The Personal Fable and Risk-Taking in Early Adolescence Amy Alberts · David Elkind · Stephen Ginsberg Received: 5 July 2006 / Accepted: 7 August 2006 / Published online: 19 December 2006 C Springer Science+Business Media, LLC 2006 Abstract Elkind’s (1967) theory of adolescent egocentrism proposes two distinct, but related, constructs – the imaginary audience and the personal fable. A corollary to the imaginary audience, the personal fable (PF) yields a sense of invulnerability and speciality commonly associated with behavioral risk-taking. When regarded as a developmental phenomenon, risk-taking is thought to be the result of cognitive immaturity. However, few adolescent health programs address the Amy Alberts, M. A., is currently pursuing a doctoral degree in Applied Child Development at the Eliot-Pearson Department of Child Development, Tufts University. Ms. Alberts’ research interests include psychosocial adjustment during the pubertal transition and adolescent-parent relationships. David Elkind, Ph.D. is Professor of Child Development at Tufts University. His research has been in the areas of social and cognitive development, building upon the research and theory of Jean Piaget. His latest book, The Power of Play will be Published by DaCapo Press in 2007. I have served on the Editorial Board of the Journal of Youth and Adolescence for almost two decades. Although Dan and I have not always agreed on adolescent issues it has always been a responsible, mutually respecting disagreement. As an editor, Dan has always been sensitive, thoughtful and supportive, a pleasure to work with and for. Stephen Ginsberg is a senior at Tufts University and is currently applying to graduate programs in clinical psychology. A. Alberts () Department of Child Development, Tufts University Lincoln-Filene Bulding, Rm. 312, Medford, MA 02155, USA e-mail:
[email protected] D. Elkind Department of Child Development, Tufts University, 205 college Ave., Medford, MA 02155, USA S. Ginsberg Tufts University, Medford, MA, USA egocentric dimension of decision making. We believe that a valid and reliable measure of PF would aid assessment of risk-taking potential and inform preventive interventions. The present paper reports the results of a newly constructed measure of PF and its relation to risk-taking behavior. The following three hypotheses were tested using data from an availability sample of 119 middle school students: 1. PF scores will increase with age; 2. males will score higher than females on the invulnerability dimension of PF; and 3. PF and risk-taking will be positively correlated. As predicted, PF scores increased significantly across the age range studied. Of the two PF dimensions, only invulnerability significantly varied across grades. Males reported significantly higher invulnerability scores than females, and PF and risk-taking were positively correlated. Suggestions for the implementation of this new and, arguably, reliable and valid scale are presented. Keywords Personal fable . Risk-taking The concept of adolescent egocentrism (Elkind, 1967) links the cognitive structures peculiar to adolescence and the behaviors characteristic of this developmental period. While the origins of adolescent egocentrism are arguably cognitive, this construct has affective, non-cognitive characteristics including self-consciousness, invulnerability, and speciality (Elkind, 1967, 1978). Indeed, the concept of adolescent egocentrism was introduced, in part at least, in an attempt to tie cognitive structures to affective facets of adolescent personality. Within the Piagetian theory of cognitive development, egocentrism is broadly defined as a lack of differentiation in subject-object relations that takes a unique form and is reflected in a unique set of thoughts and actions at each stage of mental development (Piaget, 1962). The young child, for Springer 72 J Youth Adolescence (2007) 36:71–76 example, fails to distinguish between the name and the thing. At this age, the child refuses to accept the fact that the same object can have different names and that the name can be changed. After the age of six or seven and the attainment of concrete operations, the child is freed from the egocentrism of words and things but falls prey to a new form of egocentrism. At this stage, the child fails to distinguish between a self constructed hypothesis and the facts. When playing a game that requires a change of strategies, the school age child sticks with the initial strategy and tries to make the facts fit the hypothesis. The emergence of formal operations, e.g., the capacity to mentally construct all the possibilities in a system and generate contrary-to-fact propositions (Inhelder and Piaget, 1958), frees the young adolescent from his or her hypothesis/fact egocentrism, but gives rise to a new type of failure of differentiation. Building upon Piaget’s theory, Elkind (1967) has proposed a theory of adolescent egocentrism with two distinct, but related, constructs – the imaginary audience and the personal fable. The imaginary audience is the adolescent’s assumption that his or her preoccupation with personal appearance and behavior is shared by everyone else (Elkind, 1967, 1978). Elkind contends that the presence of an admiring or fault-finding (imaginary) audience helps to account for the heightened self-consciousness characteristic of early adolescence. Elkind and Bowen (1979) constructed an Imaginary Audience Scale (IAS) and administered it to 4th, 6th, 8th, and 12th grade subjects. As expected, 8th grade participants scored significantly higher than did the other age groups. These results were replicated by other studies (Enright, Shukla, and Lapsley, 1980; Gray, 1984; Ryan, 1994). The personal fable is the corollary to the imaginary audience. Thinking of himself or herself as the center of attention, the adolescent comes to believe that it is because he or she is special and unique. “Other people will not realize their ambitions, but not me; other people will grow old and die but not me; other people will get hooked on drugs but not me.” Thanks to this personal fable, the young adolescent believes that his or her feelings and emotions are different, more intense and excruciating, than those of others. Elkind (1967) suggests that the personal fable gives rise to a sense of invulnerability and speciality with a propensity for behavioral risk-taking. While others have constructed measures to assess the personal fable (e.g., Enright, 1980), the present paper is the first attempt by Elkind and students to construct and test out their own measure of the personal fable. As the above research indicates, the theory of adolescent egocentrism predicts a curvilinear increase and decrease in adolescent egocentrism between childhood and middle-tolate adolescence. That is to say, preadolescents and late adolescents are expected to score significantly lower on the dimensions of adolescent egocentrism than those early teenagers just acquiring formal operations. Previous research has provided support for this predicted developmental pattern (Elkind, 1979; Enright, Shukla, and Lapsley, 1980; Green, Morton, Cornell, and Jones, 1986). We believe that a valid and reliable measure of the personal fable would be an invaluable aid to assessing adolescent risk-taking potential and preventive intervention. Adolescents are disproportionately represented in virtually every category of risk-taking behavior (Arnett, 1992). Substantial human and financial resources are devoted each year to developing programs that target adolescent risk behavior (Greene, Krcmar, Walters, Rubin, and Hale, 2000). Risk-taking behavior is typically conceptualized as a learned behavior, a personality characteristic, or a developmental phenomenon. When regarded as a developmental phenomenon, risk-taking is thought to be the result of cognitive immaturity. That is to say, it is assumed that adolescents are not sufficiently able to assess the risks, the costs and benefits, of engaging in risky behavior. While cognitive-social immaturity is a plausible explanation for risk-taking behavior in adolescence, few adolescent health programs take into account the egocentric dimension of decision making (Greene et al., 2000). Understanding the egocentric basis of risk-taking behavior may have important implications for identification and for preventive educational programs. Given the theoretical and intuitive link between the personal fable and engagement in risk behaviors, researchers have constructed their own personal fable measures and sought empirical evidence for their association (e.g., Greene et al., 2000; Hudson and Gray, 1986; Lapsley, 1989; Melton, 1988). Greene et al. (2000) obtained adolescent self-report data to examine the predictive power of egocentrism in adolescent risk-taking behavior. Results indicated that a high personal fable (score) was a key component in the explanation of most risk-taking behavior. Specifically, the invulnerability dimension of the personal fable was significantly associated with patterns of risk-taking behavior. In early studies conducted by Greene and colleagues (1995, 1996), the speciality dimension of personal fable was a significant predictor of adolescents’ attitudes toward risk behavior. The invulnerability dimension in these studies was inversely associated with adolescents’ perceived intentions to avoid risk behaviors. Sex effects are present for both personal fable and risk behaviors (e.g., Greene et al., 1996; Hudson and Gray, 1986). While there is general agreement that males engage in significantly more risk behaviors than females, the effect of sex on the personal fable is inconsistent in the literature. In a study by Hudson and Gray (1986), females scored significantly higher on personal fable measures than their male counterparts. However, Greene, et al. (1996) found that males scored higher on this construct than females. This apparent inconsistency in the literature warrants further attention. Springer J Youth Adolescence (2007) 36:71–76 73 Because the personal fable measures used in the studies reported above are rather long and cumbersome (49 items), we wanted to find a brief measure of the personal fable that would discriminate between age groups and relate to measures of every day risk-taking. The present paper reports the results of a newly constructed measure of the personal fable and its relation to risk-taking behavior. We tested three hypotheses. 1: that scores on the personal fable scale will increase with age over the early adolescent years; 2: that males will score higher on the invulnerability dimension of personal fable than females; and 3: that there will be a positive correlation between personal fable and risk-taking scores. Method Subjects An availability sample of 119 middle school students (66 males, mean age = 13.38 years, SD = 1.05 years; 53 females, mean age = 13.31 years, SD = .96 years) from a New England town was recruited for this study. The sample included sixth graders (n = 34), seventh graders (n = 41), and eighth graders (n = 44). In accord with theory and past research indicating the emergence of adolescent egocentrism around the ages of 11 and 12 (Elkind, 1967; Enright, Lapsley, and Shukla, 1979), students 10 years of age and below were not included in the study. This age range criterion was implemented by asking participants for their age in years and months on the day of testing. Participants’ sex was also obtained through self-report. The school draws from a lower middle class white neighborhood. Scales Personal fable The personal fable component of adolescent egocentrism was measured with the new Personal Fable (PF) scale. The PF scale is a 12-item Likert-type scale, with five anchors per item (i.e., “this is ... never, rarely, sometimes, often, and always ... true for me”). Each item presents a belief statement and participants must indicate using the 5-point scale the degree to which they regard the statement as “true for me.” Scoring for each item ranges from 1 (“this is never true for me”) to 5 (“this is always true for me”). The PF scale is comprised of two subscales, invulnerability and speciality, comprising six items each. Examples of PF scale items include: “I know I get away with a lot of stuff other kids get in trouble for” (invulnerability); “When my parents or friends tell me that they know how I feel, I don’t believe that they really do” (speciality). A total score per subscale is obtained by summing the item scores (1–5) in that subscale. The total score for each subscale, then, can range from 6–30. A composite score for the personal fable construct is obtained by summing the two subscales. Scores for the personal fable construct can range from 12–60. See Appendix for items. Although a relatively new measure, the PF scale has been field tested to investigate sex and age group differences in adolescent egocentrism (e.g., Elkind, Fallon, Maynard, Pisano, Schwartz, and Murray-Cohen, 2005). Based on a sample of 2,390 participants (males = 1179; females = 1211), Elkind et al. (2005) found that the average personal fable score for this population was 33.1. Males reported significantly higher levels of personal fable than females (F (1, 2301) = 56.71, p < .001).="" the="" internal="" reliability="" of="" the="" pf="" scale="" for="" this="" sample="" approached="" unity="" (cronbach="" alpha=".60)." risk-taking="" adolescents’="" orientation="" to="" risk-taking="" was="" assessed="" with="" the="" risk-taking="" (r-t)="" scale.="" the="" r-t="" scale="" is="" composed="" of="" 10-items,="" each="" of="" which="" presents="" the="" participant="" with="" a="" hypothetical="" situation.="" participants="" indicate="" how="" they="" would="" respond="" to="" this="" hypothetical="" situation="" by="" selecting="" one="" of="" three="" multiple="" choice,="" fixed="" responses.="" although="" the="" content="" of="" the="" response="" options="" vary="" depending="" on="" the="" given="" hypothetical="" situation,="" the="" response="" format="" is="" consistent="" in="" that="" each="" item’s="" response="" options="" represent="" similar="" degrees="" of="" orientation="" toward="" risk-taking="" (i.e.,="" would="" take="" risk,="" would="" hesitate="" or="" take="" calculated="" risk,="" would="" refrain="" from="" risk).="" scoring="" for="" each="" item="" ranges="" from="" 1="" (would="" refrain="" from="" risk)="" to="" 3="" (would="" take="" risk).="" a="" composite="" score="" for="" the="" risk-taking="" construct,="" then,="" can="" range="" from="" 10–30.="" the="" r-t="" scale="" was="" created="" to="" reflect="" common="" risk="" behaviors="" associated="" with="" developmental="" domains="" including="" social="" (e.g.,="" with="" regard="" to="" authority="" and="" peers)="" and="" physical,="" and="" relevant="" issues="" including="" novelty="" and="" social="" conformity.="" the="" following="" is="" an="" example="" of="" a="" social="" risk="" item="" related="" to="" authority:="" “when="" a="" teacher="" says="" something="" that="" i="" know="" is="" wrong,="" i="" ...”="" a)="" let="" them="" know="" that="" they="" are="" wrong,="" b)="" mention="" it="" to="" a="" friend,="" but="" not="" the="" teacher,="" or="" c)="" don’t="" say="" anything.="" an="" example="" of="" a="" physical="" risk="" item="" is:="" “if="" i="" got="" the="" chance="" to="" go="" skydiving,="" i="" ...”="" a)="" would="" definitely="" do="" it,="" b)="" might="" try="" it,="" but="" would="" be="" pretty="" nervous,="" or="" c)="" would="" say,="" “no="" way.”="" the="" following="" item="" is="" meant="" to="" capture="" risk="" associated="" with="" novelty:="" “when="" asked="" to="" play="" a="" game="" i="" have="" never="" played="" before,="" i="" ...="" ”="" a)="" give="" it="" a="" try,="" b)="" watch="" others="" before="" i="" play,="" or="" c)="" choose="" not="" to="" play.="" finally,="" an="" example="" of="" a="" social="" conformity="" item="" is:="" “if="" a="" group="" of="" my="" friends="" are="" trying="" cigarettes="" for="" the="" first="" time,="" i="" ...”="" a)="" join="" right="" in,="" b)="" want="" to="" join="" them,="" but="" decide="" not="" to,="" or="" c)="" don’t="" try="" it.="" the="" r-t="" scale="" was="" developed="" for="" the="" purposes="" of="" the="" present="" study.="" as="" such,="" psychometric="" statistics="" for="" this="" scale="" springer="" 74="" j="" youth="" adolescence="" (2007)="" 36:71–76="" are="" unavailable="" in="" the="" existing="" literature.="" however,="" using="" cronbach’s="" alpha="" of="" internal="" consistency,="" the="" r-t="" scale="" was="" shown="" to="" be="" adequately="" reliable="" for="" the="" current="" sample="" (cronbach="" alpha=".62)." design="" and="" procedures="" participants="" were="" recruited="" by="" one="" of="" the="" authors="" from="" a="" local="" middle="" school.="" six="" classrooms,="" two="" per="" grades="" 6,="" 7,="" and="" 8,="" were="" targeted="" for="" this="" sample.="" once="" permission="" had="" been="" obtained="" from="" the="" school="" district="" superintendent="" and="" principal,="" information="" packets="" were="" sent="" to="" classroom="" teachers="" for="" distribution="" among="" the="" students.="" the="" information="" packet="" contained="" a="" letter="" explaining="" the="" purpose="" of="" the="" study="" and="" a="" parental="" consent="" form.="" students="" were="" asked="" to="" deliver="" the="" information="" packet="" to="" their="" parent(s)="" for="" consideration.="" teachers="" were="" responsible="" for="" collecting="" returned="" materials="" and="" for="" keeping="" track="" of="" which="" children="" had="" parental="" permission="" to="" participate.="" participation="" was="" extremely="" high="" as="" all="" targeted="" students,="" with="" the="" exception="" of="" one,="" were="" given="" permission="" to="" take="" part="" in="" the="" study.="" data="" collection="" was="" conducted="" by="" two="" of="" the="" authors="" throughout="" a="" single="" day.="" students="" were="" excused="" from="" health="" class="" to="" complete="" the="" two="" questionnaires="" (i.e.,="" pf="" scale="" and="" r-t="" scale).="" since="" data="" were="" collected="" from="" two="" classes="" per="" grade,="" a="" total="" of="" six="" “collections”="" took="" place="" over="" the="" course="" of="" a="" school="" day.="" signed="" parental="" consent="" forms="" were="" collected="" and="" child="" assent="" was="" obtained="" prior="" to="" participation.="" verbal="" instructions="" were="" provided="" by="" one="" of="" the="" two="" authors="" in="" advance="" of="" the="" distribution="" and="" completion="" of="" the="" questionnaires.="" all="" participants="" were="" told="" that="" the="" purpose="" of="" the="" study="" was="" to="" examine="" decision-making="" and="" risk-taking="" in="" early="" adolescence="" and="" that="" their="" responses="" would="" be="" treated="" confidentially="" and="" anonymously="" (no="" identifying="" information,="" with="" the="" exception="" of="" sex="" and="" age,="" was="" required="" of="" the="" participants).="" participants="" were="" assured="" that="" there="" were="" “no="" right="" or="" wrong,="" good="" or="" bad”="" answers="" to="" any="" of="" the="" questions="" and="" were="" asked="" to="" respond="" with="" all="" honesty.="" participants="" were="" also="" instructed="" that="" they="" could="" skip="" any="" questions="" that="" they="" did="" not="" wish="" to="" answer.="" each="" of="" the="" six="" classes="" were="" group-administered="" the="" three,="" paper-and-pencil="" instruments.="" as="" a="" control="" for="" possible="" order="" effects,="" the="" sequence="" of="" the="" three="" questionnaires="" followed="" one="" of="" six="" possible="" permutations.="" one="" way="" analyses="" of="" variance="" (anova)="" were="" computed="" to="" examine="" the="" existence="" of="" an="" order="" effect="" on="" participants’="" responses="" to="" the="" pf="" and="" r-t="" scales.="" results="" indicate="" the="" absence="" of="" significant="" differences="" among="" permutations="" for="" both="" pf="" and="" r-t="" scales="" (f(5,="" 113)=".587," n.s.="" and="" f(5,="" 113)="1.179," n.s.,="" respectively),="" so="" data="" were="" combined="" for="" subsequent="" analyses.="" after="" completing="" the="" questionnaires,="" participants="" were="" debriefed="" and="" thanked="" for="" their="" participation.="" the="" entire="" procedure="" took="" between="" 15="" and="" 20="" minutes="" for="" all="" groups="" to="" complete.="" retable="" 1="" mean="" scores="" for="" personal="" fable,="" invulnerability,="" and="" speciality="" at="" grades="" 6,="" 7,="" and="" 8="" mean="" scores="" grade="" 6="" grade="" 7="" grade="" 8="" personal="" fable="" (total="" score)="" 32.12="" 33.34="" 35.75*="" invulnerability="" (subscale="" score)="" 15.88="" 16.64="" 18.30*="" speciality="" (subscale="" score)="" 16.24="" 16.88="" 18.33="" *across-grade="" comparison="" is="" significant="" at="" the="" .05="" level.="" sponses="" to="" the="" questionnaires="" were="" scored="" by="" the="" coauthors.="" all="" analyses="" were="" conducted="" using="" spss="" release="" 13.0.="" results="" scale="" reliability="" preliminary="" analyses="" were="" concerned="" with="" obtaining="" reliability="" estimates="" of="" the="" personal="" fable="" and="" risk-taking="" scales.="" using="" cronbach’s="" alpha="" of="" internal="" consistency="" both="" scales="" were="" found="" to="" have="" adequate="" internal="" reliability:="" personal="" fable=".60;" risk-taking=".62." the="" average="" personal="" fable="" score="" for="" the="" present="" sample="" was="" comparable="" to="" that="" found="" by="" elkind="" et="" al.="" (2005)="" (m="33.88)." hypothesis="" 1:="" personal="" fable="" scores="" will="" increase="" significantly="" across="" the="" age="" range="" studied.="" the="" results="" supported="" our="" first="" hypothesis.="" the="" mean="" scores="" for="" each="" grade="" level="" were="" 32.12="" (6th="" grade),="" 33.34="" (7th="" grade),="" and="" 35.75="" (8th="" grade)="" as="" shown="" in="" table="" 1,="" and="" the="" differences="" were="" significant="" at="" the="" .05="" level.="" similar="" increases="" were="" found="" for="" both="" the="" invulnerability="" and="" speciality="" subscales,="" however,="" only="" invulnerability="" subscale="" means="" significantly="" varied="" across="" grades,="" as="" shown="" in="" table="" 1.="" hypothesis="" 2:="" males="" will="" score="" higher="" than="" females="" on="" the="" invulnerability="" dimension="" of="" personal="" fable.="" as="" predicted="" there="" were="" also="" significant="" sex="" differences="" in="" performance="" on="" the="" invulnerability="" dimension="" of="" personal="" fable,="" as="" shown="" in="" table="" 2.="" males="" reported="" significantly="" higher="" invulnerability="" scores="" than="" their="" female="" counterpart,="" (f(1,="" 117)="7.284," p="">< .01).="" table="" 2="" mean="" scores="" for="" males="" and="" females="" of="" the="" invulnerability="" dimension="" of="" the="" personal="" fable="" construct="" mean="" scores="" males="" females="" invulnerability="" (subscale="" score)="" 17.97="" 15.61*="" *sex="" difference="" significant="" at="" p="">< .01.="" springer="" j="" youth="" adolescence="" (2007)="" 36:71–76="" 75="" hypothesis="" 3:="" personal="" fable="" and="" risk-taking="" scores="" will="" share="" a="" positive="" correlation.="" in="" keeping="" with="" hypothesis="" 3,="" we="" found="" a="" significant="" correlation="" between="" scores="" on="" the="" personal="" fable="" and="" risk-taking="" scales,="" r=".365," p="">< .01. discussion the results of the present study are generally in keeping with the findings of earlier research on the personal fable. based on elkind and bowen’s (1979) finding that imaginary audience scores peak in 8th grade, we expected that this construct’s counterpart, the personal fable, would likewise increase across the age range studied. indeed, personal fable scores significantly increased across grades 6, 7, and 8. consistent with greene et al.’s (1996) findings, males scored significantly higher than females on the invulnerability dimension of personal fable. the significant covariation between personal fable and risk-taking, which has been replicated by several researchers (e.g., greene et al., 2000; hudson and gray, 1986; lapsley, 1989), implicates the relevance of cognitive-social immaturity in addressing the risk behaviors of adolescents. as such, adolescent health programs might benefit from consideration of the egocentric dimension of decision making. the major contribution of the present investigation is the introduction of a short personal fable scale that appears to be both reliable and valid. as such, it may provide a useful instrument for further studies of the personal fable construct with other populations (say at risk groups) and other variables (such as impulsivity and reflectivity). limitations to the study should be noted. the sample was of a limited demographic group—predominantly white, middle class youth—and cannot be generalized to other ethnic and socio economic groups. likewise, as with all cross sectional studies, we cannot say the extent to which either the personal fable or risk-taking is a short-lived developmental phenomenon, or an abiding personality trait. while we hypothesize that the personal fable is a transient developmental phenomenon, for some young people it may well be an abiding personal trait. future research might be designed to discriminate between these transient and abiding fable and risk-taking behaviors. .01.="" discussion="" the="" results="" of="" the="" present="" study="" are="" generally="" in="" keeping="" with="" the="" findings="" of="" earlier="" research="" on="" the="" personal="" fable.="" based="" on="" elkind="" and="" bowen’s="" (1979)="" finding="" that="" imaginary="" audience="" scores="" peak="" in="" 8th="" grade,="" we="" expected="" that="" this="" construct’s="" counterpart,="" the="" personal="" fable,="" would="" likewise="" increase="" across="" the="" age="" range="" studied.="" indeed,="" personal="" fable="" scores="" significantly="" increased="" across="" grades="" 6,="" 7,="" and="" 8.="" consistent="" with="" greene="" et="" al.’s="" (1996)="" findings,="" males="" scored="" significantly="" higher="" than="" females="" on="" the="" invulnerability="" dimension="" of="" personal="" fable.="" the="" significant="" covariation="" between="" personal="" fable="" and="" risk-taking,="" which="" has="" been="" replicated="" by="" several="" researchers="" (e.g.,="" greene="" et="" al.,="" 2000;="" hudson="" and="" gray,="" 1986;="" lapsley,="" 1989),="" implicates="" the="" relevance="" of="" cognitive-social="" immaturity="" in="" addressing="" the="" risk="" behaviors="" of="" adolescents.="" as="" such,="" adolescent="" health="" programs="" might="" benefit="" from="" consideration="" of="" the="" egocentric="" dimension="" of="" decision="" making.="" the="" major="" contribution="" of="" the="" present="" investigation="" is="" the="" introduction="" of="" a="" short="" personal="" fable="" scale="" that="" appears="" to="" be="" both="" reliable="" and="" valid.="" as="" such,="" it="" may="" provide="" a="" useful="" instrument="" for="" further="" studies="" of="" the="" personal="" fable="" construct="" with="" other="" populations="" (say="" at="" risk="" groups)="" and="" other="" variables="" (such="" as="" impulsivity="" and="" reflectivity).="" limitations="" to="" the="" study="" should="" be="" noted.="" the="" sample="" was="" of="" a="" limited="" demographic="" group—predominantly="" white,="" middle="" class="" youth—and="" cannot="" be="" generalized="" to="" other="" ethnic="" and="" socio="" economic="" groups.="" likewise,="" as="" with="" all="" cross="" sectional="" studies,="" we="" cannot="" say="" the="" extent="" to="" which="" either="" the="" personal="" fable="" or="" risk-taking="" is="" a="" short-lived="" developmental="" phenomenon,="" or="" an="" abiding="" personality="" trait.="" while="" we="" hypothesize="" that="" the="" personal="" fable="" is="" a="" transient="" developmental="" phenomenon,="" for="" some="" young="" people="" it="" may="" well="" be="" an="" abiding="" personal="" trait.="" future="" research="" might="" be="" designed="" to="" discriminate="" between="" these="" transient="" and="" abiding="" fable="" and="" risk-taking=""> .01. discussion the results of the present study are generally in keeping with the findings of earlier research on the personal fable. based on elkind and bowen’s (1979) finding that imaginary audience scores peak in 8th grade, we expected that this construct’s counterpart, the personal fable, would likewise increase across the age range studied. indeed, personal fable scores significantly increased across grades 6, 7, and 8. consistent with greene et al.’s (1996) findings, males scored significantly higher than females on the invulnerability dimension of personal fable. the significant covariation between personal fable and risk-taking, which has been replicated by several researchers (e.g., greene et al., 2000; hudson and gray, 1986; lapsley, 1989), implicates the relevance of cognitive-social immaturity in addressing the risk behaviors of adolescents. as such, adolescent health programs might benefit from consideration of the egocentric dimension of decision making. the major contribution of the present investigation is the introduction of a short personal fable scale that appears to be both reliable and valid. as such, it may provide a useful instrument for further studies of the personal fable construct with other populations (say at risk groups) and other variables (such as impulsivity and reflectivity). limitations to the study should be noted. the sample was of a limited demographic group—predominantly white, middle class youth—and cannot be generalized to other ethnic and socio economic groups. likewise, as with all cross sectional studies, we cannot say the extent to which either the personal fable or risk-taking is a short-lived developmental phenomenon, or an abiding personality trait. while we hypothesize that the personal fable is a transient developmental phenomenon, for some young people it may well be an abiding personal trait. future research might be designed to discriminate between these transient and abiding fable and risk-taking behaviors.>