Back to Journal Cover Page


Volume 7, Number 6
Submitted: May 29, 2001
Resubmitted: December 12, 2001
Accepted: December 13, 2001
Publication date: January 8, 2002


Lynn E. McCutcheon
DeVry Institute - Orlando Campus


There is evidence that love styles can be regarded as personality traits. If so, we might expect that people who value a particular love style with a lover might value that same style in non-partnering, parasocial relationships. A sample of 107 college students filled out the Love Attitudes Scale (LAS) and the Celebrity Attitude Scale (CAS). There were few significant correlations. Highly significant correlations were obtained between the CAS and the game-playing and dependent love styles for a sub-sample of 39 students whose favorite celebrities were likely to be fantasy love attractions. Game-playing and dependency may be fairly stable personality traits that influence one’s choice of a favorite type of celebrity. Gender differences and possible negative implications of celebrity adulation for game-players were discussed.



The topic of dyadic love, as experienced by adult couples, has become popular among social psychologists in the last few decades (e.g. Rubin, 1973 1984). The popularity is due in no small part to the development of the Love Attitudes Scale (LAS) by Hendrick and Hendrick (1986). The LAS was designed to measure attitudes toward love on six different dimensions: passionate, game-playing, friendship, practical, dependent, and selfless. A passionate love style is marked by strong physical attraction and intense emotion. Game-playing lovers lack emotional commitment and may have simultaneous relationships with several partners. Friendship love develops slowly and lacks strong emotion, but is based on affection. A practical lover looks for someone who can meet the criteria one is searching for in a partner. Dependent love is often possessive and jealous, based on the fear that one's lover will leave. Selfless lovers are altruistic, willing to give more than they receive in return. Lee, whose theorizing was the foundation on which the LAS was constructed, has referred to adult dyadic relationships as "partnering" (1973). With the possible exception of game-playing, each love style assumes a long-term relationship with one partner marked by varying degrees of interaction and emotional intensity.

Do love styles measure enduring personality traits or are they more transient? Hendrick and Hendrick speculated that the LAS might be measuring both traits and more transient attitudes (1986). Portions of their original data supported an attitude conception of love styles. For example, students who were currently in love scored "more" on several styles than those students who reported not being currently in love. However, they also found that students who endorsed a passionate love style had higher self-esteem than students who endorsed the dependent love style (1986). Since self-esteem is generally regarded as a personality trait, this suggests a degree of stability in the love styles. They also reported test-retest correlations ranging from .60 for passionate to .82 for game-playing. The interval (4-6 weeks) suggests some stability over time. This evidence is not particularly compelling, because both attitudes and personality characteristics can have stability over time.

Subsequently, Hendrick, Hendrick and Adler (1988) reported that both men and women were moderately successful at predicting their partner's love style scores. To be even moderately successful at this task implies that love styles are fairly stable over the duration of a single partnering relationship.

Woll (1989) correlated the six love styles with 38 personality variables. A conservative interpretation of his results (using the .001 alpha level) resulted in few significant relationships, none higher than .49. For example, friendship love did not correlate significantly with any of the personality variables. On the other hand, dependent love correlated with neuroticism and the needs for aggression and social recognition. Game-playing also correlated positively with aggression, as well as the need for play, disinhibition, and extraversion.


Hendrick and Hendrick (1989) reported that several LAS subscales correlated predictably with Shaver and Hazan’s (1988) attachment styles, which have been shown to be somewhat stable over lengthy periods of time. For example, those participants who had an anxious-ambivalent attachment in their adult relationships were likely to favor a dependent love style (Hendrick and Hendrick 1989).

Frazier and Esterly found that the tendency to favor a game-playing love style declined as the length of a love relationship grew (1990). In similar fashion, others have found that married persons have lower scores on game-playing than single persons (Grote and Frieze 1994; Montgomery and Sorells 1997). This suggests that the game-playing style is either not stable enough to be considered a personality trait, or that those with a game-playing personality tend to have brief relationships with any one person until they get "hooked" on a person who "reels them in" and changes their game-playing attitude. Morrow, Clark and Brock also found that certain love styles were associated with relationship status variables (1995), suggesting that love styles are too transient to be regarded as personality traits.

The evidence does not clearly point in one direction. Perhaps the Hendricks were right when they concluded that "the love styles seem to be both personality traits and malleable attitudes. We need not be forced to choose in an either/or fashion" (1986, p. 401). One way to shed some light on this issue is to examine correlations between partnering and parasocial relationships within a sample of young adults.

A parasocial relationship can be defined as a one-sided one in which person A may know a great deal about person B, but person B may know nothing whatsoever about person A (Horton and Wohl 1956). Parasocial relationships are typical of the one-sided connections between celebrities and most of their fans.

The Celebrity Attitude Scale (CAS) was developed and refined in order to measure attitudes toward one's favorite celebrity (Ashe and McCutcheon 2001; Maltby, Houran, Lange, Ashe, and McCutcheon, in press; Maltby and McCutcheon 2001; McCutcheon, Lange, and Houran, in press). Factor analysis has revealed a three-factor solution that accounts for 46% of the variance. These three factors have been labeled social/entertainment, intense/personal, and mildly pathological. Apparently, many persons who greatly admire their favorite celebrity do so because they find that celebrity both entertaining and a topic for conversation among friends ("My friends and I like to discuss what my favorite celebrity has done"). Some persons feel a strong emotional attachment to their favorite celebrity ("I consider my favorite celebrity to be my soul mate"), while others indicate by their agreement with items like "I often feel compelled to learn the personal habits of my favorite celebrity" that their attitudes may be mildly pathological (Maltby, et al, in press).


If love styles measure personality traits then we might expect that a person who values one love style with a lover might also value that same style in non-partnering relationships. For example, if Susan adopts a selfless, all-giving, non-demanding love style with her partner, will that style typify her approach to other important social relationships? Or, will she apply the selfless love style only to those dyadic relationships that are highly similar to a partnering relationship?

Some parasocial relationships with celebrities are dissimilar to partnering relationships. For example, an 18-year-old male whose favorite celebrity is Michael Jordan probably admires Jordan for his basketball prowess and perhaps for his success as a businessman. On the other hand, the 18-year-old male whose favorite celebrity is a beautiful, young, scantily clad, female singer probably has an idealized, fantasized romantic relationship in mind.

Keeping in mind the diversity of reasons for choosing a favorite celebrity and the evidence which suggests that love styles are personality traits, I hypothesized that high scorers on the passionate and dependent LAS would tend to score high on the intense/personal subscale of the CAS. Emotional intensity appears to be strong in both passionate and dependent lovers (Hendrick and Hendrick 1986; Woll 1989). Based largely on Woll's finding that a dependent love style correlated .49 with a measure of neuroticism (1989), I hypothesized that scores on the dependent scale would correlate positively with the mildly pathological subscale from the CAS.

Furthermore, I hypothesized that the aforementioned relationships would be even stronger for a subset of participants whose choices of a favorite celebrity appeared to have been made with an idealized, fantasy love relationship in mind. My rationale was this: The person whose favorite celebrity is also a fantasy lover may adopt love attitudes toward that celebrity that would mimic their attitudes toward a real lover. In other words, it may be easy to generalize love attitudes previously directed toward a real lover to an idealized parasocial lover. It may be more difficult to generalize those same love attitudes to an equally admired celebrity who is not the object of a fantasy romance.


The participants were 71 men (mean age = 20.52 years, SD = 4.12) and 36 women (mean age = 21.78 years, SD = 5.63). All were first-year students enrolled at a technology-oriented, four-year college in central Florida. Ninety-seven were single, 5 were married, and 5 were divorced. Two additional persons failed to complete the scales.


A subset of 17 men and 22 women were selected for the second correlational analysis based on their choice of favorite celebrities. To qualify for this smaller subset a participant had to choose a favorite celebrity who met each of the following criteria: The favorite celebrity had to be of about the same age and the opposite gender of the participant. Furthermore, the celebrity had to be part of a glamorous profession. In the present study that meant that the celebrity was well-known for either singing, acting, modeling, or being a TV talk show host. The underlying assumption was that for most of the members of the subset their favorite celebrity represented a fantasized romantic parasocial relationship.

The version of the LAS used in the present study was the 4-item per subscale revision recommended by Hendrick, Hendrick and Dicke (1998). In addition to being shorter than the original LAS it has the additional advantage of having "even stronger psychometric properties than the original scale" (1998, p. 147). Strong agreement with any of the 24 items yields a score of five, and strong disagreement yields a score of one. Thus scores on each subscale could range from four to 20. High scores indicate an endorsement of a love style. In three studies alphas for the 4-item revision ranged from .68 (game-playing) to .88 (selfless), with an average alpha of .80 (Hendrick, Hendrick, and Dicke 1998).

The CAS measures the favorability of attitudes toward one's favorite celebrity. The version used here was the 23-item revision employed by Maltby, et al, (in press). Cronbach's alpha for this version was .86. A principal component analysis with oblimin rotation revealed 10 items that loaded on an "entertainment/social" factor, nine items that loaded on an "intense/personal" factor, and four that loaded on a "mildly pathological" factor. Total CAS scores have been found to correlate positively with the favorableness of attitudes toward celebrities in general. Those who believe that celebrities are deserving of their exalted status tended to score higher on the CAS than those who believe that celebrities are merely lucky. Those who believed that media coverage of the Simpson trials and the deaths of JFK, Jr., and Lady Diana was excessive had significantly lower CAS scores (Maltby, et al, in press).

About half of the participants filled out the LAS followed by the CAS as part of their activities in a critical thinking class. The other half filled them out in reverse order to minimize the likelihood of a systematic order effect. Participants were told that all responses were confidential. About three weeks later they were given general feedback about the purpose and the results of the study.



Cronbach's alpha for the CAS total score was .93, and the subscale alphas were .87, .86, and .71, respectively. Corrected item-total correlations ranged from .42 (item 10) to .73 (item 5). Means and standard deviations for males and females are shown in Table 1 below for both the CAS and LAS. None of the differences approached significance, with the lone exception being the significantly higher scores for males on the selfless love attitude scale. Means and standard deviations are similar to those of previous studies in which the CAS was used (Ashe and McCutcheon 2001), especially when age is accounted for (Maltby, et al., in press). Age correlated -.10 (ns) with CAS total. The severely restricted age range may have reduced the strength of this relationship. Previous studies employing samples with less restricted age ranges have found stronger negative relationships between age and strength of one’s interest in celebrities (Ashe and McCutcheon 2001; Maltby, et al., in press; McCutcheon, Lange, and Houran, in press). One’s interest in celebrities generally (as measured on a seven-point scale with "1" being "very weak interest" and "7" being "very strong interest" ) correlated .48 with total CAS scores and .45, .46, and .43 with CAS subscale scores. This finding is also generally consistent with previous research (Ashe and McCutcheon 2001), especially for the mild pathology subscale (Maltby, et al., in press).

Table 1. Means and Standard Deviations on the CAS and LAS for Both Genders.

  Celebrity Attitude Scale
  Female Male
Factor 1 26.72
Factor 2 17.36
Factor 3 8.11
Total 52.19
  Love Attitude Scale
  Female Male
Passionate 15.42
Game-Playing 9.31
Friendship 13.61
Practical 11.97
Dependent 10.69
Selfless 10.69

* p < .0001


Taken as a whole these results add to the growing body of research which suggests that the CAS is both reliable and valid. They also justify combining the data from females and males for subsequent analyses.

Table 2 below presents the Pearson correlations between the six love attitude scores and the Celebrity Attitude Scale and its three factor or subscale scores. The .01 alpha level was used because of the large number of coefficients. High scorers on the dependent LAS did tend to score higher on CAS intense/personal, r (105) = .27, p < .01. Scores on the dependent LAS did not correlate significantly with CAS mild pathological, r (105) = .22, although the relationship was in the predicted direction. There were also significant correlations between: practical LAS scores and CAS social/entertainment, r (105) = .25, p < .01; dependent LAS scores and CAS total, r (105) = .25, p < .01; and game-playing LAS scores and CAS mild pathological, r (105) = .28, p < .01. Taken as a whole, the results shown in Table 2 below suggest that for the most part participants did not generalize their pattern of love styles to their strongest parasocial relationship.

Table 2. Matrix of Correlation Coefficients Linking the CAS to the LAS Based on the Entire Sample.

  Passionate Game-Playing Friendship Practical Dependent Selfless
Social/Entertainment .08 .18 -.07 .25* .21 .10
Intense/Personal -.03 .17 -.04 .18 .27* .01
Mild Pathology -.09 .28* -.17 .09 .22 .02
Total CAS .01 .21 -.08 .21 .25* .06

* p < .01

Table 3 below duplicates the matrix found in Table 2. However, the sample consisted of those 39 participants whose favorite celebrity was both a member of the opposite sex and a glamorous profession, as described earlier. The results are clear. Two of the four correlations involving the dependent love style were significant, as were all four of the correlations in the game-playing column.


Table 3. Matrix of Correlation Coefficients Linking the CAS to the LAS Based on the Subset of Participants Who Selected Glamorous, Opposite-sex Favorite Celebrities

  Passionate Game-Playing Friendship Practical Dependent Selfless
Social/Entertainment .01 .47* -.22 .20 .37 .08
Intense/Personal -.05 .55** -.16 .18 .53** .20
Mild Pathology -.07 .54** -.33 -.02 .26 .10
Total CAS -.03 .55** -.24 .16 .43* .14

* p < .01 one-tailed; ** p < .001 one-tailed.

The moderate positive relationship between scores on the intense/personal subscale of the CAS and the dependent love style, is consistent with the hypothesis that both tap strong emotions. The passionate love style also taps strong emotions, but there was essentially no correlation between scores on that scale and the intense/personal subscale of the CAS. Dependent lovers have been described as "desperate and conflicted about love" (Waller and Shaver 1994 p. 269), less confident about the future of their love relationships than passionate lovers (Woll 1989), and "characterized by a rich and vivid fantasy life" (Woll 1989 p. 481).

Perhaps dependent lovers are driven by their lack of confidence and their desperate need for love to develop fantasized romantic attachments to celebrities as a back-up plan. A strong attachment to a celebrity provides an imaginary security blanket that might be needed at any time, because the breakup of one’s real partnering relationship always seems so close at hand.

Persons who score high on the game-playing love style tend to be playful, sociable, and interested in having multiple relationships (Woll 1989). The highly significant relationships between the game-playing love style scores and the CAS subscale scores seem readily understandable. For the game-player, a romantic attachment to a celebrity is merely a parasocial, fantasized extension of the real world. The physically attractive celebrity is one more conquest to be made, even if that conquest is entirely imaginary.


The present study provides little support for the view that the passionate, practical, and selfless love styles are personality traits, since correlations between scores on these scales and CAS subscales were nonsignificant under both celebrity conditions. Alternatively, these love styles may be applied to other kinds of relationships. For example, the practical lover may also apply a practical approach to the selection of friends, and the selfless lover may also be considerate and generous with friends. On the other hand, persons with dependent and game-playing love styles apparently do apply their styles to other romantic dyads, particularly parasocial ones involving fantasy relationships with celebrities. In other words, the fact that one is a game-player or a dependent person may influence that person to choose a favorite celebrity who serves as a fantasy love partner. As an aside, two gender differences are worth noting. Males scored significantly higher than females on the selfless LAS. To my knowledge this has never happened in previous studies using any of the several versions of the LAS. Since the 4-item version I used was of recent vintage (Hendrick, Hendrick, and Dicke 1998) perhaps this difference bears watching in future studies.

The other gender difference was the choice of favorite celebrity by those whose parasocial attachment was likely to be based on a heterosexual fantasy love relationship. A disproportionately high percentage of the females in my sample (61%) named favorite celebrities of the opposite sex who were in glamorous professions. By comparison, 24% of the male participants made similar kinds of choices. Indeed, gender was significantly related to choice category, phi (1) = .36, p <.001. Both Adams-Price and Greene (1990) and Stever (1995) found that females are more likely than males to be romantically attached to their favorite celebrities when they are in their teens. However, Stever found a reversal of this trend when fans are in their twenties. What the present research shows is that this reversal may not occur until the mid-twenties.

Finally, the highly significant correlation (.54) between game-playing and mild pathology scores for the celebrity-as-romantic-object condition is worth noting. The idea that celebrities are worthy of special treatment has become deeply entrenched in American and British society (Giles 2000; McCutcheon, Lange, and Houran, in press). The popularity of magazines devoted to the adulation of entertainers and athletes is so high that it seemingly overwhelms the few dissenting voices. The result mentioned above provides some evidence that celebrity worship may not always be benign. At least for game-players whose favorite celebrity serves as a fantasy romance there appears to be a strong temptation to engage in behaviors that are either questionable ("I often feel compelled to learn the personal habits of my favorite celebrity") or illegal ("If…my favorite celebrity asked me to do something illegal as a favor, I would probably do it").



Adams-Price, C. and A. L. Greene (1990). "Secondary attachments and adolescent self concept." Sex Roles, 22:187-198.

Ashe, D. D. and L. E. McCutcheon (2001). "Shyness, Loneliness, and Attitude Toward Celebrities." Current Research in Social Psychology, 6(9):124-132.

Frazier, P. and E. Esterly (1990). "Correlates of relationship beliefs: Gender, relationship experience and relationship satisfaction." Journal of Social and Personal Relationships, 7:331-352.

Giles, D. (2000). Illusions of immortality: A psychology of fame and celebrity. London: MacMillan.

Grote, N. K. and I. H. Frieze (1994). "The measurement of friendship-based love in intimate relationships." Personal Relationships, 1:275-300.

Hendrick, C. and S. Hendrick (1986). "A theory and method of love." Journal of Personality and Social Psychology, 50(2):392-402.

Hendrick, C. and S. S. Hendrick (1989). "Research on love: Does it measure up?" Journal of Personality and Social Psychology, 56:784-794.

Hendrick, S. S., C. Hendrick and N. L. Adler (1988). "Romantic relationships: Love, satisfaction, and staying together." Journal of Personality and Social Psychology, 54:980-986.

Hendrick, C., S. S. Hendrick and A. Dicke (1998). "The Love Attitudes Scale: Short form." Journal of Social and Personal Relationships, 15:147-159.

Horton, D. and R. R. Wohl (1956). "Mass communication and para-social interaction." Psychiatry, 19:215-229.

Lee, J. A. (1973). The colors of love: an exploration of the ways of loving. Ontario: New Press.


Maltby, J., J. Houran, R. Lange, D. Ashe and L. E. McCutcheon (in press). "Thou shalt worship no other gods – Unless they are celebrities." Personality and Individual Differences.

Maltby, J. and L. E. McCutcheon (in press). "The relationship between attitudes toward celebrities and authoritarianism." Psychological Reports.

McCutcheon, L. E., R. Lange and J. Houran (in press). "Evidence for non-pathological and pathological dimensions of celebrity worship." British Journal of Psychology.

Montgomery, M. J. and G. T. Sorell (1997). "Differences in love attitudes across family life stages." Family Relations, 46:55-61.

Morrow, G. D., E. Clark, and K. R. Brock (1995). "Individual and partner love styles: Implications for the quality of romantic involvements." Journal of Social and Personal Relationships, 12:363-387.

Rubin, Z. (1973). Liking and loving: An invitation to social psychology. New York: Holt, Rinehart, and Winston.

Rubin, Z. (1984)."Toward a science of relationships." Contemporary Psychology, 29:856-858.

Shaver, P. R. and C. Hazan (1988). "A biased overview of the study of love." Journal of Social and Personal Relationships, 5:473-501.

Stever, G. S. (1995). "Para-social attachments: Motivational antecedents" (Doctoral dissertation, University of Arizona, 1995). Dissertation Abstracts International, 55(7-B):3039.

Waller, N. G. and P. R. Shaver (1994). "The importance of nongenetic influences on romantic love styles: A twin-family study." Psychological Science, 5:268-274.

Woll, S. B. (1989). "Personality and relationship correlates of loving styles." Journal of Research in Personality, 23:480-505.



Correspondence should be sent to Lynn E. McCutcheon, Ed.D., 240 Harbor Dr., Winter Garden, FL 34787. I wish to acknowledge the support I have received from my colleagues, Diane Ashe, Jim Houran, Rense Lange, and John Maltby.


Dr. Lynne E. McCutcheon teaches psychology, critical thinking, and sociology at DeVry Institute of Technology in Orlando, Florida. His research interests are varied, but include such areas as sport psychology, personality, and social psychology. In conjunction with some colleagues he is working on a book about the psychological characteristics of celebrity-worshippers. E-mail: