The Journal of Positive Psychology, July 2006; 1(3): 150–159

The experience of positive emotion is associated with the automatic processing of positive emotional words

GREGORY P. STRAUSS & DANIEL N. ALLEN University of Nevada Las Vegas, USA

Abstract The current study examines the relationship between attention bias for positive emotional words and self-reported emotional experience. Previous research suggests that the experience of positive emotion momentarily broadens cognitive processes, potentially allowing individuals to build an array of enduring personal resources. However, it is unknown whether the experience of positive emotion also broadens emotional information processing. Participants included 60 healthy undergraduate students who completed measures of psychopathology, self-reported emotional experience, and an emotional Stroop task designed to measure attentional bias to positive and negative emotional information. Results indicate significant associations between reaction times for high-intensity happiness words and self-reported high levels of positive emotion and low levels of negative emotions. These associations were not present for low intensity happiness words. Findings suggest that individuals who experience high levels of positive emotion and low levels of negative emotion demonstrate an attention bias for positive information and, from an information processing perspective, provide insight into the manner in which positive emotions broaden cognitive processes. Keywords: Positive emotion, happiness, flourishing, attention bias, emotional Stroop

Introduction Increased interest in the role that positive emotions play in promoting optimal adjustment and functioning has lead to a number of investigations focusing on the impact of positive emotional experience on cognitive processing. Positive emotional experience has been shown to increase verbal fluency (i.e., identifying category words; Greene & Noice, 1988; Hirt, Melton, McDonald, & Harackiewicz, 1996), problem solving and creativity (Isen, Daubman, & Nowicki, 1987), verbal recall memory (Isen et al., 1978; Teasdale & Fogarty, 1979), attentional scope (Fredrickson & Branigan, 2005), and abstractionflexibility (Dreisbach & Goschke, 2004). These findings are consistent with the broaden-and-build theory of positive emotion proposed by Frederickson (1998, 2001), and particularly the broaden component which proposes that the experience of positive affect momentarily broadens cognitive processes, allowing a wider range of thoughts and actions to come to mind (for a recent discussion see Fredrickson & Branigan, 2005). Cognitive broadening afforded by the experience of positive emotion may have significant adaptive value, as it allows individuals to build a variety of enduring personal

resources. For example, momentary experiences of positive affect have been linked to the long-term acquisition of social resources (Aron, Norman, Aron, McKenna, & Heyman, 2000; Johnson & Fredrickson, 2005), better physical health and longevity (Danner, Snowden, & Friesen, 2001; Rozanski & Kubzansky, 2005), and increased psychological resources (Fredrickson, Tugade, Waugh, & Larkin, 2003; Tugade & Fredrickson, 2004). It also appears that as the experience of positive emotion increases relative to the experience of negative emotion then functioning also improves. For example, a number of studies that have calculated positive–negative emotion ratios have found that positive–negative ratios greater than 2.90 are associated with optimal mental health and human flourishing, while lower ratios (below 2.90) may be associated with languishing (Fredrickson & Losada, 2005; Gottman, 1994; Losada, 1999; Schwartz et al., 2002). Thus, the momentary experience of positive emotion in the absence of negative emotion is thought to be adaptive because it allows individuals to think in a broad and flexible manner that facilitates the acquisition of durable personal resources that can be drawn on in times of need. The attentional system may be central

Correspondence: Daniel N. Allen, Department of Psychology, University of Nevada Las Vegas, Las Vegas, NV 89154, USA. Tel: 702 895 1379. Fax: 702 895 0195. E-mail: [email protected] ISSN 1743–9760 print/ISSN 1743–9779 online/06/030150–10 ! 2006 Taylor & Francis DOI: 10.1080/17439760600566016

Positive words and automaticity to this process, allowing individuals to focus on a wide array of information during momentary experiences of positive emotion. Previous research is consistent with the notion that positive emotion broadens attentional scope, as it has been demonstrated that induced positive affect facilitates a global visual processing bias (Fredrickson & Branigan, 2005; Gasper & Clore, 2002). Fredrickson and Branigan (2005) proposed that the attentional broadening effect afforded by positive affect might facilitate the type of broad-minded and flexible thinking styles that promote the acquisition of enduring personal resources, such as coping, that can be drawn upon in times of need. Despite strong evidence suggesting that there is a link between broadened thinking and the building of resources (Fredrickson & Joiner, 2002), it is currently unclear whether or not there is an attentional bias for positive information that is associated with the experience of positive emotion and the acquisition of positive resources. Rather, investigations examining attentional biases for emotional information have focused almost exclusively on negative emotions. Numerous studies indicate that healthy individuals evidence an attention bias for negative information (McKenna & Sharma, 1995, 2004; Myers & McKenna, 1996; Watts, McKenna, Sharrock, & Trezise, 1986), particularly under experimental conditions that produce stress (e.g., time pressure; Sharma & McKenna, 2001). It has also been demonstrated that individuals experiencing psychopathological conditions exhibit attention biases for information that is consistent with their emotional experience (e.g., depression is associated with an attention bias for sadness; see Williams, Matthews, & MacLeod, 1996 for a review). These studies have used an emotional version of the Stroop task (E-Stroop) to demonstrate attention bias, in which emotional words are presented and the individual is instructed to ignore the meaning of the word and name the color the word is printed in, or name a simple picture on which the emotional word is superimposed. The theory behind the E-Stroop task states that words related to an individual’s preoccupations or mood state will draw attention away from the target behavior (e.g., picture or color naming), thereby causing increased reaction times (RTs) in comparison to neutral words and words less related to one’s concerns (Williams et al., 1996). Such an attention bias may take place automatically, with environmental information processed quickly, and outside of conscious awareness (Pratto & John, 1991), thus providing an adaptive advantage by promoting the most adaptive actions under threatening situations. Thus, while emotional Stroop tasks have proven very useful in understanding the manner in which the

151

attention system promotes survival during the experience of negative emotion, little is known regarding whether the attentional system facilitates adaptive functions that result from momentary experiences of positive emotion and whether such biases might lead to the building of enduring personal resources (Fredrickson, 1998). The few studies that have included positive emotional words in emotional Stroop paradigms have methodological limitations that primarily result from a failure to equate positive and negative words for emotional intensity (McKenna & Sharma, 2004; Pratto & John, 1991; White, 1996), with negative words being significantly more intense than positive words. Thus, it is not possible to determine whether differences reported between emotional conditions are due to emotional intensity or valence effects. In fact, only one known study provided an adequate methodological comparison of attention bias for positive and negative words (Martin, Williams, & Clark, 1991). Results suggest that positive emotions do have the capacity to disrupt attention. However, these findings may not generalize to the general population, as the participants in that study reported high levels of anxiety (Martin et al., 1991). A study conducted by MacLeod, Matthews, and Tata (1986) also found that healthy individuals directed their attention away from threat stimuli; however, positive words were not examined to determine whether attention was drawn toward positive information. Thus, additional studies are needed to determine whether positive information is capable of disrupting on-going attention in nonclinical samples, whether these effects are more strongly dependent on state fluctuations in positive emotions, or whether they also occur in the presence of trait features of positive mood. Finally, considering research indicating the importance of the relative balance of positive to negative mood (e.g., Fredrickson & Losada, 2005), it also seems important to examine whether effects are associated solely with levels of positive emotional experience or if the ratio between positive and negative experience is an important consideration. To address these issues, the current study examined the relationship between self-reported emotional experience and attention bias for words representing positive and negative basic emotions. Aims of the current study included: (1) to determine whether emotionally intense happiness words are capable of disrupting attention equal to or greater than negative words, (2) to determine whether attention bias differs between high and low-intensity happiness words, (3) to determine whether state and trait experiences of emotion are differentially associated with attention bias for positive words, and (4) to determine whether positive–negative mood ratios are associated with attention biases for

152

G. P. Strauss and D. N. Allen

positive words. It was hypothesized that greater attention bias (i.e., longer RTs) would result for happiness words than for words from sadness, anger, and anxiety conditions. High-intensity happiness words were also hypothesized to produce a significantly greater attention bias than less intense happiness words, and attention bias for positive words was expected to be positively correlated with the experience of happiness.

overall psychopathology. For the current study, individuals were excluded from participation if they obtained a GSI t-score "63, or if any two clinical scale scores were " a t-score of 63 (using nonpsychiatric population norms). These procedures follow the SCL-90-R rule of caseness, which is designed to screen out positive cases of psychopathology. Emotional experience

Method

Participants Participants included 65 undergraduate students between the ages of 18 and 40 who volunteered to take part in the study for which they were compensated with course credit. Individuals were excluded from the study if they were over 40 years old, spoke English as a second language, had a speech impediment, had inadequate corrected vision (>20/200), reported a current psychiatric or neurological diagnosis, or exhibited severe psychopathology based on screening using the Symptom Checklist 90-Revised (SCL-90-R; Derogatis, 1983). Individuals were also excluded if they had a history of psychiatric disorder or treatment or a history of neurological disorder including traumatic brain injury. Based on these exclusion criteria, four participants were excluded, three who evidenced significant psychopathology on screening and one who reported previous traumatic brain injury. One additional participant was excluded for failure to complete study procedures. The remaining 60 participants were 20.3 years old (SD ¼ 4.1) and had 13.5 years of education (SD ¼ 1.0). Right handed individuals composed 70% of the sample, 65% were female, and 60% were Caucasian, 25% Asian, 10% Hispanic, and 5% African American. Participants had a mean SCL-90-R Global Severity Index (GSI) score of 49.6 (range ¼ 33–62). The study was approved by the University institutional review board. Participants provided written informed consent prior to the completion of study procedures. Measures Participants completed measures designed to assess psychopathology, emotional experience, and attention bias for emotional words. Psychopathology The Symptom Checklist 90 Revised (SCL-90-R; Derogatis, 1983) was included as a screening measure for psychopathology. The SCL-90-R has nine clinical scales that assess psychopathology in specific areas, as well as a Global Severity Index (GSI) which reflects

Emotional experience was measured using singleitem state and trait ratings designed to assess the experience of 4 basic emotions including happiness, sadness, anger, and anxiety. In the current study, a distinction was made between state experience, which was assessed by examining emotional experience at the time of testing, and trait experience, which was assessed by asking the participant to rate his emotional experience on average over the past two weeks. Ratings were made on a 7-point Likerttype scale, with 1 indicating feeling the emotion ‘‘not at all’’ and 7 indicating ‘‘extremely’’ emotional. Single-item ratings were used because they have the advantage of being brief, and afford a valid means of assessing emotional experience (e.g., Lucas, Diener, & Larsen, 2003). Strauss and Allen (unpublished raw data) have also found that these singleitem measures do in fact measure constructs similar to those assessed in popular multi-item scales such as the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988). For example, in a sample of 80 individuals, happiness mood ratings were found to be highly correlated with the PANAS positive affect scales for the past few weeks (r ¼ 0.61; p < 0.001), and mood ratings of discrete negative emotions to be highly correlated with the PANAS Negative Affect scale (Sadness r ¼ 0.50, p < 0.001; Anger r ¼ 0.51, p < 0.001; Anxiety r ¼ 0.32, p ¼ 0.05) (Strauss & Allen, unpublished raw data). In addition to these discrete emotional ratings, positive–negative ratios were also calculated in order to evaluate whether positive mood was responsible for any effects on attention, or if the ratio of positive and negative mood was responsible. For both state and trait ratings, ratios were calculated using the following formula: happiness rating divided by negative composite [sadness þ anxiety þ anger/3]. These ratings reflect the proportion of positive emotion experienced relative to negative emotion. Attention bias Attention bias was assessed using a picture–word version of the emotional Stroop (E-Stroop) task in which emotion words were presented on simple

Positive words and automaticity black and white line drawings, and the participant was instructed to name the line drawing while ignoring the word. A picture-control condition was also administered to determine the interfering effects of words. This condition consisted of 12 pictures that were not altered from their original formats (i.e., stimuli appeared as simple black and white line drawings, with no word superimposed on top of the picture) that were randomly administered within the block of picture–word stimuli. A total of 90 stimuli were presented, with 84 experimental stimuli and six practice stimuli. For the experimental stimuli, six words were chosen from five emotion conditions that included happiness, anger, anxiety, sadness, and neutral. Each of the six words from these five conditions was presented twice on two different pictures, to ensure that interference effects were due to the emotional content of words, rather than the recognition of pictures. To evaluate emotional intensity effects for happiness, 12 additional stimuli were presented that included low-intensity happiness words. The six practice stimuli were constructed in the same manner as the experimental stimuli, with one practice stimulus modeled after each of the six emotional conditions. Words were printed within the pictures and appeared on the computer screen as black line drawings against a white background. Each word was written in uppercase, Arial Black font regular, size 8 point. Stimuli were presented in a semi-random order, with the restriction that no two pictures or word stimuli could be presented consecutively. This order was randomized differently for each participant. Black and white line drawings for the E-Stroop stimuli were selected from the Snodgrass and Vanderwart (1980) stimulus set, which have been normed on adults, and have nameability ratings of 90% or higher. Emotion words were chosen from two stimulus sets. The first was developed by Strauss and Allen (accepted pending revisions) and was used for words included in the happiness, anger, anxiety, sadness, and neutral stimuli. That stimulus set contains 463 words that were rated by 254 participants on a 1–7 Likert-type-scale for emotional intensity (1 ¼ not very emotional; 7 ¼ very emotional) as well as categorized into one of eight discrete emotional categories (happiness, sadness, anger, anxiety, fear, disgust, neutral, and surprise). Words selected for the E-Stroop task in the current study had categorization ratings "70% and an attempt was made to ensure that the emotion categories were equated with regard to emotional intensity. Words used in the low-intensity happiness category were selected from word lists normed for valence by Bellezza, Greenwald, and Banaji (1986). Low intensity happiness words were also rated in the Strauss and Allen (accepted pending revisions) word norm study with regard to emotional

153

intensity and categorization, and found to be highly representative of happiness. All words used in the E-Stroop task are presented in the Appendix. All word conditions were statistically similar with regard to word frequency F(5, 39) ¼ 0.6, p ¼ 0.68 (Kucera & Francis, 1967) and emotional categorization F(5, 41) ¼ 0.8, p ¼ 0.50. Word categories significantly differed with regard to word length, F(5, 41) ¼ 2.7, p ¼ 0.04. However, word length is not thought to influence interference effects in the Stroop task (Yee & Hunt, 1991). As anticipated, word categories differed with regard to emotional intensity F(4, 29) ¼ 47.5, p < 0.001, such that neutral words were significantly less emotionally intense than emotional words, allowing for a valid comparison of emotional and neutral word interference (i.e., the extent to which emotional words grab attention relative to neutral words). Emotional word categories (happiness, sadness, anger, anxiety) did not statistically differ with regard to emotional intensity F(4, 23) ¼ 2.3, p ¼ 0.11, although anxiety words showed a trend toward being less intense, suggesting that word conditions are marginally different in intensity. ANOVA also indicated that the high-intensity happiness words were significantly more intense than low-intensity happiness words F(1, 17) ¼ 14.4, p < 0.01, and that low-intensity happiness words were more intense than neutral words F(1, 17) ¼ 74.2, p < 0.001. As is common in E-Stroop studies, interference effects were represented using interference scores which were calculated for each subject using the formula Emotional Word RT – Neutral Word RT. These interference scores indicate the extent to which individuals have an attention bias for emotional words relative to neutral words. Positive scores reflect that an emotional condition is significantly more interfering than neutral words, ostensibly signifying the presence of an attention bias for that particular emotion. RTs obtained using the picture– word E-Stroop test version have been found to have high test–retest reliability (Strauss, Allen, Jorgensen, & Cramer, 2005). Additionally, picture–word E-Stroop tasks produce interference effects comparable to the color–word E-Stroop tasks (Strauss et al., 2005).

Procedure Participants were tested in a single session that lasted approximately 1 hour. All participants completed the E-Stroop task, SCL-90-R, and emotional experience ratings. Participants first received the E-Stroop task, followed by emotional experience ratings, and the SCL-90-R to ensure that E-Stroop findings were not influenced by priming effects.

154

G. P. Strauss and D. N. Allen

E-Stroop stimuli were presented on a laptop computer (15 inch monitor), with a viewing distance of approximately 60 cm from the monitor with each word appearing in the dimensions 0.6 cm high (0.6 degree of visual angle) and approximately 2 cm wide (2 degrees of visual angle). Verbal response was measured by a voice-operated microphone. To ensure that stimuli were presented at the appropriate interstimulus interval (ISI), a refresh rate detector was connected to the computer. It was found that ISI presentations were accurate to $7 ms. In the E-Stroop task, each participant was first seated in front of the computer screen, and asked to hold the voice-operated microphone in hand while the experimenter read a set of instructions from the computer screen that informed participants of the nature of the Stroop task and instructed them how to appropriately respond. Participants were then informed that a number of practice trials would ensue, and practice stimuli were then presented. After presentation of practice stimuli, participants were informed that the actual test would begin. Each participant then completed the experimental trials. For each trial, the stimulus was triggered when the participant named the picture presented on the computer screen by speaking into the microphone, where upon the experimenter then determined the accuracy of the participant’s response and recorded incorrect responses and microphone errors using a serial response box. Each response was coded as correct if it corresponded with the appropriate picture name. The response was coded as incorrect if the verbal response differed from the appropriate name. Upon the occurrence that the participant’s voice did not trigger the microphone upon sounding, when a noise other than the participant’s response triggered the microphone (e.g., background noise, cough, sneeze), or when the participant made a noise other than a response directed at naming the target (i.e., vocalized pause ‘‘um’’ ‘‘uh’’), the experimenter coded a microphone error. Results

Preliminary analyses All E-Stroop reaction time trials associated with mechanical errors were first eliminated from the data. Approximately 1.6% of the final E-Stroop data set was eliminated due to mechanical errors. Accuracy analyses were then calculated to examine differences in error rate across emotional conditions. ANOVA indicated that emotional conditions did not differ with regard to accuracy, F(6, 59) ¼ 0.9, p ¼ 0.54 (Happiness High Intensity mean ¼ 0.95, Happiness Low Intensity mean ¼ 0.94, Anger mean ¼ 0.95, Anxiety mean ¼ 0.96, Sadness

mean ¼ 0.95, Neutral mean ¼ 0.94, Picture Control mean ¼ 0.95). Incorrect responses were then eliminated to analyse RT for correct responses alone (approximately 4.0% of the dataset). Next, reaction time trials greater than 2.5 standard deviations above each condition’s mean and less than 150 ms were removed from the data as outliers. This encompassed approximately 2.8% of the final data set. A total of approximately 8% of individual RTs were eliminated from final E-Stroop analyses. Finally, for each participant, averages were calculated for RT for each condition. Interference scores (Emotional RT – Neutral RT) were then calculated for emotional conditions and the picture control to reflect the extent to which emotional words grabbed attention relative to neutral words. These interference scores formed the basic data for subsequent analyses. Interference scores were viewed as the primary dependent measure, particularly because picture-naming accuracy approached ceiling for all conditions.

Primary analyses Means and standard deviations were calculated for emotional experience conditions to confirm the existence of the positivity offset (i.e., tendency to report being in a moderately positive mood; Diener & Diener, 1996) in the current sample. Separate ANOVAs were calculated for both state and trait ratings. For state ratings, repeated measures ANOVA indicated significant differences in emotional experience, F(3, 59) ¼ 98.5, p < 0.001; Eta2 ¼ 0.63. Simple contrasts suggest that happiness was experienced significantly more than sadness, anger, and anxiety ( p < 0.001 for all; Happiness mean ¼ 5.02, SD ¼ 1.08; Sadness mean ¼ 1.90, SD ¼ 1.08; Anger mean ¼ 1.67, SD ¼ 1.63; Anxiety mean ¼ 3.02; SD ¼ 1.55). For trait ratings, repeated measures ANOVA indicated significant differences in emotional experience, F(3, 59) ¼ 28.9, p < 0.001; Eta2 ¼ 0.33. Simple contrasts suggest that happiness was experienced significantly more than sadness, anger, and anxiety ( p < 0.001 for all; Happiness mean ¼ 4.83, SD ¼ 1.18; Sadness mean ¼ 2.72, SD ¼ 1.44; Anger mean ¼ 2.67, SD ¼ 1.71; Anxiety mean ¼ 3.30, SD ¼ 1.79). Results indicate that the current sample demonstrated the positivity offset. Attention bias E-Stroop interference scores and standard errors that were calculated for each emotion condition (Emotional Word RT – Neutral Word RT) and the picture control are presented in Figure 1. Using these interference scores, a repeated measures ANOVA

E-Stroop Interference RT (ms)

Positive words and automaticity 90 80 70 60 50 40 30 20 10 0 −10 −20 −30 −40 −50 −60 −70 −80 −90

Hap._Hi

Hap_Lo

Ang.

Anx.

Sad.

155

Pic. Con.

Emotion Figure 1. Means and standard errors for emotional interference conditions. Note. Emotional conditions represent interference scores, calculated as emotional RT – Neutral RT; Negative interference scores reflect that the emotion is less interfering than neutral, while positive interference scores reflect that the emotion is more interfering than neutral, and ostensibly and attention bias; Hap._Hi ¼ Happiness high intensity; Hap_Lo ¼ Happiness low intensity; Sad. ¼ Sadness; Ang. ¼ Anger; Anx. ¼ Anxiety; Pic. Con. ¼ Picture control (i.e., picture without superimposed word).

Table I. Correlations between emotion ratings and E-Stroop interference scores. Emotional Stroop interference conditions Hap_Hi State ratings Happiness Sadness Anger Anxiety Pos/Neg ratio Trait ratings Happiness Sadness Anger Anxiety Pos/Neg ratio

0.26 %0.15 %0.24 %0.25 0.43

(0.05) (0.07) (0.056) (0.001)

0.14 %0.05 %0.15 %0.41 (0.001) 0.26 (0.046)

Hap_Lo

Anger

Anxiety

Sadness

Picture control

0.10 %0.01 %0.06 %0.14 0.22

0.18 %0.01 %0.22 %0.06 0.15

0.02 %0.04 %0.09 %0.10 0.15

0.01 %0.04 %0.10 %0.20 0.19

%0.16 %0.09 %0.12 %0.32 (0.01) 0.19

%0.02 %0.05 %0.05 %0.29 (0.03) 0.09

0.07 0.04 %0.14 %0.18 0.18

%0.01 0.11 %0.01 %0.14 0.07

0.02 %0.03 %0.03 %0.32 (0.01) 0.19

0.03 %0.15 %0.20 %0.28 (0.03) 0.20

Note: P-values presented in parentheses; State Ratings ¼ emotional experience at the time of testing; Trait ratings ¼ emotional experience on average over the past 2 weeks; Hap._Hi ¼ Happiness high-intensity words; Hap_Lo ¼ Happiness low intensity words; Pos/Neg Ratio ¼ ratio of positive to negative emotion (Happiness Rating/Negative Composite).

was conducted to assess differences in attention bias across emotion conditions. Results indicated significant differences among emotion conditions, F(5, 59) ¼ 20.1, p < 0.001 (Eta2 ¼ 0.25). Simple effects were calculated to test specific hypotheses related to attention and positive emotions. With highintensity happiness as the contrast variable, results indicated that happiness RTs were significantly faster than low-intensity happiness (F ¼ 56.6; p < 0.001; Eta2 ¼ 0.49), anger (F ¼ 22.9; p < 0.001; Eta2 ¼ 0.28), anxiety (F ¼12.1; p ¼ 0.001;

sadness (F ¼ 43.9; p < 0.001; Eta2 ¼ 0.17), Eta2 ¼ 0.43), and picture control (F ¼ 4.4; p ¼ 0.041; Eta2 ¼ 0.07) conditions. Results indicate that high-intensity happiness words were significantly less interfering than sadness, anger, anxiety, or low-intensity happiness words. Correlations between emotional experience and E-Stroop interference scores (i.e., RT emotional – RT neutral) are presented in Table I. Results indicate a significant positive correlation between state happiness and interference scores

156

G. P. Strauss and D. N. Allen

for high-intensity happiness words, suggesting that greater momentary experience of happiness is associated with greater attention bias for happiness words. The correlations between the positive/negative ratio for both affect and mood were positively correlated with high-intensity happiness interference scores, suggesting that greater experience of positive relative to negative emotion is associated with greater attention bias for happiness words. Several significant correlations were also found between interference scores and the experience of anxiety. Both state and trait ratings of anxiety were negatively correlated with the picture-control condition, indicating that higher levels of state and trait anxiety are associated with faster picture recognition. Trait anxiety was also negatively correlated with interference for high- and low-intensity happiness words and sadness words, indicating that higher levels of trait and state anxiety are associated with less attention bias for positive words. Discussion Results did not support the hypothesis that participants would evidence a greater attention bias for intense happiness words relative to low-intensity happiness, sadness, anger, or anxiety words. Rather, results indicated that intense happiness words were less interfering than other emotional conditions, potentially suggesting that the majority of individuals fail to evidence an attention bias for happiness. Correlational analyses were used to further elucidate these differences in main effects for attention bias, and to determine whether state and trait experiences of emotion are differentially associated with attention bias for positive words. Correlational analyses provided important information regarding the influence of emotional experience on attention bias, as they indicated that the momentary experience of positive affect was associated with an attention bias for happiness. These findings indicate that happiness words were processed more automatically (i.e., longer RTs) by individuals experiencing higher levels of positive affect and less automatically by individuals experiencing lower levels of positive affect. Additionally, the relationship between the experience of happiness and attention bias for happiness was unique to momentary affective experience, as correlations were nonsignificant for trait ratings of happiness. The specificity of the relationship noted for state, but not trait experience, may suggest that individuals are drawn toward positive information during intense transient experiences of happiness, but not at all times, despite typically being in a moderately positive mood. These findings may be of importance to Fredrickson’s (1998, 2001) broaden-and-build

theory of positive emotions because they imply that individuals may be primarily drawn to attend to positive information during momentary experiences of happiness that are associated with cognitive broadening. It may be that momentary experiences of positive affect broaden the scope of attentional focus, thereby allowing positive information to be automatized more efficiently. The current study provides support for this proposition because attention was disrupted by positive information, suggesting that positive information is processed automatically by individuals who experience high levels of positive affect, but not by individuals experiencing low levels of positive affect. This finding is consistent with the notion that the cognitive broadening effect afforded by the experience of positive emotion is associated with the automatic processing of positive information. Results also provide some indication that the automatic processing of positive information is related to the building of enduring personal resources. Findings provide support for this notion, as the automatic processing of happiness words was positively associated with the relative balance of positive to negative trait experience, as calculated through a positive–negative mood ratio. In previous studies, higher positive–negative mood ratios have been linked to the development of enduring personal resources, such as good interpersonal relationships (Gottman, 1994) and flourishing (Fredrickson & Losada, 2005). As such, the current findings may provide some indirect and preliminary evidence that the automatic processing of happiness is related to the development, or possibly the maintenance, of enduring personal resources. Although the current data provide no direct indication for how the automatic processing of positive information allows enduring resources to be built, it is plausible to expect that these enduring resources may be built after individuals repeatedly attend to positive information within the environment. After repeatedly attending to positive information, positive beliefs or attitudes may come to be automatized, potentially providing the type of positive thinking that facilitates the acquisition and maintenance of a variety of personal resources. For example, by repeatedly focusing on positive self-relevant information in the environment during the momentary experience of happiness, individuals may come to automatize a set of personal beliefs or strengths that can be drawn on in times of need. Findings also give some indication that the automatic processing of positive information may serve as a link between broadening and building. Although previous research provides strong support for the notion that positive emotions broaden cognitive functioning (Fredrickson, 1998, 2001;

Positive words and automaticity Fredrickson & Branigan, 2005), and that broadened states are related to the acquisition of enduring personal resources (Fredrickson & Joiner, 2002), it has yet to be determined whether there is a mechanism that allows enduring resources to be built as the result of numerous momentary experiences of positive emotion. Since the current data indicate that the automatic processing of intense happiness words is associated with both state and trait positive/ negative ratios, it can be inferred that automatic processing of positive information is related to both broadening (which occurs during state experiences) and building (which is related to the trait ratio). It may be that the momentary cognitive broadening effect afforded by positive emotion allows individuals to automatize positive information, which when done repeatedly facilitates the acquisition of a variety of enduring personal resources. Finally, findings have important implications for research on human flourishing and languishing. The current findings suggest that individuals who report a high-ratio of positive to negative mood may process intense positive information automatically, as intense happy words grabbed attention, even when the experimental task required conscious processing of other information. Since higher positive/negative ratios (>2.90; Fredrickson & Losada, 2005) have been associated with flourishing, this finding may be consistent with the notion that higher levels of flourishing are associated with greater automatization of positive information. Additionally, a lack of flourishing, which may be conceptualized as a languishing experience (Keyes & Lopez, 2002), may also be characterized by a failure to process positive information automatically, as a low positive/ negative ratio was associated with a lack of an attention bias for happiness. The failure to find an overall attention bias for intense happiness words in the current sample may be due to the greater proportion of non-flourishing to flourishing individuals examined. These findings suggest that various mental health categories may be associated with differences in emotional information processing (e.g., attention, memory), and that biases toward processing positive information may differentiate those who flourish and languish (e.g., flourishers have an attention bias for happiness while languishers do not). The current findings also provide some suggestion that mental health could be improved in individuals who are languishing by enhancing their ability to attend to positive information. This may have particular implications for psychotherapy, as clinicians may be able to re-direct attentional focus to more positive information, thereby allowing individuals who are languishing to process positive

157

information automatically. If this process were to be done repeatedly, it is possible that individuals who are languishing could build resources that would increase mental health. These statements regarding relationships between flourishing and attention bias for happiness would be strengthened by future studies that examined attention bias using Keyes’ (Keyes, 2005; Keyes & Lopez, 2002) methods for differentiating flourishing and languishing. The current findings have several limitations. First, the single-item mood ratings used in the current study may not provide a broad-based assessment of the experience of positive emotion. Findings reported here would be strengthened if additional studies replicated the relationship between automatic processing and emotional experience using multipleitem measures. Second, the use of emotional experience ratings that require use of an affective label has been met with some criticism lately (Sabini & Silver, 2005). Future studies may want to measure affective experience in relation to additional methods, such as rating emotional pictures or sounds. Third, there is some debate as to whether the E-Stroop task measures automatic attentional or memory processes. The use of additional cognitive measures, such as the dot probe task (MacLeod et al., 1986), may allow for the determination of whether automatic attention or memory is involved with broadening and building. The current findings are also limited by their correlational nature. Experimental designs utilizing mood induction procedures are needed before the theoretical interpretations posed here can be validated. Nonetheless, findings reported in the current study provide several interesting new directions for the study of positive emotions and flourishing, and point to the importance of applying cognitive information processing techniques to the field of positive psychology.

References Aron, A., Norman, C. C., Aron, E. N., McKenna, C., & Heyman, R. E. (2000). Couples shared participation in novel and arousing activities and experienced relationship quality. Journal of Personality and Social Psychology, 78, 273–284. Bellezza, F. S., Greenwald, A. G., & Banaji, M. R. (1986). Words high and low in pleasantness as rated by male and female college students. Behavior Research Methods, Instruments & Computers, 18, 299–303. Danner, D. D., Snowden, D. A., & Friesen, W. V. (2001). Positive emotions in early life and longevity: Findings from the nun study. Journal of Personality and Social Psychology, 80, 804–813. Derogatis, L. R. (1983). SCL-90-R: Administration, scoring, & procedures manual for the revised version. Baltimore: Clinical Psychometric Research. Diener, E., & Diener, C. (1996). Most people are happy. Psychological Science, 7, 181–185.

158

G. P. Strauss and D. N. Allen

Dreisbach, G., & Goschke, T. (2004). How positive affect modulates cognitive control: Reduced perseveration at the cost of increased distractibility. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30, 343–353. Fredrickson, B. L. (1998). What good are positive emotions? Review of General Psychology, 2, 300–319. Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions. American Psychologist, 56, 218–226. Fredrickson, B. L., & Branigan, C. (2005). Positive emotions broaden the scope of attention and thought–action repertoires. Cognition & Emotion, 19, 313–332. Fredrickson, B. L., & Joiner, T. (2002). Positive emotions trigger upward spirals toward emotional well-being. Psychological Science, 13, 172–175. Fredrickson, B. L., & Losada, M. F. (2005). Positive affect and the complex dynamics of human flourishing. American Psychologist, 60, 678–686. Fredrickson, B. L., Tugade, M. M., Waugh, C. E., & Larkin, G. (2003). What good are positive emotions in crises? A prospective study of resilience and emotions following the terrorist attacks on the United States on September 11th, 2001. Journal of Personality and Social Psychology, 84, 365–376. 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. Gottman, J. M. (1994). What predicts divorce? The relationship between marital processes and marital outcomes. Hillsdale, NJ: Erlbaum. Greene, T. R., & Noice, H. (1988). Influence of positive affect upon creative thinking and problem solving in children. Psychological Reports, 63, 895–898. Hirt, E. R., Melton, R. J., McDonald, H. E., & Harackiewicz, J. M. (1996). Processing goals, task interest, and the mood– performance relationship: A mediational analysis. Journal of Personality & Social Psychology, 71, 245–261. Isen, A. M., Daubman, K. A., & Nowicki, G. P. (1987). Positive affect facilitates creative problem solving. Journal of Personality & Social Psychology, 52, 1122–1131. Isen, A. M., Shalker, T. E., Clark, M., & Karp, L. (1978). Affect, accessibility of material in memory, and behavior: A cognitive loop? Journal of Personality & Social Psychology, 36, 1–12. Johnson, K. J., & Fredrickson, B. L. (2005). Positive emotions eliminate the own-race bias in face recognition. Psychological Science, 16, 875–881. Keyes, C. L. M. (2005). Mental illness and/or mental health? Investigating axioms of the complete state model of health. Journal of Consulting and Clinical Psychology, 73, 539–548. Keyes, C. L. M., & Lopez, S. J. (2002). Toward a science of mental health: Positive directions in diagnosis and interventions. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 45–59). London: Oxford University Press. Kucera, H., & Francis, W. (1967). Computational analysis of present-day American English. Providence, RI: Brown University Press. Losada, M. (1999). The complex dynamics of high performance teams. Mathematical and Computer Modelling, 30, 179–192. Lucas, R. E., Diener, E., & Larsen, R. J. (2003). Measuring positive emotions. In S. J. Lopez & C. R. Snyder (Eds.), Positive psychological assessment: A handbook of models and measures (pp. 201–218). Washington, DC: American Psychological Association. MacLeod, C., Matthews, A., & Tata, P. (1986). Attentional bias in emotional disorders. Journal of Abnormal Psychology, 95, 15–20.

Martin, M., Williams, R. M., & Clark, D. M. (1991). Does anxiety lead to selective processing of threat-related information? Behaviour Research & Therapy, 29, 147–160. McKenna, F. P., & Sharma, D. (1995). Intrusive cognitions: An investigation of the emotional Stroop task. Journal of Experimental Psychology: Learning, Memory, & Cognition, 21, 1595–1607. McKenna, F. P., & Sharma, D. (2004). Reversing the emotional Stroop effect reveals that it is not what it seems: The role of fast and slow components. Journal of Experimental Psychology: Learning, Memory, & Cognition, 30, 382–392. Myers, L. B., & McKenna, F. P. (1996). The colour naming of socially threatening words. Personality & Individual Differences, 20, 801–803. Pratto, F., & John, O. P. (1991). Automatic vigilance: The attention-grabbing power of negative social information. Journal of Personality & Social Psychology, 61, 380–391. Rozanski, A., & Kubzansky, L. D. (2005). Psychologic functioning and physical health: A paradigm of flexibility. Psychosomatic Medicine, 67, 47–53. Sabini, J., & Silver, M. (2005). Why emotion names and experiences don’t neatly pair. Psychological Inquiry, 16, 1–10. Schwartz, R. M., Reynolds III, C. F., Thase, M. E., Frank, E., Fasiczka, A. L., & Haaga, D. A. F. (2002). Optimal and normal affect balance in psychotherapy of major depression: Evaluation of the balanced states of mind model. Behavioural and Cognitive Psychotherapy, 30, 439–450. Sharma, D., & McKenna, F. P. (2001). The role of time pressure on the emotional Stroop task. British Journal of Psychology, 92, 471–481. Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning & Memory, 6, 174–215. Strauss, G. P., & Allen, D. N. [Correlations between single- and multi-item emotional experience measures]. Unpublished raw data. Strauss, G. P., & Allen, D. N. (Accepted pending revisions). Emotional intensity and categorization ratings for emotional and nonemotional words. Cognition & Emotion. Strauss, G. P., Allen, D. N., Jorgensen, M. L., & Cramer, S. L. (2005). Test–retest reliability of standard and emotional Stroop tasks: An investigation of color–word and picture–word versions. Assessment, 12, 330–337. Teasdale, J. D., & Fogarty, S. J. (1979). Differential effects of induced mood on retrieval of pleasant and unpleasant events from episodic memory. Journal of Abnormal Psychology, 88, 248–257. Tugade, M. M., & Fredrickson, B. L. (2004). Resilient individuals use positive emotions to bounce back from negative emotional experiences. Journal of Personality and Social Psychology, 86, 320–333. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality & Social Psychology, 54, 1063–1070. Watts, F. N., McKenna, F. P., Sharrock, R., & Trezise, L. (1986). Colour naming of phobia-related words. British Journal of Psychology, 77, 97–108. White, M. (1996). Automatic affective appraisal of words. Cognition & Emotion, 10, 199–211. Williams, J. M. G., Mathews, A., & MacLeod, C. (1996). The emotional Stroop task and psychopathology. Psychological Bulletin, 120, 3–24. Yee, P. L., & Hunt, E. (1991). Individual differences in Stroop dilution: Tests of the attention-capture hypothesis. Journal of Experimental Psychology: Human Perception and Performance, 17, 715–725.

Positive words and automaticity

159

Appendix Emotional words used in E-Stroop task and descriptive statistics (mean (SD)) for emotional intensity, categorization, word frequency, and word length. Emotion category Anger

Intensity Rating Categorization (percentage) Word frequency Word length

Anxiety

Sadness

Neutral

Happiness high intensity

Happiness low intensity

Angry Enemy Hatred Mad Rage Stern

Anxious Nervous Restless Tense Uneasy Urgent

Cry Gloom Grief Hopeless Sad Tragic

Boat Closet Fork Grass Lawn Saxophone

Glory Honor Joy Lively Love Smile

Angel Blossom Diploma Easter Freedom Ocean

Peace Rainbow Sunrise Sunset Triumph Warmth

05.80 (00.70) 87.00 (08.70)

05.10 (00.70) 84.00 (05.00)

05.80 (00.50) 84.00 (11.00)

01.90 (00.30) 89.00 (07.60)

05.85 (00.49) 88.00 (08.20)

04.58 (00.73) 85.00 (07.50)

43.00 (41.00) 04.67 (01.03)

20.00 (07.00) 06.50 (01.04)

28.00 (20.00) 05.00 (01.90)

47.00 (45.00) 04.60 (00.89)

60.00 (45.00) 04.67 (01.03)

46.00 (62.00) 06.40 (00.86)