009). Sac- cades that were performed between two regions of interest are referred to as interfeatural saccades; the durations of fixations performed within the regions of interest were add- ed up and are referred to as feature gaze duration.

Task and Procedure

Participants were tested in two sessions on different days (1–2 days apart). On arrival at the first session, participants gave their informed consent. At the beginning of the first session, participants were familiarized with the four re-

sponse keys (happy, sad, angry, and fearful) for the emotion recognition task. The participants learned the sequence of the keys by heart so that they would not have to look down at the keys during eye tracking. They were then assigned to one of two mood primings, happy or sad. Happy and sad mood priming were both achieved by showing the participants fragments of films. To check whether mood priming was successful, we asked the participants “How do you feel at this moment?” (on a 6-point Likert scale of 1 =extremely sadto 6 = extremely happy) after they had watched the films. Participants then performed Part 1 of the emotion recognition task, in which participants indicated by button press which emotion the presented face expressed. Each face was presented in the center of the screen for 4 s, fol- lowed by a 5 s mask (a face shape). Participants were al- lowed to respond during stimulus exposure or during the mask period (the answer time frame, thus, totaled 9 s). Part 2 was presented to the participants in the second session, in which they were primed with the alternative mood (sad, if they saw a happy film in Session 1 and happy, if they saw a sad film in Session 1). Participants’ key presses and eye movements were recorded in both sessions.

Statistical Analyses

Accuracy

Participants’ emotion recognition accuracy for the two test sets was assessed using a nonparametric measure of dis- criminability (A’). We used the formula suggested by Snodgrass, Levy-Berger, and Haydon (1985): A’ = 1/2 + [(pHit –pFA)*(1 + pHit-pFA)]/[4pHit*(1-pFA)], in which pHit refers to the proportion of hits and pFA to the proportion of false alarms. Values range from 0 to 1, with a value of 0.5 indicating chance performance.

Eye Movements

Interfeatural Saccade Ratio

We calculated a ratio comparing the amount of interfeatural saccades divided by the total number of saccades performed (in order to correct for the varying exposure times that result from the different reaction times). The higher the score on this ratio, the more global the information processing style is (i.e., the more information is integrated to form an overall impression).

Feature Gaze Duration

Feature gaze duration refers to the mean time spent looking at the features. Consecutive fixations that were performed within the same feature were first summed up, and then the mean was calculated over all features. Fixations that were Figure 1.A selection of the modified stimuli used in the

emotion recognition task (left: fearful expression; right:

happy expression).

226 P. C. Schmid et al.: How Moods Affect Information Processing

(7)

shorter than 100 ms were not taken into account. Longer feature gaze duration indicates more local processing.

To test for relations between information processing and emotion recognition accuracy (Hypothesis 1), we comput- ed correlations between the interfeatural saccade ratio, feature gaze duration, and sensitivity A’. We further calculated mixed model ANCOVAs with gender as the between-subjects factor and mood priming (happy vs. sad mood) as the within-subjects factor, and with either the interfeatural saccade ratio (global processing), feature gaze duration (local processing), or sensitivity A’ (emotion recognition accuracy) as the dependent variable (Hypotheses 2–5). Mood priming order and test part order were entered as covariates in the analyses.

To test for differential effects on the four facial expressions, we separately calculated three 2 (Mood Priming:

happy vs. sad) × 2 (Gender) × 4 (Facial Expression: happiness, sadness, anger, fear) mixed model ANCOVAs for sensitivity (A’), feature gaze duration, and the interfeatural saccade ratio as the dependent variables. Mood priming order and test part order were again entered as covariates in the analyses. Correlations of sensitivity (A’) with feature gaze duration and interfeatural saccade ratio were also calculated separately for each of the four stimulus emotions for participants in a happy and a sad mood.

Results

Table 1 presents the means and standard deviations of emotion recognition accuracy, interfeatural saccade ratio, and feature gaze duration for men and women separately and for happy and sad mood separately. The overall emotion recognition accuracy A’ was .90 (SD= .04).

Manipulation Checks

One-samplet-tests against 3.5 (neutral mood) showed that we successfully manipulated participants’ mood. Partici- pants felt happy after happy mood priming (M= 4.35),t(30)

= 8.64,p< .001, and sad after sad mood priming (M= 3.06), t(32) = 3.83,p= .001.

Relation Between Information Processing Styles and Emotion Recognition Accuracy

The interfeatural saccade ratio (global processing) correlated positively with sensitivity A’ (emotion recognition accuracy) when participants were primed with a sad mood,r(29) = .41, p= .03, but not when they were primed with a happy mood, r(29) = .17,p= .39. Longer feature gaze duration (local processing) was related to decreased Sensitivity A’,r(29) = –.44, p= .02, when participants were primed with a sad mood.

However, no relation was found for participants primed with a happy mood,r(29) = .09,p = .66. Note that there were negative correlations between interfeatural saccade ratio and feature gaze duration for happy mood,r(29) = –.50,p< .01, and for sad mood,r(29) = –.57,p< .01.

Mood Effects

Mood priming had no effect on sensitivity A’,F(1, 29) = 0.99,p = .32, but there was a significant main effect of mood priming on the interfeatural saccade ratio (global processing),F(1, 25) = 6.41,p= .02, indicating that happy mood priming resulted in more global processing (M= .40) compared to sad mood priming (M= .39). Mood priming had no effect on feature gaze duration (local processing), F(1, 25) = 0.02,p= .89.

Gender Effects

The main effect of gender for sensitivity A’ showed that women performed better than men (M= .91 for women and M= .88 for men), F(1, 29) = 3.90,p= .05. A trend was found for the interfeatural saccade ratio,F(1, 25) = 3.09,p

= .09, showing that women tended to process information more globally (M = .45) than men (M = .34). Moreover, gender significantly affected feature gaze duration, F(1, 25) = 7.99,p< .01; women processed information less locally (M= 320.40) than men (M= 657.20).

Mood × Gender Interactions

We found a significant mood priming by gender interaction for sensitivity A’,F(1, 29) = 3.85,p= .05. Men recognized Table 1

Means (standard deviations) for men’s and women’s emotion recognition sensitivity A’, interfeatural saccade ratio, and feature gaze duration in happy and sad moods

Mood Happy Sad

Sex Men Women Men Women

Sensitivity A’ .894 (.053) .907 (.049) .872 (.047) .914 (.032)

Interfeatural saccades .339 (.214) .458 (.143) .364 (.190) 417 (.202)

Feature gaze duration 602.32 (444.45) 357.05 (62.21) 621.86 (482.63) 367.88 (68.74)

P. C. Schmid et al.: How Moods Affect Information Processing 227

(8)

emotions better when they were in a happy mood (M= .90) than when in a sad mood (M= .87),t(13) = 1.82,p= .09.

Women’s mood did not affect their emotion recognition accuracy (M= .91 when they were in a happy mood andM

= .91 when in a sad mood),t(18) = .624,p= .54. The interactions between interfeatural saccade ratio and feature gaze duration were not significant,Fs < 0.11,ps > .74.

Facial Expression-Specific Analyses

No significant facial expression-specific mood priming or gender effects were found on the three dependent variables (sensitivity A’, interfeatural saccade ratio, feature gaze duration), allFs < 2.54, all ps > .06. Feature gaze duration correlated negatively with sensitivity A’ for fearful faces when participants were in a sad mood, r(29) = –.48,p = .01. Shorter feature gaze duration was, therefore, related to better recognition accuracy for fearful faces when participants were primed with a sad mood. The interfeatural saccade ratio was negatively related to sensitivity A’ for happy faces when participants were in a happy mood, r(29) = –.36,p= .05, meaning that the fewer interfeatural saccades participants in a happy mood performed, the better they recognized the happy emotions. None of the other correlations were significant (allps > .13).

Effects of Covariates

The only significant effects for the covariates were an interaction of mood priming by mood priming order for the interfeatural saccade ratio,F(1, 25) = 10.53,p= .003, and a mood priming order main effect for feature gaze duration, F(1, 25) = 4.62,p= .041.

Discussion

In this study, we investigated whether mood states, gender, and information processing styles influence emotion recognition accuracy. We expected a positive correlation between global information processing and emotion recognition accuracy and a negative correlation between local processing and emotion recognition accuracy(Hypotheses 1a and 1b). These hypotheses were only confirmed for the sad but not for the happy mood priming condition. As expected, a happy mood resulted in more global information processing than a sad mood (Hypothesis 2a). However, contrary to our expectation, a happy mood did not entail less local processing than a sad mood (Hypothesis 2b). We had predicted that participants would recognize emotions more accurately when in a happy mood than a sad mood (Hypothesis 3), and our results confirmed this hypothesis for men but not for women. As expected, we found that women outperformed men on the emotion recognition task (Hypothesis

4). In line with our expectations, women tended to process information more globally (Hypotheses 5a) and processed information significantly less locally than men (Hypothe- ses 5b).

We showed that a global information processing style is more favorable for emotion recognition than a local information processing style. However, this link could only be found when participants were primed with a sad mood, but not when they were primed with a happy mood. Existing research on the composite effect of emotion recognition has shown that global face information is difficult to ignore (Calder et al., 2000). Research on the face inversion effect suggests that hindering global information processing de- creases emotion recognition accuracy (Prkachin, 2003).

Our study further supports the assumption that global information processing is important for emotion recognition by using an eye tracking method. Unlike Calder et al.

(2000) and Prkachin (2003), we did not manipulate the information in the faces, but looked at whether participants’

eye scan pattern was related to accuracy. Using upright, nonmanipulated faces as stimuli is a strength of our study as it provides more ecological validity.

We believe that global processing was beneficial for emotion recognition only when participants were sad and not when they were happy because happy people, in general, engage in more global processing (Gasper & Clore, 2002). Therefore, if our happy-mood-priming primed participants to use a global processing style, then we might have produced a ceiling effect, explaining the lack of a correlation between global processing and accuracy when participants were happy. Note that we did, indeed, find that participants processed more globally after being primed with a happy mood compared to after being primed with a sad mood. It is already well-established that a happy mood triggers global and automatic and a sad mood local and analytic processing styles in different domains (Bless et al., 1996; De Vries et al., 2008; Gasper & Clore, 2002). More- over, it has often been argued that information processing style is responsible for the effects of sad or depressed mood on emotion recognition and interpersonal judgments in general (Ambady & Gray, 2002; Chepenik et al., 2007; For- gas, 1992). An alternative explanation for the lack of significant findings for the happy mood condition might be that the happy mood manipulation was less strong than the sad mood manipulation. Some mood induction procedures are known to produce mood effects that differ in strength;

however, this is usually not the case for the priming method we used, which employs happy and sad mood film scenes (for a review, see Gerrards-Hesse, Spies, & Hesse, 1994;

or also Gross & Levenson, 1995).

We failed to show a direct link between mood and emotion recognition accuracy in the present study. Previous studies either found mood-congruity effects or decreased emotion recognition accuracy for people in a sad mood.

What might account for our nonsignificant finding? Be- cause we used a wider range of facial expressions (happy, sad, fearful, and angry expressions) than previous studies, 228 P. C. Schmid et al.: How Moods Affect Information Processing

(9)

we may have washed out specific effects. For example, Schmid and Schmid Mast (2010), who found mood-congruity effects, only included sad and happy facial expressions as stimuli. Chepenik et al. (2007) used a stimulus set similar to the one used in the present study and found that a sad mood decreased emotion recognition compared to a neutral mood condition. However, the happy mood was not manipulated in the Chepenik et al. study and it is possible that both happy and sad moods have detrimental effects on emotion recognition accuracy.

One limitation of the present study is the lack of a control group in neutral mood; we cannot, therefore, speculate whether both happy and sad moods affected performance negatively. But we can speculate why participants did not perform better when in a happy mood (when they processed more globally, as shown by us) than when in a sad mood. Detrimental effects of mood on task performance have often been explained with the cognitive load theory (Sweller, 1988): The experience of emotions (positive and negative) might impose cognitive load and affect working memory (Um, Song, & Plass, 2007). For example, Weg- ner, Erber, and Zanakos (1993) showed that regulation of mood states after positive and negative mood priming re- quires cognitive resources that cannot be invested in other tasks. Cognitive load is known to decrease emotion recognition accuracy (Phillips, Channon, Tunstall, Heden- strom, & Lyons, 2008). In our experiment, happy participants performed more global eye movements and, therefore, gained more reliable information than participants in a sad mood, but if cognitive load was high (because participants were regulating their mood), perhaps they were not able to take advantage of this information.

In addition to the mood effects on information processing, we predicted that gender would affect information processing. In line with the literature (McClure, 2000), women outperformed men in our task. Previous research has shown that men and women differ in brain activation during emotion recognition, suggesting that they use different processing styles (Hall et al., 2003). However, no study has yet examined whether men and women differ in the way they look at the stimulus. We showed that women look at faces in a different way using a more global pattern of scanning. This suggests that gender differences in information processing not only appear in brain data, but also on a behavioral level when participants are scanning the faces.

The present study helps us to better understand the mechanisms underlying accurate emotion recognition.

The ability to recognize other people’s emotions is important for personal and professional success and well-being (Byron, Terranova, & Nowicki, 2007; Carton, Kessler, &

Pape, 1999). Thus, the current findings have important applied implications. Understanding the mechanism and moderating factors of correct emotion recognition can help us design effective training techniques aimed at im- proving emotion recognition skills.

References

Ambady, N., & Gray, H. M. (2002). On being sad and mistaken:

Mood effects on the accuracy of thin-slice judgments.Journal of Personality and Social Psychology, 83, 947–961. doi 10.1037//0022-3514.83.4.947

Bless, H., Clore, G. L., Schwarz, N., Golsiano, V., Rabe, C., &

Wölk, M. (1996). Mood and the use of scripts: Does happy mood make people really mindless?Journal of Personality and Social Psychology, 63, 585–595. doi 10.1037/0022- 3514.71.4.665

Bombari, D., Mast, F. W., & Lobmaier, J. S. (2009). Featural, configural, and holistic face processing strategies evoke different scan patterns.Perception, 38, 1508–1521. doi 10.1068/p6117 Bouhuys, A. L., Bloem, G. M., & Groothuis, T. G. G. (1995). In- duction of depressed and elated mood by music influences the perception of facial emotional expressions in healthy subjects.

Journal of Affective Disorders, 33, 215–226. doi 10.1016/

0165-0327(94)00092-N

Bouhuys, A. L., Geerts, E., & Gordijn, M. C. M. (1999). De- pressed patients’ perception of facial emotions in depressed and remitted states are associated with relapse: A longitudinal study. The Journal of Nervous and Mental Disease, 187, 595–602. doi 00005053-199910000-00002

Brosnan, M. J., Scott, F. J., Fox, S., & Pye, J. (2004). Gestalt processing in autism: Failure to process perceptual relationships and the implications for contextual understanding.Journal of Child Psychology and Psychiatry, 45, 459–469. doi 10.1111/j.1469- 7610.2004.00237.x

Byron, K., Terranova, S., & Nowicki, S. J. (2007). Nonverbal emotion recognition and salespersons: Linking ability to per- ceived and actual success.Journal of Applied Social Psychol- ogy, 37, 2600–2619. doi 10.1111/j.1559-1816.2007.00272.x Calder, A. J., Young, A. W., Keane, J., & Dean, M. (2000). Con-

figural information in facial expression perception.Journal of Experimental Psychology: Human Perception and Perfor- mance, 26, 527–551. doi 10.1037/0096-1523.26.2.527 Carton, J. S., Kessler, E. A., & Pape, C. L. (1999). Nonverbal decod-

ing skills and relationship well-being in adults.Journal of Non- verbal Behavior, 23, 91–100. doi 10.1023/A:1021339410262 Chepenik, L. G., Cornew, L. A., & Farah, M. J. (2007). The influ-

ence of sad mood on cognition.Emotion, 7, 802–811. doi 10.1037/1528-3542.7.4.802

Clore, G. L., Wyer, R. S., Dienes, B., Gasper, K., Gohm, C., &

Isbell, L. (2001). Affective feelings as feedback: Some cognitive consequences. In L. L. Martin & G. L. Clore (Eds.),The- ories of mood and cognition: A user’s guide(pp. 27–62). Mah- wah, NJ: Erlbaum.

Damasio, A. R. (1994).Descartes error: Emotion, reason, and the human brain. New York: Harper Collins.

De Vries, M., Holland, R. W., & Witteman, C. L. (2008). Fitting decisions: Mood and intuitive versus deliberative decision strategies.Cognition and Emotion, 22, 931–943. doi 10.1080/

02699930701552580

Ekman, P., & Friesen, W. V. (1976).Pictures of facial affect. Palo Alto, CA: Consulting Psychologists Press.

Forgas, J. P. (1992). On mood and peculiar people: Affect and person typicality in impression formation.Journal of Person- ality and Social Psychology, 62, 863–875. doi 10.1037/0022- 3514.62.5.863

(10)

Förster, J., & Dannenberg, L. (2010). GLOMOsys: A systems account of global versus local processing.Psychological Inquiry, 21, 175–197. doi 10.1080/1047840X.2010.487849

Freire, A., Lee, K., & Symons, L. A. (2000). The face-inversion effect as a deficit in the encoding of configural information:

Direct evidence.Perception, 29, 159–170. doi 10.1068/p3012 Gasper, K., & Clore, G. L. (2002). Attending to the big picture:

Mood and global versus local processing of visual information.

Psychological Science, 13, 34–40. doi 10.1111/1467-9280.

00406

Gerrards-Hesse, A., Spies, K., & Hesse, F. W. (1994). Experimen- tal inductions of emotional states and their effectiveness: A review. British Journal of Psychology, 85, 55–78. doi 10.1111/j.2044-8295.1994.tb02508.x

Gilligan, C. (1982).In a different voice. Cambridge, MA: Harvard University Press.

Gross, J. J., & Levenson, R. W. (1995). Emotion elicitation using films.Cognition and Emotion, 9, 87–108. doi 10.1037/0021- 843X.106.1.95

Hall, G. B. C., Witelson, S. F., Szechtman, H., & Nahmias, C.

(2003). Sex differences in functional activation patterns re- vealed by increased emotion processing demands.Brain Imag- ing, 15, 219–223. doi 10.1097/01.wnr.0000101310.64109.94 Heimberg, C., Gur, R. E., Erwin, R. J., Shtasel, D. L., & Gur, R. C.

(1992). Facial emotion discrimination: III. Behavioral findings in schizophrenia. Psychiatry Research, 42, 253–265. doi 10.1016/0165-1781(92)90117-L

Herba, C., & Phillips, M. (2004). Annotation: Development of facial expression recognition from childhood to adolescence: Behav- ioural and neurological perspectives.Journal of Child Psycholo- gy and Psychiatry, 45, 1185–1198. doi 10.1111/j.1469- 7610.2004.00316.x

Keltner, D., & Kring, A. (1998). Emotion, social function, and psychopathology.General Psychological Review, 2, 320–342.

doi 10.1037/1089-2680.2.3.320

Kimchi, R., & Palmer, S. E. (1982). Form and texture in hierar- chically constructed patterns.Journal of Experimental Psy- chology: Human Perception and Performance, 8, 521–535.

doi 10.1037/0096-1523.8.4.521

Ledoux, J. E. (2000). Emotion circuits in the brain.Annual Review of Neuroscience, 23, 155–184. doi 10.1146/annurev.neuro.

23.1.155

Lembke, A., & Ketter, T. A. (2002). Impaired recognition of facial emotion in mania. American Journal of Psychiatry, 159, 302–304. doi 10.1176/appi.ajp. 159.2.302

Leppänen, J. M., & Hietanen, J. (2004). Positive facial expressions are recognized faster than negative facial expressions, but why?

Psychological Research, 69, 22–29. doi 10.1007/s00426-003- 0157-2

McClure, E. B. (2000). A meta-analytic review of sex differences in facial expression processing and their development in in- fants, children, and adolescents.Psychological Bulletin, 126, 424–453. doi 10.1037//0033-2909.126.3.424

McKelvie, S. J. (1995). Emotional expression in upside-down faces: Evidence for configural and componential processing.Brit- ish Journal of Social Psychology, 34, 325–334. doi 10.1111/j.2044-8309.1995.tb01067.x

Moore, D. G. (2001). Reassessing emotion recognition performance in people with mental retardation: A review.American Journal of Mental Retardation, 106, 481–502. doi 10.1352/0895- 8017(2001)106

Niedenthal, P. M., Halberstadt, J. B., Margolin, J., & Innes-Ker, A. H. (2000). Emotional state and the detection of change in facial expression of emotion.European Journal of Social Psychology, 30, 211–222. doi 10.1002/(SICI)1099-0992(200003/04)30:2

<211::AID-EJSP988>3.0.CO;2-3

Nowicki, S. J., & Duke, M. P. (1994). Individual differences in the nonverbal communication of affect: The Diagnostic Analysis of Nonverbal Accuracy Scale.Journal of Nonverbal Behavior, 18, 9–35. doi 10.1007/BF02169077

Phillips, L. H., Channon, S., Tunstall, M., Hedenstrom, A., & Ly- ons, K. (2008). The role of working memory in decoding emotions.Emotion, 8, 184–191. doi 10.1037/1528-3542.8.2.184 Prkachin, G. C. (2003). The effect of orientation on detection and

identification of facial expressions of emotion.British Journal of Psychology, 94, 45–62. doi 10.1348/000712603762842093 Putrevu, S. (2001). Exploring the origins and information processing differences between men and women: Implications for advertisers.Journal of Marketing Management, 10, 47–66.

Rottenberg, J., Ray, R. R., & Gross, J. J. (2007). Emotion elicitation using films. In J. A. Coan & J. J. B. Allen (Eds.),The hand- book of emotion elicitation and assessment. New York: Oxford University Press.

Schaefer, A., Nils, F., Sanchez, X., & Phillippot, P. (2010). As- sessing the effectiveness of a large database of emotion-elicit- ing films: A new tool for emotion researchers.Cognition and Emotion, 24, 1153–1172. doi 10.1080/02699930903274322 Schmid, P. C., & Schmid Mast, M. (2010). Mood effects on emo-

tion recognition.Motivation and Emotion, 34, 288–292. doi 10.1007/s11031-010-9170-0

Schwarz, N. (1990). Feelings as information: Informational and motivational functions of affective states. In E. T. Higgins &

R. Sorrentino (Eds.),Handbook of motivation and cognition:

Foundations of social behavior(Vol. 2, pp. 527–561). New York: Guilford.

Searcy, J. H., & Bartlett, J. C. (1996). Inversion and processing of component and spatial-relational information in faces.Journal of Experimental Psychology: Human Perception and Perfor- mance, 22, 904–915. doi 10.1037/0096-1523.22.4.904 Snodgrass, J. G., Levy-Berger, G., & Haydon, M. (1985).Human

experimental psychology. New York: Oxford University Press.

Surguladze, S. A., Young, A. W., Senior, C., Brébion, G., Travis, M. J., & Phillips, M. L. (2004). Recognition accuracy and re- sponse bias to happy and sad facial expressions in patients with major depression. Neuropsychology, 18, 212–218. doi 10.1037/0894-4105.18.2.212

Sweller, J. (1988). Cognitive load during problem solving: Effects on learning.Cognitive Science, 12, 257–285. doi 10.1207/

s15516709cog1202_4

Um, E. R., Song, H., & Plass, J. (2007). The effect of positive emotions on multimedia learning. In C. Montgomerie & J.

Seale (Eds.),Proceedings of World Conference on Educational Multimedia, Hypermedia and Telecommunications 2007 (pp. 4176–4185). Chesapeake, VA: AACE. Retrieved from http://www.editlib.org/p/25979

Wegner, D. M., Erber, R., & Zanakos, S. (1993). Ironic processes in the mental control of mood and mood-related thought.Jour- nal of Personality and Social Psychology, 65, 1093–1104. doi 10.1037/0022-3514.65.6.1093

Westermann, R., Spies, K., Stahl, G., & Hesse, F. W. (1996). Rel- ative effectiveness and validity of mood induction procedures:

A meta-analysis.European Journal of Social Psychology, 26, 230 P. C. Schmid et al.: How Moods Affect Information Processing

(11)

557–580. doi 10.1002/(SICI)1099-0992(199607)26:4%3C 557::AID-EJSP769%3E3.0.CO;2-4

Yin, R. K. (1969). Looking at upside-down faces.Journal of Ex- perimental Psychology, 81, 141–145. doi 10.1037/h0027474 Zuroff, D. C., & Colussy, S. A. (1986). Emotional recognition in

schizophrenia and depressed inpatients.Journal of Clinical Psy- chology, 42, 411–416. doi 10.1002/1097-4679(198605)42:3

<411::AID-JCLP2270420302>3.0.CO;2-T

Petra Claudia Schmid

Department of Work and Organizational Psychology University of Neuchatel

Rue Emile-Argand 11 CH - 2000 Neuchatel Switzerland

[email protected]

View publication stats View publication stats