Do Consumers Seek Emotional Situations: the Need For Emotion Scale

ABSTRACT - Although it has been suggested that individuals differ in their need to seek out emotional stimuli, no construct has been developed to measure this tendency. This research focuses on the development of the concept of Need for Emotion, and describes the construction of a scale designed to capture individual differences on this dimension. The twelve item scale was found to meet reliability and validity tests.

Citation:

Niranjan V. Raman, Prithviraj Chattopadhyay, and Wayne D. Hoyer (1995) ,"Do Consumers Seek Emotional Situations: the Need For Emotion Scale", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 537-542.

Advances in Consumer Research Volume 22, 1995      Pages 537-542

DO CONSUMERS SEEK EMOTIONAL SITUATIONS: THE NEED FOR EMOTION SCALE

Niranjan V. Raman, University of Texas at Austin

Prithviraj Chattopadhyay, University of Texas at Austin

Wayne D. Hoyer, University of Texas at Austin

ABSTRACT -

Although it has been suggested that individuals differ in their need to seek out emotional stimuli, no construct has been developed to measure this tendency. This research focuses on the development of the concept of Need for Emotion, and describes the construction of a scale designed to capture individual differences on this dimension. The twelve item scale was found to meet reliability and validity tests.

The construct of need for cognition (NFC), which captures the "differences among individuals in their tendency to engage in and enjoy thinking" (Cacioppo and Petty 1982), has been well established in communication and persuasion literature. Defined as an individual's tendency to think about and elaborate on events in searching for reality, the 18-item NFC scale (Cacioppo, Petty, and Kao 1984) has been tested and found to meet standards of reliability and validity (Osberg 1987; Sadowski and Gulgoz 1992; Tolentino, Curry and Leak 1990). More importantly, the construct has found useful applications in empirical research dealing with, for example, whether individuals will process persuasive communication along the central or peripheral route to persuasion (Batra and Stayman 1990).

However, cognition represents only one mode of information processing. Individuals may also differ in their tendencies to process affective or emotional stimuli. Similar to "thinkers" who find it "fun to think" (Murphy 1947), it is possible to conceive of "experiencers" or "feelers" who enjoy "experiencing emotion". Thus a "Need for Emotion" (NFE) might be conceived of as a construct that taps into a tendency to seek out and enjoy emotional situations and as a preference to behave in a particular way (Jung 1970).

The importance of such a construct stems from the fact that many situations of interest in the field of consumer behavior may potentially be impacted by an individual's need for emotion, giving rise to a variety of important questions. For example, if certain television viewers are more likely to prefer affective stimuli, would the type of commercial (informative or emotional) have an impact on persuasion? Does product packaging have to be more affective or informational to catch consumers' attention and influence purchase? Are early adopters high or low in terms of preference for emotional stimuli, and what impact does this have on the introduction of new products? These and other questions may be better addressed through an understanding of individual differences in the need for emotion.

The aim of this paper is to introduce a scale which attempts to assess individual differences in the way people deal with emotion in a fashion analogous to the NFC scale. The definition of the NFE construct, and its distinction from existing concepts and constructs in the literature are discussed. This is followed by a description of the development of the scale, and its subsequent testing for reliability and validity. Applications and directions for future research are also suggested.

CONCEPT DEVELOPMENT

Although extensive studies have demonstrated the importance of effects of affect and moods on consumers' memories, evaluations, judgment and behavior (e.g. Edell and Burke 1987; Gardner 1985), most research in consumer behavior has focused on affective components of ads (Aaker and Bruzzone 1985; Mitchell 1986) or affective responses to ads (Holbrook and Batra 1987; Stout and Leckenby 1986). Alternatively, many studies in this area have induced a specific emotion in subjects artificially, and then examined the effects of this affect for all subjects in that condition taken together (Booth-Butterfield and Booth-Butterfield 1990). Differences across individuals regarding a need for seeking out and experiencing emotion have, for the most part, been ignored (with the exception of Allen and Hamsher 1974). This omission is surprising, given the potential for such a construct.

This research centers on the operating hypothesis that individuals vary in the degree that emotion is sought, and furthermore, that this individuality is relevant to the buyer behavior context. The rationale for this stems from two points. First, it has been established that individuals may differ in expressiveness, orientation, and intensity of experience of emotion (e.g. Allen and Hamsher 1974; Booth-Butterfield and Booth-Butterfield 1990). Based on these differences, researchers have speculated that individuals may also differ in their need to seek out emotional stimuli (Harris and Moore 1990; Larsen and Diener 1987). Second, as indicated earlier, many situations in buyer behavior such as information processing, decision-making, and impulse-buying, may be better understood by taking into account individual differences in dealing with emotions and emotional situations.

Need for Emotion

The need for emotion (NFE) is defined as the tendency or propensity for individuals to seek out emotional situations, enjoy emotional stimuli, and exhibit a preference to use emotion in interacting with the world. Emotions have been differentiated from longer affective states like moods as being more intense, attention-getting and associated with specific objects or events (Clark and Isen 1982). Individuals are usually aware of their emotions while longer emotional states such as moods are more general, less intensive, and may operate without an individual's consciousness (Gardner 1985). Since the NFE construct would mainly tap into individuals' tendencies to seek out and enjoy emotional situations which are specific events or objects, we suggest that the NFE taps mainly into short term emotion, rather than longer emotional states. However, in as much as the NFE taps into a stable tendency to seek and enjoy emotion, and process affective information, the construct should help to explain stable patterns of individual behavior. It is important to note that the NFE is conceptualized to be independent from the NFC, so that a person may be high on either, or on both affective and cognitive dimensions.

Previous research has linked cognition to masculinity, and emotions to femininity (Greenblatt, Hasenauer, and Freimuth 1980), thus suggesting an inverse correlation between the modes of processing. However, Bem (1985) suggests that the relationship is not that straightforward. She brings up the idea that although many men and women may differ along these dimensions, some people may be androgynous, or high on both the relevant dimensions. Therefore, on a cautious note, it is suggested that NFE and NFC may be independent to some degree. Nevertheless, since significant gender differences have been found in affective intensity (Flett et al. 1986a), and women have also been found to be more affectively oriented than men (Booth-Butterfield and Booth-Butterfield 1990), women are expected to have higher scores on NFE than men.

It is also important to note that NFE differs from a few constructs in this area of research that have been already examined in previous work. These constructs include affect intensity, emotional styles, and affective orientation. While these constructs are related to NFE, there are clear differences between these constructs and NFE, both theoretically, and in the way that they have been measured.

Affect Intensity

Individuals have been found to differ in the intensity with which they experience affect (Larsen and Diener 1987). Affect intensity can be described as the individual difference in the intensity of emotional response to a given level of affect stimulus (Diener, Larsen, Levine and Emmons 1985). A 40-item Affect Intensity Measure (AIM) was developed by Larsen (1984) and has been useful in separating subjects who respond more strongly to affective stimuli than those who respond less strongly. The measure includes items such as " My happy moods are so strong that I feel like I'm in heaven," as well as reverse-scored items like " Calm and cool could easily describe me." The scale has been tested for reliability and validity, and proven to be acceptable on both counts (Flett et al. 1986b; Larsen 1984; Larsen and Diener 1985).

Larsen, Diener and Emmons (1986) suggest that affect intensity is considered to be a stable individual difference factor and is presumed to be reflected in the responses of certain individuals who modulate the intensity of the stimulus such that they consistently exhibit stronger emotional reactions relative to other individuals. In support, high-intense subjects' responses to actual and hypothetical life events were found to be stronger in affect intensity than low-intense subjects' responses to the same events (Larsen, Diener and Emmons 1986). This difference was independent of positive or negative emotion inducing events.

Larsen and Diener (1987) have suggested that individuals differ in their emotional responses to situations, and attempt to maintain consistency in their emotional responses by engaging in appropriate activities. In particular, individuals high on the affect intensity dimension have been presumed to seek out and prefer emotional stimuli, and enjoy their stimulation. However, Larsen and Diener (1987) do not test this assertion, although work by Moore, Harris and Chen (1994) gives some support to this contention. In any case, the two constructs are different conceptually. AIM measures the intensity with which an individual responds to an affective stimulus while the NFE is designed to tap into an individual's tendency to seek out affective stimuli and enjoy these emotional situations. There is also the possibility that a high AIM person may be a low NFE person and vice versa for the following reasons. Because a high AIM person responds very strongly to emotional stimulus, he or she may seek to avoid emotional situations altogether. Thus, this individual may be a low NFE person. But it is also possible that another individual may seek out, enjoy, and respond strongly to emotional stimuli and hence be a high NFE as well as a high AIM person. Conversely, there may be low AIM individuals who are high NFE persons because they can't experience the affective intensity that they desire in reaction to an emotional stimulus. Thus response to emotions (AIM) and seeking out or enjoying emotions (NFE) may be both positively or negatively correlated and this would need to be empirically tested.

Test of Emotional Styles

Allen and Hamsher (1974) developed and validated a Test of Emotional Styles (TES) that included a dimension of emotional orientation. Orientation refers to "attitudes toward emotional experiences and expressions," where a positive orientation refers to a tendency to seek emotional experiences and find comfort in them while a negative orientation refers to avoidance of such situations. However, while this dimension is conceptually identical to NFE, the TES scale has some serious shortcomings. The orientation dimension of the TES scale has 30 paired items to which responses are dichotomous (i.e., true/false, or desirable/not desirable). [The entire scale is a 75-item forced-choice measure.] With such binary responses, each item can only have minimal variability (Comrey 1988, DeVellis 1991) and the scale would be restricted in its ability to adequately discriminate the degree to which individuals differ on this dimension. Furthermore, Comrey (1988) argues that with such items a phi (point-biserial) coefficient is used, and there is a likelihood of phi coefficients being quite high when there is little variance in the data set. This may happen if just one individual responded "no" to two items to which all the remaining responses were "yes", and may lead to correlational distortions (Comrey 1988). As evidence, an empirical study which used this scale (Woods, Cole and Ferrandez 1977) concluded that there was doubt as to the construct validity of the subscales of TES. Hence there is a need for a different scale which enables stronger reliability and validity (Comrey 1988).

Affective Orientation

Booth-Butterfield and Booth-Butterfield (1990) pointed out that certain individuals have a tendency to use affect as information and rely on their feelings to make decisions. Affective orientation was defined by these authors as the degree to which individuals are conscious of affective cues and use these cues to guide decision-making processes. Affectively oriented persons are considered to be persons who recognize emotional feelings and consider such emotions valuable in their decision making. On the other hand, non-affectively oriented people are those who are more focused on logic and objective facts and for whom affect is superfluous to decisions and has little to contribute to the decision process. The Affective Orientation Scale (AOS) (Booth-Butterfield and Booth-Butterfield 1990, Study 1) is designed to measure this tendency and consists of 20 items such as "I listen to what my "gut" or "heart" says in many situations" and "I try not to let my feelings guide my actions" (reverse scored).

Although the conceptual definition of affective orientation suggests a negative correlation with the NFC construct, the AOS was found to be uncorrelated with NFC (Booth-Butterfield and Booth-Butterfield 1990, Study 2), indicating independence of these constructs. In contrast, NFE is a broader measure which goes beyond the use of affect in decision making and indicates a preference to seek out and enjoy affective situations. Nevertheless since need-for-emotion and affective orientation both deal with individual differences regarding affect and emotion, we expect a strong correlation between AOS and NFE.

Summary of purpose

To summarize, we feel that the Need for Emotion represents a potentially valuable construct which can provide important insights regarding how individuals seek out situations of varying emotional intensity, process information from communications and engage in decision making. The aim of this research is to construct a scale to measure individual NFE, and test it for reliability and validity by relating it to other important constructs. Specifically, we expect a positive correlation between NFE and AOS, but no correlation between NFE and NFC. Finally, we expect women to have higher NFE scores than men.

METHODOLOGY

A pool of 48 items was generated for development of the scale. These items were developed keeping the conceptual definition of the NFE in mind. Previous research into related concepts and their attendant scales (e.g. Booth-Butterfield and Booth-Butterfield 1990; Larsen and Diener 1987), as well as the distinctions between them, served as a guide in generating these items. Thus, these items focus on (1) whether a person seeks to be involved in situations where there is a potential for emotion laden stimuli to be present; (2) whether a person is comfortable with and even enjoys experiencing such situations; (3) whether a person prefers to process emotional information; (4) on preferences and general behaviors across situations; (5) on emotion as a single dimension, rather than on various emotional sub-dimensions. None of the NFE items were imported from previous scales. To control for response bias, 23 of the 48 items were reverse-keyed.

Following this, the measure was administered to a sample of 203 undergraduate students at a large southwestern university. The sample consisted of an almost equal number of males and females. More importantly, in order to generate variation in the NFE of the developmental sample, subjects having different majors including business, advertising, engineering and liberal arts, were included in this sample.

Procedure

Subjects' responses to the NFE, NFC and the AOS were recorded concurrently on a 5-interval Likert scale ranging from Strongly Disagree to Strongly Agree. To control for response bias, all the items from these three scales and a few dummy items such as "I always use coupons" were mixed together. These were then interspersed evenly and care taken that no two items that could be perceived to be similar, appeared consecutively.

Six subjects did not complete a large part of the instrument (fifteen items or more), or did so in a manner that suggested that responses were not true (for example, marking the same response for ten or more items, some of which were reversed). Hence, these were dropped out of the analyses and the final sample consisted of 197 valid responses.

RESULTS

Responses to the 48 items under consideration were factor-analyzed by a principal components analysis in order to determine the number of underlying factors. Thirteen factors with eigenvalues values greater than one were extracted but a scree plot indicated that only the first two factors accounted for most of the variation. Hence, in accordance with applications of factor analysis in scale development (Comrey 1988; DeVellis 1991), the factor analysis was rerun specifying two factors with a varimax rotation.

The first factor (Eigenvalue 12.33) accounted for 25.7 per cent of the total variance while the second factor (Eigenvalue 3.51) accounted for 7.3 per cent of the total variance. A total of 20 items with factor loadings greater than .40 loaded on the first factor. These items such as "I look forward to situations that I know are less emotionally involving" (reverse-keyed), related to individual's desires to look forward to emotional situations. This suggested that the first factor was the need for emotion factor. Ten items with loadings greater than .40 loaded on the second factor, and these items related to generation of emotions. Items that loaded on both factors were dropped for obvious reasons.

Scale reliability

The 20 items that had strong loadings on the first factor were then subjected to reliability analyses. Item-whole correlations and alphas for the scale excluding each item successively were examined. Six items were found to have low item-whole correlations and found to reduce alpha of the overall scale. Hence these items were dropped and the remaining fourteen items were factor analyzed again. All but two items had strong loadings on the first factor. Thus the final scale consists of 12 items. Table 1 displays the need for emotion scale along with the item-whole correlations. The overall alpha for the scale was found to be .87. Unexpectedly, all the 12 items in the scale are reverse-scaled. This finding is discussed in a later section.

Scale validity

The next step of the development of this scale was to examine convergence with related constructs and discriminant validity. As mentioned previously, NFE was expected to have a high correlation with the AOS, and a low or no correlation with the NFC. A negative NFE-NFC correlation would indicate that NFE was simply measuring the opposite of NFC (Booth-Butterfield and Booth-Butterfield 1990).

Results of correlation analysis performed on the three scales are displayed in Table 2. Alphas for the three scales are also reported. As expected, a strong correlation with AOS was found. However, a moderate correlation with NFC was found which suggested that the two constructs were related.

It can be argued that the .69 correlation found between the scales is too high and may suggest conceptual redundancy. However, the two constructs are conceptualized differently and different measures developed accordingly for each. Nevertheless, to examine this issue further, the 20 items of the AOS and the 12 items of the NFE scale were subjected jointly to factor analysis. Since these two constructs are, by definition related, two factors and an oblique rotation were specified. Factor I clearly represented the affective orientation factor, with 14 items from the 20-item AOS having strong loadings on this factor. Factor II or the second factor, was found to be the NFE factor with 11 items from the 12-item NFE scale having strong loadings on this factor. Although five items from the AOS were also found to load on the second factor, the loadings of these AOS items were considerably lower than almost all of the NFE items. Hence, the two scales do appear to measure related constructs that are different from each other.

A t-test was performed to test the gender difference on NFE scores. As hypothesized, female subjects had a higher mean NFE score (Mf=46.62, S.D.=8.97) than male subjects (Mm=43.83, S.D.=8.54), and this difference was found to be statistically significant (t=2.23, p<.05, df=195).

To lend additional support to these findings, we replicated part of our study with a holdout sample (N=212) with similar characteristics as the original sample. Factor analysis and reliability analysis indicated that 10 of the original 12 items of the NFE scale loaded on a single factor. We then subjected these 10 items to a factor analysis with the AOS items to support the idea that they are two distinct measures. An inspection of the eigenvalues and the interpretability of the factors indicated that there were four factors in the solution. Hence the factor analysis was rerun specifying four factors with an oblique rotation. Similar to the initial study, we found that 9 of the 10 NFE items loaded on one factor, along with 1 item of the AOS. The remaining factors were made up of the remaining AOS items. This lends further support to our suggestion that NFE is a unidimensional construct distinct from AOS. The reliability of the scale using the holdout sample data was .84. Correlation with AOS was .67 and with NFC was .31 (p<.01).

TABLE 1

THE NEED FOR EMOTION SCALE

TABLE 2

CORRELATION BETWEEN NFE, NFC AND AOS

DISCUSSION

Although previous research has indicated that individuals differ in their preferences and responses toward emotional stimuli, most information-processing models do not account for this individual variability. A construct analogous to the need-for-cognition was proposed and its importance and relevance to the consumer behavior literature pointed out.

A scale to measure need-for-emotion was developed and this twelve-item scale was found to possess high reliability. Further, a test for convergence with the related concepts of affective orientation and gender, was performed. As hypothesized, the AOS was positively correlated with NFE, and females were found to be higher in need-for-emotion than males. Replication with a holdout sample supported our initial findings.

In terms of discriminant validity, while NFE and NFC were hypothesized to be independent constructs, a moderate correlation between these two constructs was found. We offer the following explanation for this finding. Cognition and affective systems are considered to interact with each other (Arnold 1960; Izard 1977) and not function in isolation. Although, the issue of acquisition of affect has been a controversial one in the literature (e.g. Lazarus 1982; Zajonc 1980; Zajonc and Markus 1982), the strong correlation between NFC and NFE supports the viewpoint that cognition and emotion are indeed related. Besides, both these constructs describe a tendency to "seek out stimuli", albeit different types of stimuli. Therefore, it is possible that both NFC and NFE tap into a higher order construct of need for more intensive processing. Evidently, additional research is required to make stronger inferences.

An unexpected finding of this research was that the entire NFE scale consisted of reverse-keyed items. This leads us to speculate that individuals may actually differ in their need to avoid emotional situations rather than differ in their need to seek out emotional situations. An alternative explanation focuses on the wording of the items in the scale. The word "emotion" or "emotional situation" appears frequently in the scale and there is a possibility that subjects may have interpreted these terms to connote negative emotions. This may be another reason why we found differences among individuals in their need to avoid emotional situations rather than differences in their need to actively seek out emotional situations. Evidently, the NFE construct needs to be refined further. More specifically, some items need to be rephrased to suggest emotions without suggesting a negative feeling, and the scale tested again with these modifications.

LIMITATIONS

Individuals' needs or preferences for specific emotions have not been examined in this research. For example, people may vary in their needs to feel various emotions like joy and anger, or positive versus negative affect. Our first attempt was to establish a generalized, more global need for emotion. Future studies should be conducted on specific emotions to examine these issues.

As with any research dealing with emotion, the limitations of attempting to measure emotion through cognitive means (i.e., via a questionnaire) are obvious. However, current alternative measures of emotion (i.e., physiological measures) have their own limitations such as low levels of reliability and validity. The NFE scale also asks subjects for self reports of behaviors or preferences and examines whether certain groups of subjects have consistently higher needs for emotion. This may capture stereotypical images subjects have of themselves, rather than their actual need for emotion. For example, men may be socialized into the idea that they have a lower need for emotion than women and this may show up as consistently lower scores for men on this scale. However, consistent differences obtained in the hypothesized directions on this scale between the various populations will enable a higher degree of confidence in the validity of this scale. Also, to the extent that consumers internalize a stereotypical image and behave accordingly, the predictive validity of NFE will not be affected.

Notwithstanding these limitations, the findings of this study are promising, and suggest that there is a latent construct which we label NFE, that has not been identified in previous research. However, the scale need to be refined, and more rigorous methods such as confirmatory factor analysis should be used, before any definitive conclusions can be drawn regarding this construct. Future work will address these concerns.

REFERENCES

Aaker, David A. and Donald E. Bruzzone (1985), "Causes of Irritation in Advertising," Journal of Marketing, 49 (Spring), 47-57.

Allen, Jon G. and J. Herbert Hamsher (1974), "The Development and Validation of a Test of Emotional Styles," Journal of Consulting and Clinical Psychology, 42 (5), 663-668.

Arnold, Magda B. (1960). Emotion and Personality, Vol. 1. New York: Columbia University Press.

Batra, Rajeev and Douglas M. Stayman (1990), "The Role of Mood in Advertising Effectiveness," Journal of Consumer Research, 17 (September), 203-214.

Bem, Sandra L. (1985), "Androgyny and Gender Schema Theory: A Conceptual and Empirical Integration," in T.B. Sondereggen (Ed.) Nebraska Symposium on Motivation, Psychology and Gender, Lincoln: University of Nebraska Press.

Booth-Butterfield, Melanie and Steve Booth-Butterfield (1990), "Conceptualizing Affect as Information in Communication Production," Human Communication Research, 16 (4), 451-476.

Cacioppo, John T. and Richard E. Petty (1982), "The Need for Cognition," Journal of Personality and Social Psychology, 42 (1), 116-131.

Cacioppo, John T., Richard E. Petty and Chuan Feng Kao (1984), "The Efficient Assessment of Need for Cognition," Journal of Personality Assessment, 48 (3), 306-307.

Clark, Margaret and Alice Isen (1982), "Toward Understanding the Relationship Between Feeling States and Social Behavior," in Cognitive Social Psychology, Eds. Albert Hastorf and Alice Isen, New York: Elsevier/North-Holland, 73-108.

Comrey, Andrew L. (1988), "Factor-Analytic Methods of Scale Development in Personality and Clinical Psychology," Journal of Consulting and Clinical Psychology, 56 (5), 754-761.

DeVellis, Robert F. (1991). Scale Development: Theory and Applications. Newbury Park: Sage Publications.

Diener, Ed, Randy J. Larsen, S. Levine and Robert A. Emmons (1985), "Frequency and Intensity: The Underlying Dimensions of Affect," Journal of Personality and Social Psychology, 48, 1253-1265.

Edell, Julie A. and Marian Chapman Burke (1987), "The Power of Feelings in Understanding Advertising Effects," Journal of Consumer Research, 14 (December), 421-433.

Flett, Gordon L., Paul Boase, Mary P. McAndrews, Kirk R. Blankstein and Patricia Pliner (1986a), "Affect Intensity and Self-Consciousness in College Students," Psychological Reports, 58, 148-150.

Flett, Gordon L., Paul Boase, Mary P. McAndrews, Patricia Pliner and Kirk R. Blankstein (1986b), "Affect Intensity and the Appraisal of Emotion," Journal of Research in Personality, 20, 447-459.

Gardner, Meryl P. (1985), "Mood States and Consumer Behavior: A Critical Review," Journal of Consumer Research, 12 (December), 281-300.

Greenblatt, L., J. Hasenauer and V. Freimuth (1980), "Psychological Sex Type and Androgeny in the Study of Communication Variables: Self-Disclosure and Communication Apprehension," Human Communication Research, 6, 117-129.

Harris, William D. and David J. Moore (1990), "Affect Intensity as an Individual Difference Variable in Consumer Response to Advertising Appeals," in Marvin E. Goldberg, Gerald Gorn and Richard W. Pollay (Eds.) Advances in Consumer Research, 17, 792-797.

Holbrook, Michael B. and Rajeev Batra (1987), "Assessing the Role of Emotions as Mediators of Consumer Responses to Advertising," Journal of Consumer Research, 14, 404-420.

Izard, Carroll E. (1977). Human Emotions. New York: Plenum Press.

Jung, Carl G. (1970). The Development of Personality. Princeton, New Jersey: Princeton University Press.

Larsen, Randy J. (1984), "Theory and Measurement of Affect Intensity as an Individual Difference Characteristic, " Dissertation Abstracts International, 85, 2297B. (University Microfilms No. 84-22112).

Larsen, Randy J. and Ed Diener (1985), " A Multitrait-Multimethod Examination of Affect Structure: Hedonic Level and Emotional Intensity," Personality and Individual Differences, 6, 631-636.

Larsen, Randy J. and Ed Diener (1987), "Affect Intensity as an Individual Difference Characteristic: A Review," Journal of Research in Personality, 21, 1-39.

Larsen, Randy J., Ed Diener and Robert A. Emmons (1986), "Affect Intensity and Reactions to Daily Life Events, " Journal of Personality and Social Psychology, 51, 803-814.

Lazarus, Richard S. (1982), "Thoughts on the relations between emotion and cognition, " American Psychologist, 35, 1019-1024.

Moore, David J., William D. Harris and Hong C. Chen (1994), "Exploring the Role of Individual Differences in Affect Intensity on the Consumer's Response to Advertising Appeals," in Chris T. Allen and Deborah Roedder John (Eds.) Advances in Consumer Research, 21, 181-187.

Mitchell, Andrew A. (1986), "The Effect of Verbal and Visual Components of Advertising on Brand Attitudes and Attitude toward Advertisement," Journal of Consumer Research, 13, 12-24.

Murphy, G. (1947). Personality: A Biosocial Approach to Origins and Structure. New York: Harper.

Osberg, Timothy M. (1987), "The Convergent and Discriminant Validity of the Need for Cognition Scale," Journal of Personality Assessment, 51 (3), 441-450.

Russell, J. A. (1980), " A Circumplex Model of Affect," Journal of Personality and Social Psychology, 39, 1161-1178.

Sadowski, Cyril J. and Sami Gulgoz (1992), "Internal Consistency and Test-Retest Reliability of the Need for Cognition Scale," Perceptual and Motor Skills, 74 (2), 610.

Stout, Patricia A. and John D. Leckenby (1986), "Measuring Emotional Response to Advertising," Journal of Advertising, 15 (4), 35-42.

Tolentino, Eugenia, Lisa Curry and Gary Leak (1990), "Further Validation of the Short Form of the Need for Cognition Scale," Psychological Reports, 66 (1), 321-322.

Woods, Donald J., Sandra Cole and George Ferrandez (1977), "Dream Reports and the Test of Emotional Styles: A Convergent-Discriminant Validity Study," Journal of Clinical Psychology, 33 (4), 1021-1022.

Zajonc, Robert B. (1980), "Feeling and Thinking: Preferences Need no Inferences," American Psychologist, 35, 151-175.

Zajonc Robert B. and Hazel Markus (1982), "Affective and Cognitive Factors in Preferences," Journal of Consumer Research, 9, 123-131.

----------------------------------------