Perceptions of stigmatization, attributional style and adjustment in adolescents with craniofacial conditions.


Material Information

Perceptions of stigmatization, attributional style and adjustment in adolescents with craniofacial conditions.
Physical Description:
vii, 189 leaves : ill. ; 29 cm.
Usitalo, Ann Marie
Publication Date:


Subjects / Keywords:
Research   ( mesh )
Craniofacial Abnormalities -- Adolescent   ( mesh )
Craniofacial Abnormalities -- psychology   ( mesh )
Self Concept   ( mesh )
Department of Clinical and Health Psychology thesis Ph. D   ( mesh )
Dissertations, Academic -- College of Health Professions -- Department of Clinical and Health Psychology -- UF   ( mesh )
dictionary   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 2002.
Bibliography: leaves 169-188.
Statement of Responsibility:
by Ann Marie Usitalo.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002809306
oclc - 51763277
notis - ANT7768
System ID:

This item is only available as the following downloads:

Full Text








I would first and foremost like to thank the adolescents and parents who

participated in this study and who were the inspiration for it. My appreciation also goes

to the staffs of the University of Florida, the Nemours Children's Clinic, and All

Children's Hospital craniofacial clinics who allowed me access to their patients and who

helped me, in numerous ways, to collect the data for this study. Particular thanks are

given to Dr. John Nackashi, Ms. Virginia Dixon-Woods, and Ms. Teri Barcia at the

University of Florida; Dr. Peggy Greco, Ms. Joanne Denicola, and Ms. Kelly McDonell

at Nemours Children's Clinic; and, Dr. Mary Pavan, Dr. Michael Gallant, Dr. Ernesto

Ruas, Ms. Cathy Conley, and Ms. Betty Graves at All Children's Hospital. I would also

like to thank my committee, Drs. James Rodrigue, James Algina, James Shepperd, and

William Williams for their general support and specific contributions. Special thanks go

to my chairperson, Dr. Stephen Boggs, who introduced me to pediatric psychology and

served as a role model and very patient and knowledgeable mentor during my entire

graduate school career.

I also wish to acknowledge the extraordinary amount of help, love, and

encouragement I have received from family, friends, and teachers throughout this long

but ultimately rewarding process. I am especially thankful to my mother, Mabel Usitalo;

my late father, Richard Usitalo; and my sister, Susan Usitalo, for their affectionate

confidence and life-long support. Finally, I am more grateful than I can properly express

to my husband Paul Pitel and my children, Jonathan, Stephen, and Andrew for their love,

endless patience and encouragement, and for the many personal sacrifices they have

made on my behalf.


ACKNOW LEDGMENTS...................................... ...............-.. .. ............... i

ABSTRACT......................... ......... .....


1 INTRODUCTION. ................ .... ............ ..................1

2 LITERATURE REVIEW.............................. .........................8

Appearance......................... ........ .... .. ... .. .......... ....... ........ 8
Stigmatization................ ....... ........... ............ ... ............ .23
Adjustm ent............ .......................... ............ ... ...............30
Self-Protective Strategies and Stigmatization.................... ............ .........43

3 SUMMARY AND OBJECTIVES.................... ....................... 53

4 M ETH OD S............................................... ............................. 59

Participants and Settings....... ... .. ....... ........ ...................... 59
M easures........... ... .............. .... .. ...........................................65
Procedure.... .................................... ... ...... ..... ..............72

5 RESULTS................................. ................ ..................75

Description of Sample Variables ......................................................... 75
Group Comparisons.............................. ............ ..................... 85
Intercorrelations of Variables.................................. ..... .........................87
Tests of the Hypotheses.......................................................94

6 DISCUSSION .................................... .......... ..... ................. 114

Perceptions of Stigmatization........................................... .................. 14
Attributional Style ................... ............... .... .................... .......... 120
Tests of the Hypotheses............................................... .................. 128
Limitations and Strengths of the Study..................................................151
Conclusion........................... .............................. .................153



A SOCIAL PERCEPTION QUESTIONNAIRE........................ ..................157

B PEER INTERACTION REPORT .......................................................159

C DEMOGRAPHIC QUESTIONNAIRE.................. ... ..................163

ITEM RESULTS.....................................................165

ITEM RESULTS........ .................................................167


BIOGRAPHICAL SKETCH.........................................................189

Abstract of Dissertation Presented to the Graduate School of the University of
Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy



Ann Marie Usitalo

May 2002

Chairperson: Stephen Boggs, Ph.D.
Major Department: Clinical and Health Psychology

This study investigated the relationships among demographic and condition

variables, perceptions of stigmatization, attributional style, self-concept, and social and

behavioral adjustment in adolescents with craniofacial conditions (CFC). Of particular

interest was whether attributional style acts to mediate the relationship between perceived

stigma and outcome variables.

Participants were 53 male and 30 female adolescent/parent dyads. Adolescents

were between 13 and 18 years of age with a visible, congenital facial difference (cleft lip,

cleft lip/palate, craniofacial syndrome, or other CFC such as hemangioma) and no

developmental delay or profound physical disability. Independent variables were age,

gender, number of functional impairments (hearing or speech problems, learning

disability), self-rated appearance (Self-Description Questionnaire-II/SDQ-II) and

perceived stigmatization. Outcome variables were self-concept (SDQ-II), social skills as

measured by the Social Skills Rating System (SSRS-Student), extent of peer network


(Peer Interaction Record-PIR), and externalizing and internalizing behavioral problems


Of the adolescents, 83% reported some experience of stigmatization in their social

interactions; parents' and adolescents' perceptions were significantly correlated but

adolescents reported significantly higher levels of stigmatization. Greater satisfaction

with physical appearance and more optimistic attributional style were significantly and

positively correlated with self-concept and social skills, and negatively correlated with

behavioral problems. Higher perceived stigmatization was negatively correlated with

self-concept and social skills and positively correlated with behavioral problems.

Demographic (age, gender, SES) and condition-related variables (diagnosis, number of

functional impairments) were generally not significantly predictive. However, learning

disabilities were associated with fewer social skills, more internalizing problems, and a

more pessimistic attributional style. Results show that a more positive attributional style

mediated the relationship between perceptions of stigmatization and social skills, self-

concept, and internalizing behavior problems; it had no effect on either extent of peer

network or externalizing behavior problems. Both practical and theoretical implications

of these results are addressed, as are directions for further research.


Craniofacial conditions represent one of the most commonly occurring birth

defects. Clefts of the lip and/or palate alone occur in approximately 1.5 out of 1,000

births and the prevalence of all other craniofacial conditions together equals or exceeds

that of cleft lip and/or palate (Eder, 1995). Craniofacial conditions are extremely

heterogeneous in terms of severity of facial deformity, associated features (e.g., problems

with speech and hearing, mental retardation), and complexity of treatment. Children with

craniofacial conditions and their families face an intricate set of challenges in coping

with the rigorous treatment demands and psychological sequelae of having an obvious,

chronic disability.

Immediately after the birth of a child with a craniofacial condition parents must

cope with their reaction to the facial disfigurement, increased care-giving demands (e.g.,

feeding problems), multiple diagnostic procedures, and the reaction of friends and family

to their child's appearance. Treatment may extend from infancy through late adolescence

or young adulthood and frequently involves numerous surgeries and hospitalizations,

extensive dental and orthodontic work, and therapy for speech and hearing impairments.

Although multidisciplinary diagnostic and treatment advances have led to significant

improvements in outcome, a child born with a craniofacial condition may have visible

scars and facial distortions throughout life; a "normal" appearance is often impossible to

achieve. Paradoxically, the advancements in plastic surgery and orthodontic techniques

used to correct not only major disfigurements but also common, minor deviations from an

"ideal" have likely led to a narrower definition of what is an acceptable appearance and

thus accentuated the facial differences of children with craniofacial conditions.

Given the nature of treatment for craniofacial conditions and the importance of

facial appearance in human interaction, it is not surprising that parents and professionals

are concerned about the impact of the condition on the child's or adolescent's

psychological well-being. Psychological issues associated with facial impairment have

been the focus of considerable study. However, the nature of the research, the questions

asked, and the methodologies used have shifted considerably over the years and parallel

the developments in pediatric psychology as a whole.

Much research has focused on comparing children with craniofacial conditions to

normal controls on various psychological measures and is based on the dual assumptions

that significant differences do exist, and that facially disfigured children will perform

more poorly and appear more pathological than would normal children. However, the

overwhelming conclusion from this body of research is that children with craniofacial

conditions, like other children with chronic and/or life threatening illnesses, do not differ

substantially from healthy children in terms of disease-specific personality patterns or

prevalence of severe emotional or adjustment disorders (Eder, 1995; LaGreca, 1990,

1992; Lavigne & Raier-Routman, 1992; Williamson, 1996).

Discrepant results have been obtained on dimensions such as self-concept or self-

esteem. Some studies have found diminished self-concept in children with craniofacial

conditions (Arndt, Travis, Lefebvre, Munro, & Niec, 1986; Broder, Smith & Strauss,

1994; Broder & Strauss, 1989; Kapp, 1979) while others have shown that they view


themselves in a positive light and score similarly to normal controls on most standardized

measures of self-concept or self-esteem (Kapp-Simon, Simon, & Kristovich, 1992; Eder,

1995; Leonard, Brust, Abrahams, & Sielaff, 1991; Williamson, 1996). The disparity in

results has been attributed to methodological problems such as small samples and

inadequate measures, or to differences in the samples or variables assessed (e.g., global

versus specific self-concept). Despite some negative findings, most children with

craniofacial conditions do quite well in the face of difficult life situations, and as Drotar

(1981) noted, the problems that do exist may often be interpreted as normal reactions to

real life stressors.

Chronic illness or disability does, however, function as a stressor that may, in

combination with other variables, contribute to increased risk for the development of

various problems. There is evidence that children and adolescents with craniofacial

conditions may be at higher risk for behavioral problems (Palkes, Marsh, & Talent, 1986;

Pertschuk & Whitaker, 1985; Schneiderman & Auer, 1984; Speltz, Morton, Goodell, &

Clarren, 1993), poorer school achievement (Brantley & Clifford, 1979; Richman, 1976;

Richman, Eliason, & Lindgren, 1988), and difficulties with social adjustment and peer

relations (Birch & Lindsay, 1971; Kapp-Simon, 1986; Pertschuk & Whitaker, 1985;

Pillemer & Cook, 1989; Pruzinsky, 1992). And, within the population of facially

disfigured children and adolescents, considerable variability exists in functioning, with

some individuals adjusting well and others adjusting poorly or inadequately. It is also

recognized that specific areas of functioning such as social adjustment and peer relations,

or developmental stages (e.g., adolescence) may prove especially problematic (Clifford &

Clifford, 1986; Kapp-Simon et. al., 1992; La Greca, 1990, 1992; Rubin & Wilkinson,


Research has thus turned to exploring sources of vulnerability and resistance in

children and their families, and to delineating factors that differentiate those who do well

from those who do not (Drotar, 1997; National Advisory Mental Health Council, 1996;

Wallander, 1993; Wallander & Thompson, 1995; Wallander, Vari, Babani, Banis, &

Wilcox, 1989). Efforts are being made to expand the scope of inquiry to include the

complex network of individual characteristics (Shepard & Magni, 1995; Speltz,

Galbreath, & Greenberg, 1995), family and peer relations, and environmental (e.g., SES)

and medical factors that structure the existence of a child with a craniofacial condition. A

more theoretically based orientation, using general frameworks such as the "risk and

resistance" model proposed by Wallander et al. (1989), and/or specific theories from

clinical, developmental, personality, and social psychology as heuristic devices has been

encouraged (Faier-Routman & Lavigne, 1992). However, most of the research to date

has rested, either implicitly or explicitly, on the assumption that a direct relationship

exists between facial disfigurement and psychological problems. Relatively little

attention has been paid to moderating and mediating factors or to theories that challenge

this basic assumption.

Problems in self-concept and adjustment in children with craniofacial conditions

have traditionally been attributed to their abnormal facial appearance and the stigma

associated with it. Treatment decisions regarding the timing and extent of surgical

corrections are often made based on beliefs about the effects of the child's appearance on

his or her adjustment. Indeed, abundant evidence exists that physical appearance affects

the way individuals are perceived and treated by others (for a review, see Eagly,

Ashmore, Makhijani, & Longo, 1991). Unattractive individuals are viewed more

negatively relative to attractive persons, most notably in terms of social desirability and

people's expectations and judgments about their traits and capacities. In fact, it has been

argued that appearance-based discrimination is severe enough to warrant protection under

handicap law (Harvard Law Review, 1987). But research also shows that unattractive

individuals are credited with many positive qualities, and variables such as familiarity

and similarity of interests may lessen the influence of the physical attractiveness

stereotype (Berscheid & Walster, 1974). These issues have not been explored in relation

to children with craniofacial conditions. Furthermore, as craniofacial disfigurement

typically involves more salient and significant disturbances of normal facial structure

than does ordinary "unattractiveness," at least until corrective surgery is performed, it is

not clear that findings from the literature on attractiveness are generalizable to this


Theories as to the effect of social stigma on psychological adjustment have also

guided research in this area. Individuals with craniofacial conditions are, based on

appearance alone, generally considered to possess a "discredited" condition that is

stigmatized by society and thus seen by others as negative, unfavorable, or in some way

unacceptable (Crocker & Major, 1989; Harvard Law Review, 1987; Goffinan, 1963; Reis

& Hodgins, 1995). Several assertions are fundamental to stigma theory and the role

stigmatization plays in the development of problems among people with facial

disfigurements. First, if the world holds a negative view of a particular attribute, others

will respond unfavorably to the person who possesses that attribute. This social

disapproval is the hallmark of stigma. Second, the effects of such negative social

feedback about the self on the marked person lead to (a) the belief that the attribute

carries a stigma and is a negative characteristic to possess, and (b) diminished self-worth

and self-esteem.

Although traditional stigma theory assumes that social reactions to a discredited

attribute are almost universally negative, some evidence exists that people may react

positively or ambivalently to stigmatizing conditions associated with illness or handicap

as a result of societal pressure to be kind or helpful to those with disabilities (Bennett &

Stanton, 1993; Carver, Glass, & Katz, 1977; Katz, 1981; Weiner, Perry, & Magnusson,

1988). Thus, variations in social feedback might contribute to the development of higher

or lower levels of self-concept in individuals with craniofacial conditions. Unfortunately,

no one has addressed the range of reactions individuals with craniofacial conditions may

encounter. Likewise, although the relationship between perceived stigma and adjustment

has been studied in other chronic conditions (Arston, Droge, Norton & Murray, 1986;

Hermann, Whitman, Wyler, Anton, & Vanderzwagg, 1990; Ireys, Werthamer-Larsson,

Kolodner, & Gross, 1994; Ryan, Kempner, & Emlen, 1980; Westbrook, Bauman, &

Shinnar, 1992), no investigation has empirically examined perceptions of stigmatization

and the impact of such perceptions on the psychological health of facially disfigured


In attempting to explain the ability of stigmatized individuals to maintain a

positive self-concept despite discrimination and diminished opportunities for success,

Crocker and Major (1989) described three self-protective strategies that are "afforded by

membership in a group that is stigmatized ... (a) attributing negative feedback to

prejudice against their group, (b) selectively comparing their outcomes with those of

members of their own group, and (c) selectively devaluing those attributes on which their

group typically fares poorly and valuing those attributes on which their group excels"

(p.612). Despite evidence that these self-protective strategies mitigate the effects of

discrimination and negative social reactions in other stigmatized groups, little is known

about their functioning among the group of individuals with craniofacial conditions.

Assessment of the effects of craniofacial conditions on the psychosocial

adjustment of children and adolescents is extremely difficult. As noted above, a number

of variables may contribute to increased risk or resilience in any individual, and the task

of understanding them and their interactions is daunting. However, it appears that

progress in this area has been hampered by an over-reliance on certain theories and would

benefit from a shift in approach. The research described here attempts to explore the

relationship between perceptions of stigmatization and the self-protective strategy of

attribution on the adjustment of adolescents with craniofacial conditions.


"He had but one eye and the popular prejudice runs in favor of two"
--Charles Dickens, Nicholas Nickleby


Concern about the role of facial disfigurement in the adjustment of children with

craniofacial conditions arises from a recognition of the social significance of facial

appearance and the strength of the physical attractiveness stereotype. Research on

appearance and its impact on everyday life has focused on examining the multitude of

inferential relations that exist between physical attractiveness and personal attributes, as

well as the treatment accorded both attractive and unattractive individuals.

Ratings of Attractiveness

The stability, reliability, and universality of attractiveness preferences is

remarkable. Children and adults from different cultures use comparable criteria in

evaluating facial attractiveness, very young infants can differentiate attractive from

unattractive faces, and an individual's perceived attractiveness remains stable over a

period of years. Research in this area refutes previously held assumptions that standards

of beauty are culture-bound and transmitted to children through a long process of


Infants as young as 3 months of age can discriminate between attractive and

unattractive female faces and, in general, appear to prefer attractive to unattractive faces.

(Langlois et al., 1987). Subsequent research replicated this finding (Samuels,

Butterworth, Roberts, Graupner, & Hole, 1994), and extended it to demonstrate that

infant preferences for attractive faces generalize to adult Caucasian males, adult African-

American females, and other infants (Langlois, Ritter, Roggman, & Vaughn, 1991).

Other studies support an emotional "liking" versus simple visual preference interpretation

of the differential looking time infants showed toward attractive and unattractive faces

(Langlois, Roggman, & Rieser-Danner, 1990 as cited in Langlois, 1995).

Evidence also shows that a wide range of ethnic and cultural groups agree in their

judgments of facial attractiveness. High levels of agreement were found among African-

Americans, Hispanics, and Caucasians rating African-American, Hispanic, and Caucasian

infants and children whether they rated targets from their own or other groups (Kleck,

Richardson, & Ronald, 1974; Stephan & Langlois, 1984). Cross-cultural ratings of

attractiveness were also remarkably consistent, although not as highly reliable as those

obtained from individuals within the same culture (Langlois, 1995; Barden, 1990). The

high level of agreement across ages, ethnic groups, and cultures suggests that certain

structural features may be universally perceived as attractive. Cunningham (1986) found

that this is, indeed, true. He took precise measurements of 24 facial features from

photographs of 50 female faces and found that, although ethnic differences do exist,

African-American, Asian, and Caucasian faces judged to be attractive share many similar

structural features.

Significant agreement exists on ratings of attractiveness, and the reliability of

these ratings within cultures is extremely high. For example, Iliffe (1960) reported a

study in which attractiveness judgments were collected from more than 4,000 male and

female raters from 15 to 55 years old. Reliability coefficients were in the very high .90s


for raters of the same age and sex, and in the high .80s and low .90s for dissimilar raters

(e.g., teenage girls and middle-aged or older men).

Differential Perceptions and Treatment of Attractive and Unattractive Individuals

Substantial evidence from social and developmental psychology make it clear

that facial appearance exerts a marked and pervasive influence on judgments about and

reactions to others. Children and adults prefer attractive over unattractive individuals and

attribute more positive traits, behaviors, and skills to more attractive people (for reviews

see Berscheid & Walster, 1974; Bull & Rumsey, 1988).

Research with adult subjects

In their classic study on the physical attractiveness stereotype, Dion, Berscheid,

and Walster (1972) concluded that "what is beautiful is good." College students rated

facial photographs previously judged as low, medium, or high in physical attractiveness

on 27 different personality traits using a 6-point bipolar adjective scale. They also ranked

them on personality traits by indicating which stimulus person possessed "the most" and

"the least" of a given trait. Finally, the students were asked to decide which stimulus

persons were expected to lead better lives in terms of personal happiness and

occupational success. Physically attractive persons were ascribed more favorable

personality traits and successful life outcomes than were unattractive subjects with one

exception: attractive individuals were not expected to be better parents.

This paradigm has guided many subsequent investigations of the "beauty is good"

stereotype. Students' initial impressions and expectations of attractive and unattractive

professional counselors has been shown to conform to the physical attractiveness

stereotype (Cash, Begley, McGown, & Wise, 1975). College students asked to judge the

level of "psychological disturbance" in their peers from tapes constructed to reflect low

and high levels of maladjustment attributed greater disturbance and poorer prognosis to

unattractive than attractive individuals (Cash, Kehr, Ployson, & Freeman, 1977). And, in

a predictive study of the readjustment of discharged psychiatric patients, attractive

patients stayed outside of the hospital longer, and female patients judged as relatively

more attractive were viewed as better adjusted by a community informant (Farina, Bums,

Austad, Bugglin, & Fischer, 1986). Physical attractiveness was still significantly related

to days out of the hospital and the informant's rating of adjustment when age, education,

diagnosis, frequency of hospitalization, and predischarge adjustment were controlled.

Gross and Crofton (1977) tested and found support for a corollary to the physical

attractiveness stereotype. They asked college students to rate the attractiveness of photos

that were paired with either a favorable, average, or unfavorable personality description.

They concluded that, not only are attractive individuals assumed to have positive

qualities, but that individuals viewed as having positive qualities are judged as more

attractive and thus "what is good is beautiful."

Although the relationship between attractiveness and the attributions of positive

qualities is strong, evidence suggests that some negative attributes are also associated

with attractiveness. In Tanke's (1982) study, male freshman rated attractive female

students higher than unattractive ones on items measuring social and sexual excitement,

but unattractive females were rated higher on items measuring modesty, concern for

others, and interpersonal sensitivity. Both Dermer and Thiel (1975) and Cash and Janda

(1984) explored the negative side of the physical attractiveness stereotype. Their results

indicate that attractive people are perceived as more likely than unattractive people to

possess the negative attributes of vanity and selfishness. And, in a more recent study

(Agnew & Thomason, 1994) subjects were shown pictures of either an attractive or

unattractive man who was fictitiously described as having been diagnosed HIV-positive.

The attractive man was judged to be more to blame for his condition than was the

unattractive man.

Among adults, attractiveness influences one's desirability as a dating partner,

chance of employment, and even whether one will be convicted of a crime. In one study,

Walster, Aronson, Abrahams, and Rottman (1966) found that attractiveness was, by far,

the strongest predictor of liking among 376 randomly matched couples who evaluated

their blind dates at a college dance. It was also the strongest predictor of asking the

partner for another date, regardless of the attractiveness of the person doing the asking.

Intelligence and achievement were not significantly correlated with liking, and

personality variables were only weakly correlated.

Attractive individuals appear to maintain their advantage in social interactions

even when their attractiveness is hidden from their interaction partner (Goldman &

Lewis, 1977). College students engaged in a 5-minute telephone conversation with

opposite-sex partners whom they had never seen, and then rated their partner's social

skills and likeability. Even when their appearance was not visible, attractive individuals

received higher ratings of social skills and likeability. The authors suggested that

differential socialization, with attractive individuals having a history of more positive

attention and thus more opportunities to develop social skills, might account for the


Research suggests that numerous other benefits may accrue to attractive

individuals: attractive defendants were less likely to be judged guilty and given milder

punishments than unattractive defendants in a simulated jury trial (Efran, 1974);

individuals were more likely to help an attractive than a unattractive person by mailing a

"lost" graduate school application with a photograph attached to it (Benson, Karabenick,

& Lerner, 1976) and to provide them with greater quantities of assistance (West &

Brown, 1975). In sum, attractive individuals are both perceived more favorably and are

likely to experience significantly more positive social interactions, at least on initial

encounter, than are unattractive individuals.

The enormous amount of research on the physical attractiveness stereotype led

Eagly et al. (1991) to conduct a meta-analysis of published results. They concluded that,

although subjects ascribed more favorable personality traits and more successful life

outcomes to attractive than unattractive targets, the average magnitude of effect was

moderate and the strength of the effect varied considerably from study to study. Physical

attractiveness had its greatest impact on attributions of social competence, an

intermediate impact on adjustment, potency, and intellectual competence, and little

impact on attributions of integrity and concern for others. Partial support for the

assertion of a "dark side" to physical attractiveness was obtained: attractive people are

perceived as more vain and less modest than unattractive people, but not more egotistical.

Feingold (1992) also analyzed research on the physical attractiveness stereotype

and obtained comparable results. Attractive people of both sexes are perceived as more

sociable, dominant, sexually warm, mentally healthy, socially skilled, and less modest

than unattractive persons, but not as possessing greater character. Neither meta-analysis

examined the effects of attractiveness on perceptions of individuals who are acquainted,

or included studies using children as either rater or target.

Thus the research on the physical attractiveness stereotype shows that attractive

individuals are perceived more favorably and are likely to experience significantly more

positive social interactions, at least upon initial encounter, than are unattractive

individuals. As more meaningful information about others may be difficult for

individuals to discover due to the mobility and anonymity of modem life, the

instantaneous effects of appearance may be greater than ever before. Initial impressions

have been shown to significantly affect the dynamics of subsequent interactions and

encounters (Bull & Rumsey, 1988).

Research with children as subjects

Children' and adolescents' physical attractiveness also affects how they are

perceived by others, the inferences made about their behavior, and their interactions with

other children and adults. Research shows that preschoolers and young elementary

school children hold the same behavioral stereotypes as adults do, and believe that

unattractive children are more likely to engage in troublesome behaviors than attractive

children. Langlois and Stephan (1977) found that 6- and 10-year-old children rated

attractive children, regardless of ethnic origin, as smarter, more likeable, and more

socially skilled. These results, together with those from Kleck et al. (1974) suggest that

physical attractiveness is extremely powerful in determining peer preference ratings and

behavioral expectations.

Most studies deal with initial judgments about others who are not known to the

subject and about whom little is known. However, evidence indicates that attractiveness

may be a powerful predictor of preference over a long period of time and not simply in

the initial stages of interaction. Dion and Berscheid (1974) showed that, even in

preschool, children choose attractive peers as friends over unattractive ones. In this

study, peer nominations on a number of variables were obtained for 77 preschool children

(ages 4-6) rated for attractiveness by adult judges. Attractive children received, in

general, higher popularity ratings and were perceived as more independent than were

unattractive children. Unattractive male children received more nominations for

aggressive behavior, and unattractive children of both sexes received more nominations

for being "scary" than did attractive children. This research suggests that similar

stereotypes of attractiveness exist for acquainted as well as unacquainted children, and

that appearance is associated with specific child character attributes as early as the

preschool years. Low to moderate correlations between physical attractiveness and

popularity were also reported by Kleck et al. (1974).

Vaughn and Langlois (1983) examined the relationship between attractiveness

and peer relation during the preschool years. Two sociometric ratings (popularity and

social competence) and one measure of attention received from classmates were obtained

for 59 children previously rated for attractiveness. The attention each child received was

not associated with attractiveness, but physical attractiveness and popularity were again

significantly related. These findings parallel those obtained from research with adults:

physically attractive individuals are generally preferred, appearance is most significant in

social interactions, and the positive valence associated with attractiveness spreads to

judgments of personal characteristics.

Differential treatment of attractive and unattractive children extends beyond peer

interactions to adult-child interactions. Infants and children who are attractive tend to

receive preferential treatment from adults. Pictures of attractive infants were looked at

significantly longer than were pictures of infants rated as unattractive (Hildebrandt &

Fitzgerald, 1978, 1981, as cited in Barden, 1990). Parke and Sawin (as cited in Tobiasen,

1984) found that normal 2-day-old infants and neonates perceived as attractive were held

and touched more than were unattractive newborns and neonates. Stephan and Langlois

(1984) found strong and consistent differences in behavioral expectations for attractive

and unattractive infants. Attractive infants were rated as more smart, likeable, and good;

less active; and causing fewer problems to their parents. Given these data, it is possible

that differential treatment of attractive and unattractive infants begins almost immediately

after birth.

In an early and influential study, Dion (1972) gave undergraduate women

photographs of attractive and unattractive children and a description of a misbehavior

attributed to the child. The undergraduates rated the misbehavior of unattractive children

more negatively than the same behavior by a more attractive child, and were more likely

to perceive it as being part of the child's personality style. In contrast, misbehavior by

attractive children was seen as an isolated incident. Unattractive children were also rated

as dishonest, unpleasant, and chronically antisocial compared with attractive children.

The preference for attractive children extends to nonsocial as well as social

situations. Clifford and Walster (1973) asked fifth grade teachers to rate unfamiliar

children on the basis of identical report cards, to which a small picture had been attached.

Attractive children were rated as having greater academic ability, having better social

relations and adjustment, and as more likely to become successful in life than unattractive


Physically attractive children are generally favored over unattractive children in

terms of the discipline they receive (Dion, 1972,1974), grades on assignments and report

cards (Felson, 1980; Salvia, Algozzine, & Sheare, 1977), and expectations for academic

achievement and social success (Adams & Cohen, 1976). However, some discrepant

results have been obtained and interactions with gender and race have been noted

(Adams, 1978; Adams & LaVoie, 1974; Dion, 1974; Marwit, Marwit & Walker, 1978;

Rich, 1975). Although these conflicting results may be attributed to chance or variations

in methodology, it is likely that they reflect an interaction among appearance, the

attribute being evaluated, and personal characteristics of the child or observer.

Appearance affects inferences as to social competence significantly more than it affects

inferences about intellectual competence and integrity. The variations among studies

may also reflect the difficulty of reconciling competing stereotypes: the general

association between attractiveness and positive personal and interpersonal attributes, and

the specific association between intelligence and a relatively unattractive, studious


Ratings of Attractiveness and Adjustment

Feingold's (1992) meta-analysis of the correlational research showed that physical

attractiveness was related to personality traits associated with social behavior such as

loneliness, social anxiety, and self-consciousness. It was not, however, related to the

personality traits of sociability, dominance, and mental health, or to the character-related

traits of self-absorption and manipulativeness. For example, one study found a

surprisingly weak relationship between attractiveness and self-esteem among adults:

highly attractive individuals discounted the validity of performance-related praise and

attributed positive evaluations to their attractiveness and not to their internal abilities or

qualities (Major, Carrington, & Camevale, 1984). Another interesting finding was that

self-rated physical attractiveness was more positively correlated to a wider range of

attributes than was other-rated physical attractiveness.

Self-rated attractiveness may be more important than other-rated attractiveness

among children as well. Kenealy, Frude, and Shaw (1989) examined the relationship

between attractiveness and psychosocial adjustment in 1018 children aged 11 to 12.

Attractiveness ratings were obtained from the child's teacher, a panel of judges, an

interviewer, and the child. The child also completed the Piers-Harris Self-Concept Scale

(Piers & Harris, 1984), behavior ratings were obtained from parents and teachers, and a

popularity score was derived from peer nominations in the child's classroom. There was

a strong positive relationship between self-rated attractiveness and self-concept, but a

weak or nonexistent relationship between child's self-esteem and attractiveness ratings by

teachers and judges. As teacher ratings of attractiveness were associated with positive

child attributes (e.g., popularity, confidence, leadership), these results suggest that

appearance as judged by others has an effect on children's behavior and popularity but

that self-esteem may be mostly associated with children's perceptions of their own


Similar results were obtained among a sample of children and adolescents with

spina bifida (Appleton et al., 1994; Lefebvre & Munro, 1986). On the other hand, a study

by Cooper (1993) found that other-rated physical attractiveness had an effect on the self-

esteem of learning disabled children. Cumulative data suggest that appearance and self-

esteem are linked. However, the relationship is complicated and one's view of one's own

appearance is at least, if not more, important than ratings by other people. As one

researcher has noted, "perhaps Beauty lies in the eye of the Beheld as much as in the gaze

of the Beholder" (Lefebvre & Munro, 1986, p.60).

Appearance and Behavior

Approaching the question of differential treatment from another angle, several

researchers questioned whether unattractive individuals actually bchai e di ffercnil, froni

artrjtii\ e indii iduAk ilid thus elicit negative social responses. And, if so, is this because

there is a subtle and unidentified difference in biological or neurological functioning

among people with different levels of attractiveness or is it a reflection of different

socialization experiences, including the misinterpretation of others' behaviors?

Langlois and Downs (1979) observed peer interactions of 3- and 5-year-old

children who varied on ratings of attractiveness. They found few differences in affiliative

behaviors but observed an increase in hyperactivity among unattractive children. Most

interesting were their findings on aggression: no differences were found in 3-year-olds,

but 5-year-old unattractive children displayed more aggression toward their peers than

did attractive children. They hypothesized that age differences in levels of aggression are

attributable to the activation of self-fulfilling prophecies: unattractive children may be

negatively stereotyped and learn over time the behaviors associated with unattractiveness.

Kleck and Strenta (1980) designed two ingenious experiments in which subjects

were randomly assigned to one of three conditions: asthma, epilepsy, and facial

disfigurement. A large "scar" was applied to subjects in the facial disfigurement

condition using theatrical make-up. Under the guise of "conditioning" the scar to make it

more realistic, it was surreptitiously removed. Thus the subjects thought they had a facial

disfigurement when, in fact, there was no change in their appearance. After interacting

with another person, these subjects were asked to comment on those aspects of the other

person's behavior that appeared linked to their appearance. Subjects who thought they

were disfigured (but were not) perceived a strong reaction to their facial appearance in

the other person. The results suggest that an expectancy/perceptual bias mechanism may

influence the social behaviors of facially disfigured individuals. Thus differences in

behavior may result from individuals with craniofacial conditions molding their actions

to fit the negative stereotypes of others and/or changing their behavior to meet their

expectations about how others will act, and thus eliciting the behaviors they expected.

Facial Disfigurement, Behavior, and Social Interaction

An essential question, of course, is whether facially disfigured individuals elicit

the same judgments and social responses as do unattractive individuals. Although

researchers and clinicians have noted that patients complain of discrimination, teasing,

difficulties making friends, and problems in initial encounters with strangers (Bull &

Rumsey, 1988; Lefebvre & Munro, 1986; Pertschuk & Whitaker, 1984), little direct,

empirical evidence exists about the attitudes ofnondisabled children to their peers with

facial disfigurements. Most of the research with children is based on judgments of

drawings and pictures of children with cleft lip and/or palate that are assumed to be

generalizable to children with other craniofacial conditions. As many of the other

conditions involve greater disfigurement, research on children with cleft lip/palate may

underestimate the negative perceptions and interactions experienced by children with

more extensive disfigurement.

In a series of studies examining children's reactions to disabled or handicapped

peers, Richardson and colleagues (Richardson, 1970, 1971, 1983; Richardson & Royce,

1968) presented children the following pictures: (1) child with no disability, (2) child

with crutches and a brace on left leg, (3) child in a wheelchair, (4) child with left hand

missing, (5) child with facial disfigurement, (6) a child with obesity. The children were

then asked, "Who do you like best," and rankings of preference were obtained. Across

age and cultural groups, the rankings were remarkably consistent and clearly supported

Richardson's conclusion that nondisabled children preferred nondisabled peers. The

rankings were also consistent across studies, with the facially disfigured child

consistently ranked next to last, just above the child with obesity.

Recognizing that children respond to other children in a specific social context,

Harper and colleagues (Harper, 1995; Harper, Wacker, & Seaborg Cobb, 1986)

investigated the impact of situational characteristics upon rankings. They altered

Richardson's methodology slightly, presented different pictures, eliminated the picture of

the child without a hand, and placed questions of preference in the context of various

social and play situations (e.g., "Who would you like to go to a movie with?", "Who

would you like to ride a bike with?"). They found that rankings varied by type of

disability pictured, sample tested, and the type of activity specified. For example, a child

in a wheelchair was least likely to be chosen as someone to "ride a bike with." Children

in the United States uniformly demonstrated low preference for cosmetic disability (facial

disfigurement and obesity) versus orthopedic disability. This was not true for children

from different cultures (Middle East, North Africa, Asia). Children in less developed

countries, where greater emphasis is placed on functional survival, showed a low

preference for mobility impairments. In more developed countries, emphasis upon

thinness and appearance may contribute to the low preference for obese and facially

disfigured children.

When the rankings were summed over all conditions, school-aged children were

less likely to choose a child with a facial disfigurement as the preferred friend or

playmate than any other handicapped children. Harper and colleagues (1995) estimate

that less than 1% of the 1,000 children interviewed selected the child with the facial

disfigurement as most preferred. Their assessment of the attributions and reactions to

facial disfigurement what the participating children actually said about the various

conditions was revealing. They found that the children engaged in an "illogical spread"

of negative reactions, assuming that a problem in one area leads to deficiencies in other

areas (e.g., "can't play because of her mouth," "wouldn't know how to play the game").

They also noted that many of the children's comments expressed, quite directly, their

feelings about children with facial disfigurements (e.g., "awkward", "scared", "dirty") and

their perceptions about their appearance (e.g., "cat lips", "looks bad"). Harper interpreted

these results as indicating the magnitude of stigmatization associated with craniofacial

conditions and the potential difficulties in social interactions for children with

craniofacial conditions.

In other studies, ratings of facially disfigured children based upon photographs

have been obtained. Schneiderman and Harding (1984) asked grade school children to

rate color slides of children with normal facial features and those with repaired unilateral

and bilateral clefts of the lip. Photographs of children with clefts were rated more

negatively on several measures including boring, stupid, sad, dirty, mean, and bad.

Tobiasen (1987) used two versions of the same photographs. In one version, the child

was pictured with his/her cleft and in the other version the cleft was corrected. The

subjects' ratings of children with facial deformities were consistently negative, and

pictures of girls with facial deformities were judged more negatively than pictures of


Several researchers studied the effects of facial conditions upon teacher

expectations and employment opportunities. Richman (1978) showed that teachers

underestimated the intellectual ability of the cleft lip/palate children judged to have

severe facial disfigurement. Academic achievement in children with facial clefts and

normal intelligence is often lower than expected on standardized tests and on parents' and

teachers' reports (Brantley & Clifford, 1979; Tobiasen, 1984). Evidence exists for low

academic/intellectual self-concept as well (Broder & Strauss, 1989). Given Richman's

findings, it may be that many children with low academic achievement and self-concept

suffer from the self-fulfilling prophecy of low teacher expectations. Scheuerle, Guilford,

and Garcia (1982) found that employers responded negatively to an individual with a

cleft lip/palate. When a speech disorder was paired with the cleft lip/palate, the

negativity of judgments increased. The accumulated evidence suggests that facial

anomalies are a stigmatized condition: individuals with a craniofacial condition are

viewed more negatively by society and are likely to experience instances of prejudice and




Stigmatized individuals or groups are not defined by their possession of an

attribute that makes them different from others but by the devaluation society places on

the attribute and, by extension, on those who possess it. Numerous writers (Amtson et

al., 1986; Dell, 1986; Goffman, 1963) have emphasized the social nature of stigma that

it can be defined only in terms of a dynamic relationship between an individual and

society. An attribute becomes stigmatized when it interrupts the normal flow of

interaction and

... an individual who might have been received easily in ordinary
social intercourse possesses a trait that can obtrude itself upon attention
and turn those of us whom he meets away from him, breaking the claim
that his other attributes have on us (Goffman, p.5).

Goffman distinguishes between discredited and discredible conditions. A

discredited condition is one with visible cues while a discredible condition presents no

visible cues but, if it is revealed, the individual may be stigmatized. Epilepsy, mental

illness, and being HIV-positive are examples of the latter; facial disfigurement is a classic

example of the former.

Goffman (1963) argues that stigmatization requires the acceptance and

internalization of society's devaluation the stigmatized person must believe that he or

she has a discredible condition and feel some shame about possessing it. It is this last

proposition that implicitly fuels the extensive body of research on self-esteem and self-

concept in children and adolescents with craniofacial conditions, and the expectation that

society's negative views are incorporated into one's sense of self and self-worth.

Psychological, behavioral, and social problems may stem either from diminished self-

concept or self-esteem, or more directly from the accumulation of negative experiences

and limited opportunities.

Many researchers have tried to identify the variables most central to

understanding stigmatization. Hermann et al. (1990) noted that current conceptions have

dichotomized stigma into two types: enacted stigma, or actual instances of

discrimination; and felt stigma, or the perception that others, whether in direct or subtle

ways, devalue the person and/or express a reluctance to interact with them but do not

engage in explicit discriminatory behavior. Jones et al. (1984) outlined six dimensions

they believed most critical for understanding the impact of a given stigma on social

interactions. These are: (1) concealability; (2) pattern of change over time and ultimate

outcome; (3) disruptiveness, or the extent to which the stigma adds to the difficulty of the

interaction; (4) aesthetic factors, or how pleasing or distasteful the stigma is to the senses;

(5) origin of, or responsibility for, the condition; and (6) peril posed to others by the

condition. Stigma theory postulates that the greater the number and/or intensity of these

attributes in any given person or condition, the more potent the stigma associated with

them and the greater the risk for psychosocial difficulties.

Research discussed previously emphasized the negative judgments and

evaluations made of individuals with craniofacial conditions based upon photographs,

drawings, or videos. Additional evidence for stigmatization comes from studies of the

behaviors displayed toward those with facial disfigurements. Goffman (1963) noted that

avoidance, anxiety, uncertainty as to how to interact, and invasion of privacy by staring

and asking intrusive questions characterize encounters between stigmatized individuals

and the general public. In a series of encounters with 450 members of the general public,

Rumsey, Bull, and Gahagan (1982) found that people stood significantly further away

from a confederate in three disfigured conditions (facial birthmark, facial scarring, and

bruising consistent with a traffic accident) than in a nondisfigured condition. Subjects

also chose to stand significantly more often on the confederate's nondisfigured side (the

disfigurement only appeared on one side of the confederate's face). Finally, if facially

disfigured individuals attempted to engage others in brief encounters on the street, many

members of the public tried to avoid them by increasing their pace, averting their gaze,

and attempting to ignore their presence (Rumsey, as cited in Rumsey & Bull, 1986; Bull

& Rumsey, 1986). Other researchers have noted that members of the general public stare

quite openly at disfigured people (Macgregor, as cited in Bull & Rumsey, 1988).

Stigmatization in Epilepsy

There is a paucity of empirical research on the experience of stigmatization and

its perceived impact among individuals with craniofacial conditions. However, the

stigma associated with epilepsy has been investigated and the findings appear relevant to

understanding the experience of facially disfigured individuals. Although the two

conditions differ on one of the attributes (i.e., concealability), they are comparable on

most of the remaining. Individuals with epilepsy and craniofacial conditions are not

(usually) held responsible for their condition nor thought to present any risk to others.

Furthermore, the course over time, disruptiveness, and aesthetic factors vary considerably

within each condition.

Scrambler and Hopkins (as cited in West, 1986) examined the meaning of

epilepsy to people in terms of stigma, stigma strategies, stigmatization, and consequences

for identity. They studied 94 adults with epilepsy and found that only a minority recalled

any instances of stigmatization at all, a pattern contrasting markedly with the pervasive

sense of shame and fear of negative reaction that made up their sense of felt stigma. As a

result, the authors proposed that the lives of people with epilepsy are characterized less

by "enacted stigma" than by "felt stigma," a concomitant sense of shame about their

condition and a predisposition toward negative interpretations of other's behavior.

A more mixed view of the experiences associated with epilepsy emerges from

Schneider and Conrad's (1980) study. While some subjects reported a sense of shame,

stigmatization, and a concomitant desire to conceal their condition, others did not. Many

subjects were willing to selectively reveal they had epilepsy and noted that their

willingness to do so was related to what others, as "stigma coaches" had taught them to

do. In their work, Schneider and Conrad emphasized the role of parents as "stigma

coaches," noting that the greater the parents' sense of shame, the less likely their

offspring were to disclose any information.

Additional evidence that epilepsy is not invariably experienced as stigmatizing

was found in the Ryan et al. (1980) study of 445 individuals with epilepsy. Using a

quantifiable measure of perceived stigma they created, the investigators found

considerable variation in the amount of stigmatization experienced. In addition, the

finding that the relationship between seizure severity and perceptions of stigmatization

due to the disorder was mediated by other individual characteristics lent support to their

proposed sociopsychological model of stigma in epilepsy.

Two other studies empirically examined the role of perceived stigma and

psychological disturbance in individuals with epilepsy. In the first (Arnston et al., 1986),

perceived stigma was significantly and meaningfully correlated with attitudinal variables

(helplessness, self-esteem, and life satisfaction) and measures of psychopathology

(depression, anxiety, and somatic symptoms) in members of epilepsy organizations

throughout the country. Using both quantitative and qualitative measures, they

concluded that their respondent's social-psychological environment was more important

than the physical characteristics of seizure activity. Furthermore, actual seizure

frequency was less important to a person with epilepsy than the perception of how

severely the seizures affect one's life. Possible limitations of the study include a

nonrepresentative sample (all were active in epilepsy groups), the inability to infer

causality from the correlations (does the perceived stigma lead to psychopathology or

vice versa?), and inaccurate reporting of seizure rates.

Hermann et al. (1990) also found a relationship between self-reported feelings of

stigma and psychopathology. Of the seven variables they found predictive of

psychopathology, six were psychosocial in nature (perceived stigma, elevated number of

stressful life events, poor adjustment to epilepsy, financial stress, vocational problems,

and an external locus of control). Earlier onset of epilepsy was the only neurological

variable associated with psychopathology. Number, type, and duration of seizures were

not significant.

Despite observations (West, 1986) that the attitude of parents appears to be

crucially related to the extent of the child's "felt" stigma and the strategies adopted to

manage it, little research exists on families and stigma. West (1986) conducted a

qualitative inquiry of the relationship among parents' experience of "felt" stigma, the

strategies they adopted to manage it, and their effectiveness in helping their child achieve

a "normal" identity. Loosely structured interviews lasting between 2 and 4 hours were

administered to 20 families on at least two occasions over the course of about 1 year.

Due to the nature of the research and the small sample, the author presented only

tentative conclusions. Nonetheless, his findings concurred with those of other studies in

recognizing the importance of felt stigma in adjustment to epilepsy. The majority of

parents (15 of 20 cases) indicated a perception of felt stigma and engaged in various

information management strategies. In general, those who experienced stigma most

strongly were most likely to practice concealment and to maintain high levels of felt

stigma regardless of the success of concealment. Selective disclosure appeared to reduce

perceptions of felt stigma. Furthermore, those parents committed to a policy of

concealment minimized their children's participation in activities outside of the family.

As isolation and restriction of activities are not usually thought of as conducive to normal

development, West noted that the child with epilepsy may experience detrimental effects

as a direct result of the successful use of a strategy chosen for the very purpose of

avoiding negative outcome.

Westbrook et al. (1992) explicitly tested a conceptual model grounded in stigma

theory to investigate the association of stigma to self-esteem among adolescents with

epilepsy. The model hypothesized relationships among several characteristics of epilepsy

(seizure type, seizure frequency, and duration of epilepsy), perceived stigma,

management of disclosure, and self-esteem. Seizure type and frequency predicted low

self-esteem, as did the belief that epilepsy is stigmatizing. Younger subjects (12-16 years

old) had significantly higher stigma scores than did older subjects. However, most of the

subjects did not report feeling stigmatized by their condition. In explaining the negative

findings, the researchers noted that stigma theory does not clearly incorporate personal

subject characteristics such as personality traits, self-perceptions, cognitive evaluations,

or coping styles that may buffer or accentuate the effect of condition-related attributes on

perceived stigma and management of disclosure.

Stigma theory as traditionally formulated postulates strong and direct

relationships between variables: the greater the number and severity of stigma-related

dimensions, the more intense the stigmatization and negative its effects. However, the

accumulated epilepsy and appearance research suggests that the relationships are

complex and mediated by characteristics of the individual such as personality traits and

coping styles, the particular social situation, or the social group within which the

individual is acting. For example, Ireys et al. (1994) assessed how selected condition

characteristics (e.g., indices of severity, symptom predictability, prognosis, age of onset,

and visibility of condition) increased the risk of psychological symptoms in a sample of

286 young adults with chronic illness. They concluded that perceptions of impact

mediate the association between selected risk factors and mental health and should be

seen as a type of cognitive appraisal that shapes the relationship between condition

characteristics and mental health. Similar "self-protective strategies" are presumed to

mediate the association between stigmatization and self-concept in a variety of social

groups or categories of people.

As noted previously, no research has directly examined the subjective experiences

of individuals with craniofacial conditions with regard to perceptions of stigma. As a

result, stigmatization's effect on adjustment has generally been presumed to be negative

and direct with no mediating and few moderating variables (e.g., age and sex) blunting its



Findings from the literature on stigma and physical attractiveness have

contributed to concerns that children and adolescents with craniofacial conditions will

experience emotional and social dysfunction as well as poor self-concept. The prediction

ofpsychosocial difficulties was grounded in the application of social psychological

theories on prejudice, discrimination, and self-esteem to this population. For example,

the theory of "reflected appraisals" posits that children and adolescents with craniofacial

conditions are likely to incorporate society's negative view of individuals with facial

disfigurement into their self-concept. Research on self-fulfilling prophecies suggests that

mistaken beliefs about these children's capabilities would lead to expectations by others

of poor performance and differential behavior toward them. As a result, children with

craniofacial conditions would come to behave and view themselves in a manner

consistent with these negative stereotypes. Fears that discrimination would result in

limited opportunities for developing a sense of competence in many domains and

diminished efficacy based self-esteem added to apprehensions about the social and

psychological impact of craniofacial disfigurement.

Research on the adjustment of children and adolescents with craniofacial

conditions has concentrated on self-concept or self-perception, parent and teacher ratings

of behavior, and social adjustment. Parent and self-ratings of appearance and speech

have been used to assess both levels of satisfaction with treatment outcome and the

relationship, if any, between appearance and adjustment. Most studies have focused on

children and adolescents with cleft lip and/or palate as they comprise the vast majority of

patients with craniofacial conditions. It is not clear whether data gathered from children

with cleft lip and/or palate can be generalized to children with more severe conditions.


Self-concept, or the perceptions one has of oneself, exerts a profound influence on

behavior and much research has focused on understanding its development and

functioning. More complicated than it intuitively appears, self-concept includes both

global feelings of self-worth and judgements about oneself in specific domains such as

academic ability, social skills, and physical appearance (Bracken, 1992; Eder, Gerlach, &

Perlmutter, 1987; Harter, 1986; Marsh, 1986,1993). Thus, one may possess a relatively

high global self-concept and still evaluate oneself negatively on a particular dimension.

Self-esteem is often considered the evaluative component of one's self-concept, and self-

image, which comes from social interactions, is thought of as the internalization of others'

reactions to one's self and behavior. Self-concept, in all its manifestations, plays a critical

role in psychological functioning (Taylor & Brown, 1988) and adjustment, and concerns

about possible disturbances in self-concept among children with craniofacial conditions

have been prominent.

Overall, the research on self-concept has produced few significant differences

between children with craniofacial conditions and typical children, and remarkably little

evidence for diminished global self-concept. Kapp (1979) used the Piers-Harris Self-

Concept Scale (Piers & Harris, 1984) to measure the self-evaluations of 34 cleft lip

and/or palate adolescents (ages 11 to 13) with a matched normative group. No difference

in global self-concept (total scores) was obtained, although on specific subscales, girls

with clefts reported greater anxiousness and general unhappiness and dissatisfaction than

did same-sex controls.

Brantley and Clifford (1979) actually found that adolescents with cleft lip and/or

palate reported a higher level of self-esteem than did obese adolescents or adolescents

with no identified problems. They administered an extensive battery of behavioral

adjustment instruments to three groups of 10 to 18 year-old adolescents. Of 18 body-

image and 15 self-concept variables, significant group differences were found on only 2:

self-esteem and perceived acceptability by parents at birth. Adolescents with clefts

reported lowered perceived acceptability by parents at birth. There were no differences

among the groups on the body-image variables. The authors suggested that the

experience of coping successfully with a cleft and perceptions of unacceptability resulted

in heightened self-esteem in these subjects.

Adolescents with cleft lip and/or palate again scored at or above the norm for

global self-concept in another study (Leonard et al., 1991) using the Piers-Harris

Children's Self-Concept Scale. Further analysis showed an interaction between gender

and age: adolescent girls experienced a more negative self-concept in comparison to

younger girls and adolescent boys experienced higher self-concept compared to younger

boys. These results concurred with Brantley and Clifford's (1979) and Kapp's (1979)

findings of high overall self-concept combined with problems in specific domains or

among particular age-gender combinations.

The self-concept of children (8 to 11 years) was also found to be at or above

normal in the study by Leonard et al. (1991), a result discrepant from other studies. For

example, Kapp-Simon (1986) compared the self-concept of primary school-age children


(5 to 9 years) with clefts to a control group of nonaffected peers using the Primary Self-

Concept Inventory (PSCI) (Muller & Leonetti, 1974). The children with clefts reported a

lower global self-concept and perceived themselves as less socially adept and more

frequently sad and angry than their peers. However, although the differences in self-

concept score were significant, only 54% of the group with clefts scored in the "at risk"

range for global self-concept compared to the 74% classified as "at risk" on the social self


Broder and Strauss (1989) also used the Primary Self-Concept Inventory to

investigate the impact of visible versus invisible defects on children's self-concept. They

divided their sample of 6 to 9-year-old children into three groups -- cleft lip only, cleft lip

and palate, and cleft palate -- and compared their scores to those of peers with no

conditions. All groups with clefts scored significantly below the controls and in the "at

risk" range on the total self-concept score, both factors of the social domain, and the

physical factor of the personal self-domain.

Thus, while there is some inconsistency in results, one pervasive finding is that

global self-concept is marginally, if at all, affected by having a craniofacial condition.

However, specific domains of the self-concept appear vulnerable, and interactions of age

and gender may be significant. Some of the discrepancies can be attributed to the

differences in age groups and measures. The focus on either children or adolescents is a

significant strength as it acknowledges developmental differences in self-concept and

adjustment to facial disfigurement (Damon & Hart, 1982; Speltz,, 1995; Speltz,

Greenberg, Endriga, & Galbreath, 1994).

Psychological and Behavioral Adjustment

Several studies have focused on parent and teacher reports of behavioral

adjustment (Richman, 1976; Schneiderman & Auer, 1984; Tobiasen & Hiebert, 1984). In

an early review of the literature on the intelligence, achievement, and psychological

adjustment of children with clefts of the lip and/or palate, Richman & Eliason (1982)

noted that research suggested that children with clefts were more inhibited than their

peers without clefts but displayed no evidence of significant psychopathology. Richman

(1976) gathered teacher Behavior Problem Checklist (BPC) reports for 44 children with

clefts and 44 matched controls ages 9 to 14. Children with clefts had significantly higher

scores on the internalizing scale of the BPC than controls, but did not differ on the

externalizing scales.

More introversion or internalizing behavior problems has been reported in other

studies as well (Pertschulk & Whitaker,1985; Pillemer & Cook, 1989). In a post-surgical

study of 25 children, Pillemer & Cook found significant problems with introversion and

peer relations. However, these problems emerged on projective measures only. Ratings

on objective measures were within normal limits. Pertschulk & Whitaker studied 43

craniofacial patients aged 6 to 13 years using a combination of standardized measures

and structured interviews. They found the patients to have poorer self-concept, greater

anxiety, and, as rated by parents and teachers, more negative social encounters and

problematic classroom behavior than matched control children. Emphasizing that the

children with craniofacial conditions were not functioning in a psychosocially deviant

range, the authors noted their results indicated limitations rather than marked deficits. No

consistent pattern of psychosocial problems was found within the whole group, and there

were no significant differences among children with different diagnoses. However, a

subset of patients with problematic responses showed a pattern of behavioral inhibition

and limited social contacts and/or frequent negative social encounters.

Schneiderman and Auer (1984) used both parent and teacher reports on the

Behavior Problem Checklist (BPC) to evaluate the adjustment of 58 children with cleft

lip and palate from preschool through grade nine. They found that very young children

of both sexes and girls in junior high school tended to be categorized as having

"personality problems," whereas elementary and junior high school boys were more

likely to be seen as having "conduct problems." Parents perceived more conduct

problems among males than did teachers. However, in Tobiasen and Hiebert's (1984)

study of 2 to 12-year-old children, parents of children with facial clefts reported the same

frequency of conduct problems as did parents of children without clefts.

Problems in family interactions, school productivity, and peer relationships were

also found in six often children (ages 5-15 years) with mixed craniofacial conditions

assessed by Palkes et al. (1986). This study explored the effects of the attitudes of

parents upon the child's well being. Parental attitudes toward their children were

favorable, and most expected their child to have positive self-esteem. Thus, the high

percentage of children with behavioral problems was startling and raised the possibility

that children with conditions other than clefts may differ in their adjustment. However,

the extremely small sample size prohibited any definitive conclusions.

As with self-concept, it is clear that children and adolescents with craniofacial

conditions do not display any single pattern of behavioral difficulties. The absence of

consistent group findings characterizing much of the literature has led some researchers

to investigate differences among children with facial disfigurements. Both

clinical/anecdotal (Pruzinsky, 1992; Tobiasen, 1995) and research (Birch & Lindsay,

1971; Clifford & Clifford, 1986; Pertschuk & Whitaker, 1984; Richman, Holmes, &

Eliason, 1985; Starr & Heiserman, 1977) data show that, within the craniofacial

population, some children and adolescents have poor adjustment, while others adapt quite

well to their condition.

Differences in Adjustment

Starr and Heiserman's research (1977) showed a relationship between acceptance

of disability and self-esteem (Linkowski, 1971) among adolescents with clefts of the lip

and/or palate. Those adolescents who indicated the highest level of acceptance on the

Scale to Measure Acceptance of Disability had higher self-esteem scores on the Self-

Esteem Scale (Simmons, Rosenberg, & Rosenberg, 1973) and scored significantly lower

on aggression, activity level, somatization, and sleep disturbance as measured by the

Missouri Children's Behavior Checklist (Sines, Pauker, & Sines, 1971). There was no

relationship between type of cleft and attitude toward their disability. The authors

concluded that complex psychosocial issues are more important than physical severity in

the acceptance of one's facial disfigurement, and that adolescents with difficulty

accepting their disability are at high risk for disturbances in behavior and self-concept

and may benefit from psychosocial interventions.

Perceptions of educational and social functioning among adolescents with cleft lip

and palate varied in accordance with their personality adjustment as measured by MMPI

scores (Richman, 1983). Significantly more adolescents with abnormal MMPIs felt

dissatisfied with their educational and social functioning than did those with normal


MMPIs. Groups of well and poorly adjusted adolescents with cleft lip and palate, based

on parent Behavior Problem Checklist (Quay & Peterson, 1969) ratings, were also

compared in the Richman et al. 1985 study. They examined the relationships among self-

ratings of facial appearance, teacher ratings of appearance, and parent and child ratings of

adjustment as measured by the BPC. While the ratings of adolescents in the well-adjusted

group agreed with those of parents and teachers, adolescents in the poorly adjusted group

agreed only with parental ratings on the conduct problems dimension of the BPC. They

rated themselves as having significantly fewer internalizing personality problems than

parents, and as significantly more attractive than teachers. The authors hypothesized that

poorly adjusted individuals use denial of facial disfigurement as a defense mechanism

and that their social withdrawal is a consequence of their denial. However, adolescent's

self-ratings on all three measures were not significantly different between the groups, and

those in the poorly adjusted group had more severe facial disfigurement. Thus, although

differences in adjustment exist, the relationship among severe facial disfigurement,

adjustment, introversion, and a greater use of denial remains ambiguous.

Effects of specific condition related variables (appearance, speech, hearing) on

adjustment have been explicitly examined in several studies. As noted above, some

researchers found no significant differences in adjustment among children with various

craniofacial malformations or levels of attractiveness (Kapp-Simon, 1986; Pertschuk &

Whitaker, 1985; Starr, 1980; Starr & Heiserman, 1977). Others have asserted that those

children with major deformities (not just cleft lip and/or palate) or who are less attractive

do more poorly (Pillemer & Cook, 1989; Pruzinsky, 1992; Richman, et al., 1985).

Padwa, Evans, & Pillemar (1991) examined the effect of symmetric and asymmetric

(hemifacial microsomia, facial asymmetry, unilateral coronal orbital synostosis, and

hemihypertrophy) craniofacial deformities and functional impairments (hearing and

speech problems) on measures of depression, social/emotional development, and

behavior. Those with symmetric deformities scored significantly poorer on all measures

of adjustment, but differences among the groups with or without functional impairments

were not significant. Williamson (1996) also found that speech, hearing, and level of

attractiveness were not related to behavior problems or self-esteem. These results were

somewhat surprising given evidence that speech and hearing problems can also have a

negative influence on adjustment (reviewed in Williams, 1996) and assumptions about

the additive effects of having both functional and visible impairments.

Appearance and Adjustment

Although the relationship between severity of facial disfigurement or level of

attractiveness and adjustment is not clear, both patients and professionals remain

concerned about facial appearance (Broder, Smith, &Strauss, 1992; Strauss, Broder,

Helms, 1988; Noar, 1991) and many surgeries are performed for cosmetic reasons alone.

Pre-and postoperative comparisons of facial appearance and adjustment have generally

shown that patients perceive improvement in both areas. In one study (Lefebvre &

Barclay, 1982), a large percentage of patients reported better psychosocial adjustment

after surgery to improve their appearance, suggesting that improved appearance may have

had a beneficial effect. Amdt et al. (1986) found that patients rated their appearance,

self-esteem, and social acceptance and adeptness as noticeably improved after surgery.

As independent raters observed only relatively subtle changes in appearance, the authors

surmised that the patients' improved self-evaluations increased their level of social

comfort and thus enhanced their quality of life. Similar results were reported by

Tobiasen and Hiebert (1993) in their study of children and adolescents with facial clefts:

self-ratings, not peer ratings, of severity of impairment strongly predicted positive

psychosocial adjustment in several areas including global self-esteem, positive mood, and

social competence. When evaluating changes in appearance, it is important to remember

that impressions of individuals are formed by factors other than physical attractiveness

including vocal qualities, gesturing, and other social skills (Canady, 1995). Furthermore,

patients' judgments of outcome appear to depend as much on their expectations as on any

objective criteria.

Social Adjustment and Peer Interactions

Healthy social interactions are essential to good adjustment and satisfaction with

one's life. Several retrospective studies (Birch & Lindsay, 1971; Bjornsson &

Agustsdottir, 1987; Heller et al., 1981) of long-term adjustment in patients treated for

craniofacial conditions noted specific problems in adult social functioning. For example,

Heller, Tidmarsh and Pless (1981) studied 96 young adults (ages 18 to 28) born with

clefts and 56% of all respondents expressed some degree of dissatisfaction with their

social interactions, social activities, and number of friends. As patterns of social

interaction are known to begin early in life and remain relatively stable (Rubin &

Wilkinson, 1995), concerns about social and peer adjustment of children and adolescents

with craniofacial conditions are expressed in almost every paper written. Unfortunately,

one consistent finding is that the probability of problems in social and peer functioning is

quite high, and the internalizing behavioral problems often associated with facial

disfigurements include many that are social in nature: shyness, anxiety, hypersensitivity,


self-consciousness, social inhibition, and negative self-perception (Bull & Rumsey, 1986;

Kapp-Simon et al., 1992: Rubin & Wilkinson, 1995; Tobiasen, 1995; Tobiasen &

Hiebert, 1993).

Problematic social interactions are emblematic of stigmatization. Goffman (1963)

noted that the stigmatized can never be certain about how they will be received and thus

may avoid contact with nonstigmatized individuals. However, a desire for social

interaction may remain, and this desire is what distinguishes social withdrawal or

avoidance and dysfunctional shyness from the personality dimension of introversion

(Cheek & Melchior, 1990). And, as Rubin and Wilkinson (1995) noted, the social

withdrawal and anxiety that may predict peer rejection for craniofacial, handicapped and

other stigmatized children is different from the aggression that is associated with

rejection for unattractive children.

The social experiences of individuals with craniofacial conditions may be

organized into the three components of affect, cognition, and observable behavior.

Affective and cognitive dimensions include the symptoms noted above: anxiety, self-

deprecating thoughts, self-consciousness, and hypersensitivity to evaluation. Direct

observational data (McGuire, Kapp-Simon, & Simon as cited in Kapp-Simon, 1995) and

secondary reports (Rumsey as cited in Bull & Rumsey, 1986) have identified several

behaviors common to individuals with craniofacial conditions. These include infrequent

eye contact, a more monotonous tone of voice, slouched posture, fewer attempts at

initiating conversations with others, and less effective approach behaviors. As a result,

increasing attention has been paid to the importance of identifying the social skills of


individuals with craniofacial conditions and working to increase their social competence

(Bull & Rumsey, 1986; Kapp-Simon, 1995; Tobiasen, 1995).

Noting that the relationships among self-concept, social skills, inhibition, and

adjustment in children with craniofacial conditions had never been explicitly examined,

Kapp-Simon and colleagues (1992) conducted such a study with 45 young adolescents.

They concluded that children with craniofacial conditions appear at risk for psychological

adjustment problems but not for poor self-concept, and suggested that self-concept scores

may not be the best indicators of overall psychological functioning. Adolescents in the

study scored in the normal range on self-perception, social skills, and inhibition, but 1.5

SD below the average on a measure of adjustment. Regression analyses were conducted.

Social skills were the primary predictor variable and, together with athletic competence,

accounted for 73.5% of the variance in adjustment. According to the authors, these

results suggest that adjustment in children with craniofacial conditions is related to social

skills and social behaviors rather than feelings about appearance, school performance, or

even their own sense of self-worth. However, other researchers have found that self-

perceptions, particularly of appearance, play an important role in psychosocial

functioning and continue to emphasize the need to understand both maladaptive

cognitions and coping styles in addition to behavioral deficits (Richman, 1983; Richman

et al.,1985; Tobiasen & Hiebert, 1993).

Research on the adjustment of children and adolescents with craniofacial

conditions is often confusing and inconsistent due to disparities in variables measured,

instruments used, and age and gender differences among subjects. However, several

patterns have emerged. As a group, children with craniofacial conditions are well

functioning and not appreciably different from their nonaffected peers. Specific

problems have been noted in subgroups of children and areas of functioning, for example,

social and peer interactions, school achievement, a tendency toward internalizing

behavior problems, and diminished self-concept in the social and physical attractiveness

domains. However, global self-concept is usually comparable to, or only slightly lower

than, that of nonaffected peers. Given the expectations of poor adjustment derived from

literature on the physical attractiveness stereotype and stigmatization combined with

social psychological theories on the development of self-esteem, the generally high level

of functioning is somewhat surprising. Although the need for increased knowledge has

been emphasized (Bull & Rumsey, 1986; Pruzinsky, 1992; Rubin & Wilkinson, 1995;

Tobiasen, 1995), little is known about specific risk and resiliency factors, including the

possible use of "self-protective strategies" in lessening the impact of negative stereotypes

and instances of prejudice or discrimination.

Self-Protective Strategies and Stigmatization

A significant body of research over many years has led to the rather surprising

conclusion that prejudice against members of stigmatized or oppressed groups generally

does not result in lowered self-esteem for members of the group. Without denying that

prejudice and discrimination may be harmful in other ways, Crocker and Major (1989)

proposed that three mechanisms, or "self-protective strategies," mediate the association

between stigmatization and global self-esteem among members of discredited groups.

Although the relationship has not been studied, it is possible that these strategies also

help buffer the self-esteem and self-concept of children and adolescents with craniofacial

conditions from the effects of negative perceptions and treatment by others. Furthermore,

differential use of these strategies may contribute to the observed variability in

psychosocial functioning among these children and adolescents.

Two of the self-protective strategies or mechanisms outlined by Crocker and

Major (1989) are those of"ingroup comparisons" and "selective devaluation." As

members of stigmatized groups are often disadvantaged on numerous dimensions,

comparing oneself to similarly stigmatized others ingroupup comparison") rather than to

members of advantaged groups may protect one's self-esteem. Ingroup comparisons are

often forced on stigmatized individuals as a result of segregated environments but may

occur because people tend to compare themselves to others who are similar to them on

important dimensions. The wish to enhance self-esteem by comparisons with alike or

even less fortunate individuals may also motivate ingroup comparisons. The importance

of reference groups has been shown in assessing children's self-concept in academic

settings (Harter, 1986; Marsh & Parker, 1984) and in the maintenance of self-esteem

among women with breast cancer (Taylor, 1983; Taylor, Wood, & Lichtman, 1983).

Researchers have also shown that both self-esteem and affective state are related to

beliefs about how one's abilities and attributes compare with those of others (Crocker,

Alloy, & Kayne, 1988; Tabachnik, Crocker, & Alloy, 1983).

Selective devaluation, or regarding as less important to one's self-concept those

dimensions on which one does poorly and valuing those on which one or one's group

excels, may also serve to protect the self-esteem of stigmatized individuals. Since

William James (1890/1981), the idea that it is necessary to know whether an individual

values a quality in order to know whether that individual's self-esteem will suffer as a

result of deficiencies in that quality has guided thinking about the self-concept and the

preservation of self-esteem. Harter (1986) showed that self-esteem was linked to

children's abilities to discount areas in which they were not competent. As it may be

difficult to completely discount the importance of certain attributes in a given culture or

sub-culture (e.g., appearance or academic success in the United States today), one self-

protective strategy used by members of disadvantaged groups may consist of relatively

overvaluing those attributes or domains in which they excel rather than devaluing those

in which the dominant group excels.

Attributing negative feedback or relatively poor outcomes to prejudice against

one's group is the third self-protective strategy suggested by Crocker and Major (1989). It

is also the focus of this research. For example, if an individual with a craniofacial

condition is not invited to party or fails to get a job, he or she may be uncertain whether

the negative outcome is due to actual personal inadequacies or because the other person is

prejudiced. This uncertainty, a characteristic reaction of stigmatized individuals

(Crocker, Voelkl, Testa, & Major, 1991; Goffman, 1963), allows one to attribute

rejection to external causes such as prejudice rather than to personal deficiencies even if

no prejudice exists, and thus preserve one's self-esteem.

The idea that people seek causal explanations for events, particularly those that

are negative, unexpected, or important, and that one's thoughts or perceptions about the

causes of events has affective and behavioral consequences is central to attribution theory

and the focus of much theoretical and clinical research (Abramson, Seligman, &

Teasdale, 1978; McFarland & Ross, 1982; Metalsky & Abramson, 1981; Schoenherr,

Brown, Baldwin, & Kaslow, 1992; Weiner, 1988; Weiner et al., 1988). For example,

according to both the reformulated account of learned helplessness (Abramson et al.,

1978) and the hopelessness theory of depression (Abramson, Metalsky, & Alloy, 1988;

Abramson, Metalsky, & Alloy, 1989), the causal attributions people make for

uncontrollable events they experience influences their helplessness and hopelessness,

self-esteem, and level of depression. Weiner, another prominent attributional theorist,

places attributions at the center of his theory of motivation and emotion in both

achievement and affiliative domains (Weiner, 1986; Weiner, 1988).

Considerable evidence from the attributional theory literature indicates that

making external attributions for negative events bolsters self-esteem, and, conversely,

attributing negative outcomes or feedback to internal causes (i.e., lack of ability or being

overweight) leads to lower self-esteem among individuals and groups (Bradley, 1978;

Crocker, Comwell, & Major, 1993; Ickes, 1988; Tennen & Herzberger, 1987). A

tendency to blame others for negative outcomes, whether justified or not, has been

observed among stigmatized groups, including those with craniofacial conditions (Bull &

Rumsey, 1986; Crocker & Major, 1989; Goffman, 1963; MacGregor, Abel, Bryt, Lauer,

& Weissman, 1953). The danger, of course, to this approach is that externalizing

responsibility for poor outcomes may protect self-esteem in the short-term but have long-

term negative effects if it prevents one from making necessary changes.

In addition to mediating self-esteem, causal attributions may mediate the affective

and behavioral responses to positive and negative outcomes among both children and

adults. Attention has been paid to the role of attributions in fostering depression, social

anxiety, and loneliness, and their influence on perceptions of social competence and

achievement motivation (Crocker et al., 1993; Dweck & Goetz, 1978; McFarland &

Ross, 1982; Michela, Peplau, & Weeks, 1982; Mikulincer, 1988; Seligman et al., 1984;


Sobol & Earn, 1985; Tennen & Herzberger, 1987; Weiner, 1988; Vaux, 1988). Much of

the research in this area has, in fact, focused on the role of causal explanations in the

development of the same social and internalizing behavior problems associated with

having a craniofacial condition.

Individuals are considered to exhibit a particular attributional style to the extent

that they make similar attributions for different events. Several researchers (Ickes, 1988;

Ickes & Layden, 1978; Metalsky & Abramson, 1981; Weiner, 1986) have asserted that

certain characteristic styles of attribution may predispose an individual to psychological

difficulties. Factors such as utilization of situational information, strength of attributions,

and differentiation of causal explanations across events and time contribute to the

enormous variability in attributional style among and within people. While

acknowledging the importance of situational effects, the trait aspects of attributional style

are the primary focus here, and an understanding of differences in attributional styles

requires consideration of the dimensions usually associated with causal explanations:

internal-external, stable-unstable, controllability-uncontrollability, and global-specific.

Although essentially similar, Metalsky and Abramson's (1981) and Weiner's

(1986) attributional style theories differ in their views about the controllability and

globality dimensions. Metalsky and Abramson, for example, maintain that people who

tend to attribute bad outcomes to internal, stable and global factors and good outcomes to

external factors such as "good luck" have a negative attributional style and should be

most prone to depression and low self-esteem. Those with positive attributional style

generally attribute good outcomes to internal factors and bad outcomes to external

factors. Although he substitutes controllability for globality, Weiner also postulates

fundamental differences between individuals with negative and positive attributional

styles in achievement and affiliative domains. As noted previously, evidence suggests

that external attributions for negative events are linked to high self-esteem and, like

overly positive self-evaluations, may actually promote mental health (Taylor & Brown,


The hypothesis that attributions mediate affective and behavioral reactions to

positive and negative outcomes is consistent with empirical evidence on self-esteem and

psychological well being in both adults and children. In a series of studies, Ickes and

Layden (1978; Ickes, 1988) explored the interrelation among attributional style, self-

esteem, and sex. For negative outcomes, low self-esteem participants made internal

attributions while high self-esteem participants made external attributions. The reverse

was true for positive outcomes. Regarding sex differences, they found that irrespective

of self-esteem level, males tended to attribute failure to external factors and success to

internal factors while females did the opposite. A separate study examined the

relationship between ability attributions and self-conceptions and found that attributing

negative outcomes to a lack of ability was positively associated with depression (Ickes,

1988). Individual differences in self-esteem and affective states have also been linked to

attributional differences in other studies. Tennen and Herzberger (1987) found that

higher self-esteem subjects attributed positive, but not negative, outcomes to internal and

stable causal factors whereas low self-esteem subjects tended to make internal

attributions for failure and external, unstable, and specific attributions for success. In

Mikulincer's (1988) experiment, subjects were exposed to a series of unsolvable

problems. Individuals with an internal attributional style for negative outcomes were

shown to experience more depression and a decrease in performance than were those

with an external style.

Abramson, Metalsky, and Alloy (1988, 1989) proposed a hopelessness theory of

depression with a diathesis-stress component. This model predicts that a style of

attributing negative life events to internal, stable, and global causes and viewing these

events as important (the diathesis) interacts with the actual occurrence of negative life

events (stress) to increase the probability of the hopelessness subtype of depression.

Although never examined, it may be that a similar process influences the development of

internalizing behavior problems in individuals with craniofacial conditions or stigmatized


Empirical research has, in fact, shown that the stigmatized may protect their self-

esteem and decrease negative affect by attributing negative feedback to external factors

such as prejudice rather than to internal deficiencies (Crocker et al., 1993; Crocker et al.,

1991). In the first study, overweight women who received negative social feedback from

a male evaluator and who attributed the feedback to their weight and not to the

evaluator's prejudice experienced more negative affect and depression than women who

did not make internal attributions for social rejection (Crocker et al., 1993). In another

study, women who received negative feedback from a prejudiced evaluator and attributed

the feedback to his prejudice (an external factor) reported less depression than did women

who received negative feedback from a nonprejudiced evaluator (Crocker et al., 1991).

In both situations, women who made external attributions for negative outcomes were

better able to protect their self-esteem and decrease their likelihood of depressed affect.

A related experiment (Crocker et al., 1991) with African-American and White

participants and a White evaluator showed both a self-protective effect and a differential

response to positive versus negative feedback. African-American subjects discounted the

negative feedback they received from a White evaluator and, in so doing, maintained a

high level of self-esteem. However, when African-American participants received

positive feedback from a White evaluator, their self-esteem suffered. The authors noted

that these findings are congruent with Goffman's (1963) hypothesis that interactions

between stigmatized and nonstigmatized individuals are often governed by ambivalence,

uncertainty, and difficulty interpreting the motives of others. Thus, stigmatized

individuals may be particularly sensitive to the motives underlying positive feedback.

Such feedback may have positive consequences for self-esteem only when the

stigmatized are certain the feedback reflects their deservingness, and not special

consideration or a fear of appearing prejudiced on the part of the nonstigmatized.

Affect and attributional style appear linked in children as well. In the Seligman et

al. study (1984), children who attributed bad events to internal, stable, and global causes

were more likely to report depressive symptoms than were children who attributed these

events to external, stable, and specific causes. Attributional style also predicted

depression six months later, suggesting that it is a risk factor for affective disorders rather

than vice versa. The possibility that children may copy their parents' attributional style

was suggested by the high correlations between styles of mothers and their children.

Attributional style also predicted depressive symptomatology in children with chronic

illnesses (Schoenherr et al., 1992).

While most of the research examining the relation of attributional style to

problems in living has focused on depression, some attention has been paid to loneliness,

shyness, and difficulties in social and peer interactions. Although they were primarily

interested in evaluating the dimensions of attributional style, Anderson and Riger (1991)

found that a negative attributional style, as measured by the Attributional Style

Questionnaire (Peterson et al., 1982) was predictive of loneliness in interpersonal

situations. And, in discussing the connections among shyness, self-esteem, and self-

consciousness, Cheek and Melchior (1990) reviewed the research literature indicating

that shy people tend to accept personal responsibility for failure and attribute successes to

external, unstable causes. Noting that shy people blame themselves for social difficulties

and tend to ignore, resist, or doubt the accuracy of positive evaluations, the authors

asserted that "the cognitive tendencies involved in shyness are persistently maladaptive

before, during, and after social interactions (p.71).

The relationship between attributional style and success and failure in the social

relationships of children has also been examined. Dweck and Goetz (1978) examined

children's attributions for social rejection and found they were able to distinguish

between internal, stable causes such as ability, and external and unstable causes such as

the mood of the rejector. Children who attributed rejection to internal and stable causes

had more difficulty continuing to pursue their social goals than did children who

presented external explanations for rejection. Additional studies support the influence of

attributions on children's social and peer interactions and on their self-esteem.

Furthermore, there is evidence that altering mistaken attributions for poor social


interactions may lead to better social relations in the future (Dodge, McClaskey, &

Feldman, 1985; Sobol & Earn, 1985; Sobol, Earn, Bennett, & Humphries, 1983).

Causal attributions clearly reflect an important aspect of an individual's cognitive

style or appraisal of events and situations and may, as one researcher has noted, even act

as unconscious influences on behavior (Dodge et al., 1985). As presented here,

attributions may serve as "protective" strategies and mediate affective reactions to

positive and negative outcomes. Evidence suggests that a positive attributional style, in

which stable, internal factors are seen as responsible for success and external ones for

failure, may contribute to the generally high levels of self-esteem and good psychological

adjustment found among children and adolescents with craniofacial conditions.

Inadequate use of these strategies may also help explain the low self-esteem and

internalizing behavior problems displayed by some children with facial disfigurements.


Literature on the physical attractiveness stereotype and stigmatization combined

with social psychological theories on the development of self-esteem has led to

predictions of major difficulties for individuals with craniofacial conditions. It does not

predict the reality of relative strength and psychological health characteristic of most of

these children. Like their peers with other chronic and/or life threatening conditions,

children with facial differences do not differ substantially from healthy children in their

emotional or psychological functioning. However, they may be at higher risk for specific

behavioral problems, difficulties in social adjustment, and poorer school achievement.

Furthermore, evidence suggests there are significant differences in functioning within

subgroups of children with facial disfigurements.

The fact that specific problems have been noted within subgroups of these

children and adolescents is not surprising. Research shows the presence of clear

normative standards of appearance, and differential perception and treatment of attractive

and unattractive individuals. These differences apply both to peer- and adult-child

interactions. Individuals with craniofacial conditions are viewed more negatively and are

likely to experience instances of prejudice and discrimination. Thus, the stress of coping

with a chronic medical condition may be exacerbated by instances of both felt and

enacted stigmatization. However, the discrepancy between predictions from the literature

and the observed health of many of these children and adolescents highlights how little is

known about specific risk and resiliency factors and variables that may mediate the

impact of negative experiences.

As noted earlier, theoretical models such as Thompson's transactional stress and

coping model and Wallander and Vami's risk and resiliency model (Thompson, Gil,

Burbach, Keith, & Kinney, 1993; Thompson, Gustafson, Hamlett, & Spock, 1992; Vami

& Wallander, 1988; Wallander et al., 1989; Vami & Setoguchi, 1996; Wallander &

Thompson, 1995; Wallander, Feldman & Vari, 1989; Wallander & Vami, 1989),

provide a heuristic framework for exploring sources of vulnerability and resistance in

children and their families. Using such a multivariate, conceptually based approach, the

model developed for this research (See Figure 1) specifically explores the roles of

perceived stigma and attributional style within a larger framework of individual, family,

and social factors. Given the large number of variables in the model, many variables of

both theoretical and practical interest cannot be examined in this study. These include

psychosocial stressors such as major life events, daily hassles, instances of enacted

stigmatization, prejudice, and discrimination, and condition-related variables. Other

variables of interest crucial to individual functioning of individuals but not specifically

examined in this study include social support; family environment and patterns of

interaction; and intrapersonal variables such as temperament, intelligence, and

competence (Benjamins, Hillman, & Saddler, 1993; Benson, Gross, Kellum, Messer, &

Passmore, 1991; Campis, DeMaso, & Twente, 1995; Greenberg & Quamma, 1994;

Hamlett, Gustafson, Spock, & Thompson, 1992). What are examined are the

relationships among certain demographic and condition variables, perceptions of

stigmatization, attributional style, self-concept, and specific indices ofpsychosocial

Psychosocial Stressors
Condition-related problems
Life events

Figure 1. Proposed Model Using Perceived Stigma and Attributional Style.
(Model variables in italics are examined in this study)

adjustment in adolescents with a variety of craniofacial conditions. Independent

variables include age, gender, SES, perceived appearance, number of functional

impairments, special school placement or remediation, and perceptions of stigmatization.

Adjustment or outcome variables include global self-concept, internalizing and

externalizing behaviors, social adjustment and peer interactions. Attributional style is

hypothesized to mediate the relationship between perception of stigmatization and the

outcome variables. The following questions are addressed:

* What is the level of perceived stigma among adolescents with craniofacial

* What is the attributional style of adolescents with craniofacial conditions in
regard to positive and negative outcome of events?

* What is the relationship between perceptions of stigmatization, attributional
style, and general measures of self-concept?

* What is the relationship between perceptions of stigmatization, attributional
style, and behavioral, social, and academic adjustment?

Developmental differences in self-concept and an awareness of the age-related

effects of certain variables led to the focus upon adolescents in this study. The normal

developmental tasks of adolescence include the formation of a sense of personal identity,

the establishment of satisfactory peer relations, and the successful negotiation of school

and educational requirements directed toward future career considerations. Adolescents

consistently identify these areas, together with physical appearance, as their primary

concerns (Violato & Holden, 1988; Vami & Setoguchi, 1996). These areas are known to

be among the most difficult for children and adolescents with craniofacial conditions, and

adolescents may be particularly vulnerable to the development of problems.

Furthermore, positive peer relations and satisfactory social relations are of paramount

importance to healthy development: they have been shown to influence academic

success in high school and presage the mastering of intimate adult relationships. They are

also likely to be significantly influenced by experiences of rejection and discrimination,

and by the attributions one makes for negative experiences.

Previous models of adjustment in this population ignored both subjective

experiences and internal mediating factors. Experiences with stigmatization are

presumed to influence adjustment but nothing is known about an individual's subjective

experience. Furthermore, cognitive processes and coping strategies are known to

influence outcome among chronically ill children but we have limited understanding of

these factors in this population. Incorporating such variables into research on children

and adolescents with craniofacial conditions may help explain disparities in previous

studies and increase the knowledge necessary for identifying children at high risk for

psychological maladjustment. Factors promoting greater resiliency in these children and

their families may also be identified.


Based upon the preceding literature review, the hypotheses of this study are as

I. Age, gender, SES, and number of functional impairments will not significantly
predict perceived stigma, general self-concept, social adjustment, peer relations, or
externalizing and internalizing behavior problems.

II. Greater satisfaction with perceived facial and physical appearance will predict
higher general self-concept, social adjustment, and peer relations. Greater
satisfaction with perceived facial and physical appearance will predict lower
externalizing and internalizing behavior problems.

III. Higher perceived stigma will predict lower general self-concept, social adjustment,
and peer relations. Higher perceived stigma will predict higher externalizing and
internalizing behavior problems.

IV. Perceived stigma will contribute to the prediction of general self-concept, social
adjustment, peer relations, and externalizing and internalizing behavior problems
even when other independent variables (age, sex, gender, SES, number of
functional impairments, perceived appearance) are used in predicting the outcome

V. The relationship between perceived stigmatization and the adolescent
characteristics of general self-concept, social adjustment, peer network and
internalizing and externalizing behavior problems will be mediated by the
adolescent's attributional style.


Participants and Settings

Eighty-three dyads composed of adolescents (53 male, 30 female) with a visible

facial difference and a parent or guardian participated in this study. Participants were

recruited from comprehensive craniofacial clinics in three Florida cities. Of those

participating, 36 (43.37%) were affiliated with the University of Florida Craniofacial

Center (UF) in Gainesville, 25 (30.12%) were affiliated with the Nemours Children's

Clinic (NCC) in Jacksonville, and 22 (26.51%) were affiliated with the University of

South Florida Craniofacial Programs (USF) in Tampa-St. Petersburg.

Adolescents were eligible to participate if they were between 13.0 and 18.0 years

of age with a visible facial difference/craniofacial anomaly, were not enrolled in an

educable or trainable mentally handicapped classroom, had no profound physical

disability such as blindness or inability to walk, and had no more than a moderate hearing

loss (speech reception threshold of 40-60dB) in their better ear. Additionally, both

adolescent and participating parent/guardian were required to speak and read English.

Participants were screened in three stages to insure that eligibility requirements

were not breached. First, clinic records were reviewed to identify potential adolescents.

Secondary screening occurred when potential participants were contacted and invited to

participate in the study. Finally, after informed consent was obtained, each adolescent's

medical record was reviewed to determine if inclusion criteria were met. Of the 133


families identified and actually contacted, 16 were ineligible due to developmental delay

or mental handicap, parent unable to read English, or adolescent not living at home, and

19 declined to participate. Of the parents who declined, three stated that their adolescent

was well-adjusted and did not think of themselves as having a facial difference and three

reported their child had psychosocial problems related to their craniofacial condition and

the child did not wish to discuss anything related to it. Eleven parents or adolescents

declined due to lack of time or interest, and two parents reported significant family crises

(recent death or divorce) that precluded participation. Thus 98 (84%) of the 117 eligible

families contacted agreed to participate. However, eleven did not sufficiently complete

one or both sets of questionnaires and data from four families were discarded because it

was later determined that the adolescent did not meet inclusion criteria. Although the

parents of these four adolescents denied any problems during the initial screening

interview, subsequent review of the medical records indicated that their adolescent had

either severe hearing loss or developmental delay that made them ineligible to participate.

The final sample included 83 adolescent/parent dyads, a participation rate of 71%.

Adolescent participants ranged from 13 to 18 years with a mean age of 14.6 years

(S = 1.57 years). Sixty-eight (81.9%) of the adolescents were Caucasian, 6 were

African-American (7.2%), 6 were Hispanic (7.2%), and 3 (3.6%) were from other ethnic

groups. Forty-one adolescents (49.8%) had a diagnosis of unilateral cleft lip

with/without cleft palate (UCL/P), 16 had bilateral cleft lip with/without cleft palate

(19.27%), 13 (15.66%) were identified as having a craniofacial syndrome, and 13

(15.66%) reported various other craniofacial diagnoses. Individuals with a particular

craniofacial syndrome (e.g., Apert or Crouzon) have a recognized pattern of differences

that occur in various areas of the body and are considered to have a single and specific

cause. The features characteristic of a specific syndrome can occur to different degrees,

and a child can have many, some, or only a few of the features. Specific craniofacial

diagnoses such as cleft lip, cleft lip and palate, or hemangioma also vary in severity but

these conditions involve only the structures of the head and face the bones and soft

tissues such as nerves, muscles and fat. Table 1 gives a complete list of the conditions

represented. For some analyses, participants were grouped into three general diagnostic

categories: Group I Unilateral and bilateral cleft lip with/without cleft palate (n=57,

68.7%); Group II Craniofacial syndromes (n=13, 15.7%); Group III Other diagnoses

(n=13, 15.7%).

Of the participating parents or guardians, 67 (80.7%) were mothers, 14 (16.9%)

were fathers, and 2 (2.4%) were guardians or other biological relatives. Fifty-six (67.5%)

of the participating adults were married, 18 (21.7%) were divorced, and 9 (10.8%)

reported some other marital status. Families averaged 2.78 children (n_= 82, SD = 1.52).

Mean SES score, calculated using the Hollingshead four factor index of social status

(Hollingshead, 1975), was 40.13 (n=. 82, SD = 11.71, Mdn = 39.57), corresponding to

the lower end of the minor professional/technical classification. SES raw scores were

normally distributed and ranged from 11 to 64. Demographic and condition-related

information is provided in Tables 2 and 3.

Group comparisons were conducted to determine if there were any significant

differences on selected demographic and condition-related variables among participants

recruited from the three institutions. Chi-square analyses were used for the categorical

variables of gender and diagnostic group. Analyses of variance (ANOVAs) were used

Table 1

Craniofacial Diagnoses of Participants by Institution

Diagnosis (n=36) (n= 25) (= 22) ( = 83)

Unilateral Cleft Lip/Palatea 20 10 11 41

Bilateral Cleft Lip/Palatea 6 6 4 16

Craniofacial Syndromesb 5 4 4 13
Treacher-Collins 0 0 1
Goldenhar 1 1 0
Microtia 0 2 0
Hemifacial microsomia 2 0 2
Crouzon's 1 0 0
Pierre-Robin 0 0 1
Velocardiofacial (VCFS) 1 0 0
Stickler 0 1 0

Other' 5 5 3 13
Ectodermal Dysplasia & CL/P 2 0 1
Bilateral dysmorphic ears 0 2 0
Hemangioma 1 0 0
Parry-Rhomberg syndrome 1 0 0
Facial Nevus 0 0 1
Macrostomia 1 0 0
Facial asymmetry 0 2 1
Scleroderma with Cleft Lip 0 1 0

Note. In selected subsequent analyses, participants are grouped by diagnosis in the
following manner: aGroup I Unilateral and bilateral cleft lip; bGroup II Craniofacial
syndromes; Group III Other diagnoses.

Table 2

Demographic Information by Institution

Variable (n=36) (n=25) (n=22) (N=83)

Age (in years)
M 14.80 14.45 14.47 14.61
SD 1.58 1.50 1.65 1.57

Male (n) 26 16 11 53
Female (n) 10 9 11 30

# Hospitalizations
M 6.23 6.28 6.61 6.34
SD 3.62 3.16 5.22 3.94

# Surgeries
M 6.10 6.20 6.95 6.36
SD 3.41 3.15 5.00 3.80

SES Score
M 44.57** 38.56 34.66** 40.15
SD 12.07 9.24 11.29 11.86

Mother's Educationa
(years completed) M 14.19 13.12 12.57 13.45
SD 2.54 1.39 2.56 2.34

Father's Educationb
(years completed) M 14.49 13.00 12.73 13.64
SD 3.87 2.02 2.05 3.10

#Children in family'
M 3.00 2.84 2.33 2.78
SD 1.96 1.18 0.86 1.52
Note. n = 82. bn= 75.
**difference is significant at p < .01.

Table 3

Rates of Special Educational Placement and Treatment-Related Variables

Variable Number Percentagg

History of Special Educational Placement
Specific Learning Disability 19 22.90
Speech 17 20.50
Gifted or Advanced 8 9.60
Speech and SLD 4 4.8
Speech and Gifted 2 2.4
Hearing Impaired 3 3.6
None 29 34.90

Number repeated grade in school 20 24.10

Number receiving treatment for
Hearing 12 14.5
Speech 8 9.6
Dental or Orthodontic 68 81.1

Number anticipating future surgery 59 71.1

Number of current functional impairments per adolescent
0 55 66.3
1 16 19.3
2 10 12.0
3 2 2.4

for age, SES score, and number of hospitalizations and surgeries. Participants from the

three institutions did not differ on any of the measures except for SES score, E (2,80) =

5.82, 1 = .004. Families from UF/Gainesville obtained significantly higher mean SES

score (44.57) than did families from USF/Tamp-St. Petersburg (34.66). The difference

between Gainesville and Jacksonville participants' SES was not significant.

The three diagnostic categories were also compared on demographic and

condition-related variables. Chi-square analysis indicated that gender and diagnosis were

significantly related, X2 (2, N = 83) = 7.81, 1 = .02, with proportionately more males and

fewer females than expected in the cleft lip/palate group (Group I) and proportionately

fewer males than females in Group III (Other diagnoses). The greater number of males in

the cleft lip/palate group is as expected given the 2:1 sex ratio of males to females

identified for this condition by epidemiologic studies (Owens et al., 1985). Analyses of

variance indicated the differences among diagnostic groups on SES score, number of

functional impairments, and number of surgeries were not significant. However, number

of hospitalizations was significantly different for adolescents in Diagnostic Groups I and

II, F(2,75) = 3.22, p < .05. Adolescents in Group I (UCL/P and BCL/P) averaged 6.95

hospitalizations (SD=3.49) while those in Group II (Craniofacial syndromes) averaged

3.85 (SD= 3.49). Mean number of hospitalizations for those in Group III (Other

diagnoses) was 6.38 (SD=5.97).


Measures Completed by Adolescent Participants

Physical appearance

Self Description Questionnaire II, Physical Appearance Scale (SDQ-

II; Marsh, 1990). The construct of general perceived physical appearance was measured

using the eight-item Physical Appearance subscale of the Self Description Questionnaire-

II (Marsh, 1990). This subscale assesses the degree to which the adolescents are happy

with the way they look, think they have a nice looking face and body, and compare their

appearance to their peers. Internal consistency (coefficient alpha) reliability of the

Physical Appearance subscale is .91 (Marsh, 1990). Test-retest reliability over seven

weeks was .75 (Marsh, 1990). The SDQ-II is based upon a multifaceted, hierarchical

model of self-concept and designed for individual scale interpretation. Scores range from

8 to 48, with higher scores indicating greater satisfaction with one's physical appearance.

Perceived stigma

Social Perception Questionnaire. The construct of perceived stigma was

operationally defined as the subjective belief that having a facial difference/craniofacial

anomaly negatively affects social relationships and interactions. There is no standardized

measure of perceived stigma designed for use in this population. However, a

questionnaire designed to assess perceived stigma among individuals 12 to 20 years old

with epilepsy has been used in several studies (Westbrook et al., 1992) and was modified

for use in this research (see Appendix A). The Westbrook et al. questionnaire included

four items that, in a factor analysis, loaded on a single factor with relatively large

coefficients; varimax factor loadings were .91, .90, .75, and .52. Interitem correlations

were generally moderate (average .46) and internal consistency was .78, as measured by

Cronbach's alpha. The wording of the items was changed so that "facial

difference/craniofacial anomaly" was substituted for the word "epilepsy". In addition, the

wording of one of the items was changed to replace "sexual relations" with "can date or

have intimate relationships". Questions were scored on a 0-5 point Likert (visual analog)

scale. Zero-point responses indicate no perceived stigma, while the adolescents graded

positive perceived stigma responses from 1 to 5. Higher numbers indicated greater

perceived stigma. Items were summed to produce a total score for use in analyses; 20

was the maximum possible score.

Attributional style

KASTAN Children's Attributional Style Questionnaire (CASQ; Seligman et

al., 1984) is the major instrument utilized by researchers to examine the ways youth 7 to

18 years attribute causality for positive and negative events. The 48 events tap three

attributional dimensions: interal-external, stable-unstable, global-specific. The internal-

external dimension refers to whether the locus of causality is internal (due to self) or

external (due to others or circumstances). The stability dimension refers to the degree to

which the child views the cause of the event as being stable over time. The globality

dimension refers to the degree to which the child views the cause of the event as being

generalizable across situations. Each item comprises a situation (e.g., "You get good

grades") and two possible attributions to explain why the situation occurred (e.g., "I am a

hard worker" vs. "School work is easy"). Children are instructed to choose the alternative

that best describes why the event happened to them. Half the situations represent good

outcomes; half represent bad outcomes. A composite score for positive and negative

events is derived as well as an overall composite score. The lower the difference score

between positive and negative event composite scores (summary score), the more

pessimistic is the child's explanatory style for causes of events. Adequate psychometric

properties, including test-retest reliability and internal consistency have been

demonstrated (Kaslow, Rehm, & Siegel, 1984; Seligman et al., 1984).


Self Description Questionnaire-II, General Self Subscale (SDQ-II; Marsh,

1990). The General Self subscale of the SDQ-II was used as a measure of global self-

concept. The SDQ-II is a 102 item questionnaire designed to measure self-concept in

adolescents in grades 7 through 12. Each of the 11 SDQ-II domains contains 8-10 items,

half of which are negatively worded to disrupt positive response biases. The SDQ-II

subscales are: Physical Abilities, Physical Appearance, Opposite-Sex Relations, Same-

Sex Relations, Parent Relations, Honesty-Trustworthiness, Emotional Stability, Math,

Verbal, General School, and General Self. In addition, a Total Scale score is calculated

by summing the individual subscale scores. The coefficient alpha estimate of reliability

for each subscale varies from .83 to .91, whereas the average correlation among the

factors is modest. The scale is based upon a theoretical model that conceives of the self-

concept as a multidimensional, hierarchical entity. Although normed on a group of 5,494

Australian adolescents, the SDQ-II has been used extensively throughout the world,

including within the United States. It produces scores comparable to those obtained from

a multidimensional self-concept scale (Bracken, 1992) normed in the United States

(r=.80), and factor structure and construct validity have been consistently supported.

Internal consistency (coefficient alpha) reliability for the General Self subscale is .88 and

test-retest reliability over a seven-week interval is .85 (Marsh, 1990). Scores on the

General Self subscale range from 10 to 60, with higher scores indicating higher overall

self-worth and self-satisfaction. As only intragroup comparisons were made, raw scores

were used in all calculations.

Social adjustment

Social Skills Rating System (SSRS; Gresham & Elliott, 1990). The SSRS

samples the domains of social skills, academic competence, and problem behavior. It

was designed for use with normal and slightly handicapped students for whom social

skills deficits might limit academic performance. The Social Skills sub-scale of the


Student Form was used as a measure of adolescent social adjustment. The SSRS offers

Parent, Teacher and Student Forms at three developmental levels (i.e., preschool,

kindergarten through sixth grade, seventh through twelfth grade). Only the Parent and

Student forms for grades seven through twelve were used. The student version is for

youth who are capable of reading at the third grade level and assesses only social skills

while the parent version assesses both social skills and problem behaviors. The SSRS

standardization sample included 4,170 self-ratings of children and youth, 1,027 parents,

and 259 teachers. The sample represented White, Black and Hispanic individuals drawn

from all regions of the United States, and urban, rural, and suburban communities.

Internal reliability (coefficient alpha) is excellent for all forms in regard to the Social

Skills Scale (.83 .94) and Problem Behavior Scale (.73-.88). Coefficient alphas are .83

for the Secondary student self-report Social Skills scale and.90 for the Parent report

Social Skills scale (Gresham & Elliot, 1990). Test-retest reliability for the self-report

Social Skills scale is .68, computed for a 4-week period using the Elementary

standardization sample. Content, social, criterion-related, and construct validity are

adequate. Raw scores were used in all statistical analyses.

Peer Interaction Record (PIR;Thompson, 1994). The number of adolescents'

social interactions with peers was assessed using a modification of the Health and Daily

Living Form (Moos, Cronkite, Billings, & Finney, 1984) entitled the Peer Interaction

Record (PIR; Thompson, 1994). The PIR asks the adolescent to estimate how often they

engaged in each of 12 typical peer activities during the past week. The number of

different activities engaged in and the number of different peers mentioned across the 12

activities were calculated to estimate the adolescent's activity level and peer network.

The PIR has been shown to be sensitive to differences in the quality and quantity of peer

interactions in adolescents with chronic illnesses (Pendley, Dahlquist, & Dreyer, 1997),

and discriminates between same- and opposite-sex interactions, and between group (e.g.,

Boy Scouts, organized religious activities) and peer-initiated activities (See Appendix B).

Measures Completed by Parent or Guardian

Demographic information

Demographic Questionnaire. Parents completed a brief questionnaire (see

Appendix C) identifying marital status, parents' education and occupation, ethnicity, and

number of children. Information regarding their adolescent's school status and current

and past special educational placements was also obtained. Condition-related variables

such as number of hospitalizations and surgeries, and number of functional impairments

(e.g., speech or chronic hearing problem, learning disability) requiring therapy or

treatment at the time of the study were also assessed by parent report. Number of

functional impairments was totaled with one point given for each impairment reported by

the parent or recorded in the medical record. Scores ranged from 0 (no functional

impairments) to a maximum of 3 (1 point each for hearing impairment, speech

impairment, and special educational placement).

Perceived stigmatization

Social Perception Questionnaire. The parent or primary caregiver completed

the Social Perception Questionnaire, described previously, giving their perceptions of

their adolescent's experience of stigmatization. The wording of the items was altered to

reflect this change, i.e., replacing "your appearance/facial difference" with "your child's

appearance/facial difference". Internal consistency of the original four-item scale was

.78, as measured by Cronbach's alpha correlation coefficient (Westbrook et al, 1992). A

copy of the questionnaire is included in Appendix A.

Attributional style

The Attributional Style Questionnaire (ASQ; Peterson et al., 1982). The ASQ

presents 12 hypothetical events 6 good and 6 bad for which participants are asked to

make attributions and rate these attributions on internal-external, stable-unstable, and

global-specific dimensions. Composite scores for both positive and negative events are

derived, as well as an overall composite attributional style score. Internal reliability for

the positive and negative subscales, estimated using Cronbach's coefficient alpha, are .75

and .72, respectively.

Social and behavioral adjustment

Social Skills Rating System, Social Skills and Behavior Problems Scales

(SSRS; Gresham & Eilliot, 1990). The Behavior Problem scale of the SSRS, discussed

above, served as the measure of behavioral adjustment. This scale was completed by the

parent/guardian and assessed both internalizing and externalizing behavioral difficulties,

as well as the parent's perception of their adolescent's social skills. Coefficient alphas

are .82 for the Externalizing subscale, .72 for the Internalizing subscale, and .81 for the /

total Problem Behavior scale. Test-retest reliabilities, calculated over a 4-week period for

the Elementary standardization sample, were .58 for Externalizing behaviors and .48 for

Internalizing behaviors (Gresham & Elliott, 1990). The Social Skills scale is a parent-

rated assessment of their child's competence in social interactions and is comparable to

the student form of the SSRS discussed previously.

Peer Interaction Record (PIR; Thompson, 1994). The number of adolescents'

social interactions with peers was assessed using a modification of the Health and Daily

Living Form (Moos, Cronkite, Billings, & Finney, 1984) entitled the Peer Interaction

Record (PIR; Thompson, 1994). The PIR asks the parent to estimate how often their

adolescent engaged in each of 12 typical peer activities during the past week. The

number of different activities engaged in and the number of different peers mentioned

across the 12 activities were calculated to estimate the adolescent's activity level and peer

network. The PIR has been shown to be sensitive to differences in the quality and

quantity of peer interactions (Pendley et al., 1997).


Appropriate Institutional Review Board approval was obtained from all

participating institutions as well as from the Florida Department of Health Review

Council for Human Subjects. Adolescents with facial differences were recruited either

during a routine clinic visit or by telephone from lists provided by the cooperating

institutions. Craniofacial clinic patient rosters were examined and potential participants

identified. Letters explaining the study were sent to the parents or guardians of eligible

patients. Approximately one to two weeks later, either the principal investigator (PI) or a

trained research assistant (RA) called to answer any questions and solicit participation. If

the parent and adolescent were interested in participating, arrangements were made to

collect data either during a clinic visit, at the family's home, or by mail. All participants

were given a copy of the appropriate informed consent form to keep and invited to

receive a summary of study results after it is completed. Each participant was informed

of their right to refuse to answer any question and their right to discontinue their

participation at any time without jeopardizing their family's medical care. Order of

presentation of measures was counterbalanced and the measures were divided into two

groups to minimize order effects. Adolescents were compensated $10.00 for each set of

questionnaires completed or $20.00 for completing all measures. Participating parents

were compensated $5.00 for each set completed or $10.00 total.

Data collected during a clinic visit. A research team member met both

adolescent and parent/guardian, reviewed procedures, and obtained signed Informed

Consent and Assent forms. Participants then completed the first set of measures.

Remaining questionnaires were completed either during a later clinic visit or, if they

preferred, participants were provided with a stamped, addressed, return envelopes and

permitted to finish them at home. Follow-up phone calls were made to review

instructions, answer any questions, and encourage return of all questionnaires.

Data collected by mail or at home. Participants were mailed the appropriate

Informed Consent and Assent, that they were asked not to sign until contacted by the

investigator; detailed instructions; the first set of questionnaires; and a stamped return

envelope. Approximately 5-7 days later the family was contacted by phone and

Informed Consents and instructions were reviewed with both the adolescent and parent.

After the first set was returned, the second set of questionnaires and instructions was

mailed to the family. Again, phone calls were made to ensure measures were

appropriately completed and returned. Two families were mailed the first set of

questionnaires and completed the second set at home, in the investigator's presence.

After obtaining informed consent, appropriate medical records were reviewed for

information regarding the adolescents' craniofacial diagnosis and results of their most


recent hearing and speech tests. All questionnaires were coded by number and kept

separately from the informed consents in files accessible only to the researchers.

Measures were scored according to standard procedures by either the PI or RA and

entered into the database. SES score and category were calculated using the four-factor

index of social position (Hollingshead, 1975).


The first section of this chapter contains descriptive statistics for demographic and

test variables, comparisons between the craniofacial sample means and the

standardization sample means for selected measures and gender, institution, and

diagnostic group comparisons (Tables 3-6). Due to the large number of comparisons

conducted, a more stringent significance level of2=.01 was adopted to minimize the

chance of Type I errors for these analyses.

Intercorrelations of variables are presented next (Tables 7-8), followed by results

of multiple regression tests of the hypotheses (Tables 9-18). All statistical analyses of

the data were performed using the Statistical Package for the Social Sciences (SPSS/Base

10, 1999).

Description of Sample Variables

Adolescent Characteristics

Adolescents ranged in age from 12.80 to 18.00 years with a mean age of 14.61

(SD = 1.57). Mean family SES was 40.13 (SD = 11.71) using the Hollingshead (1975)

four-factor index of social status. Details of condition-related variables are given in

Table 3. In calculating the total number of functional impairments, one point was given

for each impairment (chronic hearing, speech, learning disability) requiring treatment or

special educational placement at the time of the study. Based on these criteria, 55

(66.3%) of the adolescents reported no functional impairments, 16 (19.3%) reported one,

10 (12%) reported two, and 2 (2.4%) reported three functional impairments.

Forty-five (54.2%) of sample adolescents had a history of enrollment in special

classes for speech, specific learning disability, or hearing impairment, and 20 (24.1%)

had repeated a grade in school. At the time of the study, 68 (81.9%) were undergoing

dental or orthodontic procedures, 12 (14.5%) were receiving some treatment for hearing

problems (i.e., antibiotics for ear infections, hearing aids, etc.), and 8 (9.6%) were in

speech therapy. Participants reported a mean of 6.34 hospitalizations (n=83, SD= 3.94)

and 6.34 surgeries (n=83, SD=3.80). A majority of participants (71.1%) anticipated

undergoing additional surgical procedures at some time in the future.

Distribution of Variables

Means, standard deviations, and Cronbach's alpha reliability coefficients for

adolescent variables and questionnaires are presented in Table 4. Those for the parent

questionnaires are given in Table 5. Scores were normally distributed for SES, both raw

and standardized scores on Student and Parent versions of the SSRS Total Social Skills

Scale, (SSRS-Student, SSRS-Parent) and the Parent Attributional Style Questionnaire

(PASQ). The remaining variables were not normally distributed and could not be

transformed to normal distributions. Thus, nonparametric or distribution-free statistical

procedures were used in those analyses where the assumptions required for the use of

parametric techniques were violated.

Adolescent Measures

Attributional style

The mean for the Children's Attributional Style Questionnaire (CASQ) composite

positive events score (M= 13.71; SD =3.63) was not significantly different from the

mean for the groups on which the CASQ was developed (M = 13.46; SD =3.91 ), t(81)=

Table 4

Sample Means and Standard Deviations Adolescents

Variable n M SD a

Adolescent Characteristics

Age in Years
Number of Functional Impairments

Adolescent Ouestionnaires

Attributional Style
CASQ Composite Positive
CASQ Composite Negative
CASQ Overall Score
Physical Appearance
SDQ-II Physical Appearance

SDQ-II General Self 82

Perceived Stigma
Social Perception Questionnaire 83

Social Adjustment
SSRS Student Raw Social Skills 82
SSRS Student Standardized Social Skills 82
PIR Peer Network 82
PIR # Activities 82

14.61 1.57
40.13 11.71
0.51 0.80

13.71 3.63 .59
7.68 3.04 .52
6.02 5.52 .66

33.54 10.14 .91

52.21 8.76 .78

6.58 5.25 .92

52.16 8.62 .82
102.98 14.97
5.31 4.87 --
6.13 2.78 .73

.62, p = .54. However, the difference for the total negative events score was significant.

The craniofacial sample obtained a mean of 7.68 (SD = 3.04) compared to the original

sample's mean of 6.34 (SD = 2.79), suggesting that this group of adolescents with facial

differences were more likely to make internal, stable, and global attributions for negative

events, t (81)= 4.0, p = .000. The CASQ overall composite attributional style score, used

in all statistical analyses, was calculated by subtracting the total score for negative events

Table 5

Sample Means and Standard Deviations Parent Ouestionnaires

Variable n M SD a

Attributional Style
ASQ Positive Composite 80 5.34 .74 .75
ASQ Negative Composite 80 4.13 .83 .78
ASQ Overall Composite 80 1.21 .99

Perceived Stigma
Social Perception Questionnaire 83 5.33 5.05 .96

Behavioral Adjustment
SSRS Parent Total Behavior Problems 82 9.04 4.21 .82
SSRS Behavior Problems-Standardized 82 102.84 14.20
SSRS Parent Externalizing Problems 82 4.39 2.42 .78
SSRS Parent Internalizing Problems 82 4.74 2.45 .76

Social Adjustment
SSRS Parent Raw Social Skills 82 54.26 10.39 .91
SSRS Social Skills Standardized 82 98.62 14.94
PIR Peer Network 81 5.33 4.89

from that for positive events, with a lower score representing a more pessimistic

attributional style. In this sample, the CASQ overall attributional style score (M = 6.02,

SD = 5.52) was slightly lower than, but not significantly different from, that of the

reported sample's mean of 7.12, t (81) = -1.78, 2 = .08. Cronbach's alpha reliability

statistics for the craniofacial sample's composite positive and composite negative events

scales were .59 and .52. Average reported internal consistency reliabilities are .69 and

.52, respectively. Cronbach's alpha reliability for the craniofacial sample's composite

attributional style score was .66, which is within acceptable limits and comparable to

other reported reliability statistics (Seligman et al, 1984).

The three dimensions of attributions (Internal-Exteral, Stable-Unstable, Global-

Specific) for the negative events subscale were compared to those of the normative

sample in order to better understand the factors contributing to craniofacial sample's

more pessimistic attributional style. T-tests indicated no significant group differences on

the Stability dimension, t (79)= .971, p = .334. However, the results indicated that these

adolescents were significantly more likely to attribute negative events to internal (t (79)=

2.69, 1 <.00) and global causes (t (79) = 8.24, p < .00).

Physical appearance

Adolescents' perception of their physical appearance was measured using the

eight-item Physical Appearance subscale of the Self Description Questionnaire-II (SDQ-

II Appearance). The sample mean of 33.45 ( = 10.14) was not significantly different

from the standardization group mean of 31.2 (S = 10.3), (80), =2.0, p = .05. There

were also no statistically significant differences between the craniofacial sample male (M

= 34.26, SD = 10.54) and female (M = 32.16, SD = 9.28) scores and the standardization

group male (M =34.2, S = 9.6), t (49) = .04, p = .97, and female scores (M = 28.4, SD =

10.2), t (30) = 2.26, 2 = .03. Internal consistency reliability for the sample's SDQ-II

Appearance subscale (a = .91) exactly equaled the previously reported coefficient alpha.


Mean self-concept score for the craniofacial sample, as measured by the SDQ-II

General Self subscale, was 52.21 (D = 8.76), significantly higher than that of the

standardization mean of 48.3 (S = 9.7), (80) = 4.02, < .01. When examined by

gender, there were no significant differences in self-concept scores for males in the

craniofacial sample (M.= 52.1, SD = 9.1) versus the standardization group (M = 49.1, SD

= 9.6), t (50) = 2.4, 2 = .02. However, the sample mean for females (M = 52.23, SD =

8.30) was significantly higher than that of the standardization group's mean (M= 47.4,

SD = 9.7), t (29)= 3.2, = .003. Coefficient alpha for this administration was .78,

somewhat lower than the previously reported internal consistency reliability of.88.

Perceived stigma

Perceived stigma was operationally defined as the belief that one's craniofacial

condition and appearance negatively affected social interactions. The Adolescent Social

Perception Questionnaire (ASPQ), composed of four items, was used in an attempt to

quantify this construct. Zero-point responses on each item indicated no perceived stigma,

while positive responses were graded on a 1 to 5 scale. Items were summed to yield a

total score that was used in all statistical analyses. Possible scores ranged from 0 to 20,

with higher scores indicating greater perceived stigma. Results for individual questions

were tabulated and are presented in Appendix D.

Scores ranged from 0-5 on each item, and from 0-20 for the full questionnaire.

Thus, scores for this sample of adolescents were distributed across the entire range of

possible values. Sample mean was 6.58 (SD = 5.25) and Cronbach's alpha correlation

coefficient was .92, indicating high internal consistency. Seventeen percent of the

adolescents had a total score of 0 while 83% of adolescents reported some experience of

stigmatization in their social interactions, i.e., they had a total score of one or greater.

However, while the vast majority of adolescents perceived some negative impact, many

participants did not report any belief that their facial difference led to rejection or

stigmatization in specific areas of social functioning. For example, 19% had a score of

"0" on the question asking whether they thought their facial difference affected whether

others wanted to be friends with them; 25% responded negatively to the question as to

whether they thought it affected whether people liked them; 34% felt their facial

difference did not affect their inclusion in social events; and 30% reported feeling their

appearance had no negative effects upon dating and intimate relationships.

For descriptive purposes, adolescents were divided, based upon their total scores,

into three categories according to whether they reported None Low (0-6), Moderate (7-

13), or High (14-20) perceptions of stigmatization. Forty-six (55.4%) of this sample

reported a Low level of perceived stigma, 29 (34.9%) reported experiencing a Moderate

level of stigmatization, and 8 (9.6%) of the adolescents reporting feeling a High level. In

sum, these results suggest significant variability in perceptions of stigmatization among

this sample of adolescents, with virtually all reporting some negative experiences and

almost 50% perceiving moderate to high levels of social rejection and isolation secondary

to their craniofacial condition. As noted previously, gender, age, diagnostic category,

number of functional impairments, and racial/ethnic identity were not significantly

related to perceptions of stigmatization.

Social adjustment

Although raw scores were used in all statistical analyses, standardized scores and

behavioral levels were calculated in order to describe and compare the social skills levels

of craniofacial participants. Standardized mean scores (M = 102.98, SD = 14.97) for

adolescent self-reported social skills (SSRS-Student Social Skills scale) were not

significantly different from the normative sample standard scores (M = 100, SD = 15), t

(82) = 1.79, = .08. Cronbach's alpha coefficient was .82 and also compares favorably

to the normative sample coefficient alpha of .83. Raw scores were also translated into

descriptive Behavioral Levels, derived from cut-off points based upon the SSRS

standardization sample. Raw scores within one standard deviation of the sample mean

are considered to be in the Average range, those above one standard deviation are labeled

More, and those one standard deviation below the mean are labeled Fewer.

Classification of the craniofacial adolescents into these Social Skills Behavioral Levels

resulted in 12 (14.5%) classified as displaying Fewer social skills, 55 (66.3%) as

displaying an Average level, and 15 (18.3%) as exhibiting More social skills than the

average for the standardization group.

The number of different peers mentioned across the 12 activities listed on the PIR

was calculated to estimate the adolescent's peer network. Mean for this variable was

5.31 (S = 4.87), and scores ranged from 0 to 28.

Parent/Guardian Measures

Attributional style

Means and standard deviations for composite positive, composite negative, and

overall attributional style scores on the Attributional Style Questionnaire (ASQ) are

presented in Table 5. Differences between craniofacial parent means and the

standardization sample's means on the ASQ were not statistically significant [(Positive -

M = 5.25, SD =.62, t (79) = 1.09, p = .28); (Negative M = 4.12, SD = .64, t (79)= .097,

p =.92); (Overall M = 1.14, t (79)= .74, p =.46)]. Respectable alpha coefficients of.75

and .78 were obtained for the composite positive and negative subscales, respectively.

These compare favorably with standardization coefficient alphas of .75 and .72. for

positive and negative subscales.

Perceived stigma

Total scores on the Parent Social Perception Questionnaire (PSPQ), used to assess

parental perceptions of their adolescent's experience of craniofacial-related

stigmatization, ranged from 0-19 with a sample mean of 5.33 (SD = 5.05). Cronbach's

alpha correlation coefficient for the four-item scale was .96. The format for this

questionnaire exactly paralled that completed by the adolescents, and thus possible scores

ranged from 0-5 for each item and 0-20 for the whole scale. Individual item results are

presented in Appendix E.

Adolescent and parent perceptions of stigmatization, using the total scale scores,

were compared using the Wilcoxon test for matched-subjects. The results indicated a

significant difference, with adolescents reporting higher levels of perceived

stigmatization than their parents: z =-2.13, p<.05. When responses to individual

questions were examined, parents reported less perceived stigma related to their

adolescent's facial difference than the adolescents reported. Twenty-seven percent of

parents thought their child's condition had absolutely no effect on whether others wanted

to be friends with their adolescent, 35% of parents thought it had no effect upon whether

others liked their adolescent, 35% felt it did not affect social invitations and activities,

and 43% believed their adolescent's facial difference would have absolutely no impact on

dating and intimate relationships. Twenty-five percent of parents viewed their

adolescent's facial difference as having absolutely no negative impact on social

interactions (i.e., total score summed over all questions was 0). When parents were

grouped according to amount of stigmatization they perceived their adolescent to

experience, 52 (62.7%) fell into the None-Low level category, 23 (27.7%) reported a


Moderate level, and 8 (9.6%) reported perceiving a High level. Thus, overall, just 37%

of parents saw their children as being moderately or highly stigmatized in social

situations as a result of their facial difference.

Behavioral and social adjustment

The mean of the craniofacial parents' SSRS-Parent Social Skills Scale

Standardized scores (M = 98.62, SD = 14.94) did not differ significantly from the

standard scores of the normative sample (M = 100; SD = 15), t (81) = -.835, = .41.

When classified according to descriptive behavioral levels, 17 (21.00%) parents reported

that their adolescents exhibit Fewer social skills than the standardization group, 52 (63%)

reported an Average level, and 13 (16%) viewed their children as having More social

skills. Internal reliabilities for the craniofacial sample and the standardization sample

were .91 and .90, respectively

Parent reported externalizing and internalizing behavior problems were also

assessed and a total behavior problem score calculated. Cronbach's coefficient alphas

were .78 for the externalizing scale, .76 for the internalizing scale, and .82 for the total

behavior problem scale. These compare favorably with those from the standardization

sample, which were .82, .72, and .81, respectively. There is no statistically significant

difference between the obtained mean of 102.84 (SD = 14.20) on the Total Behavioral

Problems Scale and the normative sample mean of 100 (SD = 15), t(80) = 1.8, = .08;

standardized scores are not calculated for subscales. Behavioral Level placements for the

externalizing and internalizing subscales and the total behavioral problems scale are

shown in Table 6.

Table 6

Behavioral Problem Levels for Exteralizine. Interalizin. and Total Behavioral
Problems (N= 83)

Exteralizing Internalizing Total
Behavioral Level n % n % n %

Fewer 7 8 9 11 6 7

Average 64 77 62 75 57 69

More 10 12 10 12 18 22

Group Comparisons

Institution, diagnostic category, and gender comparisons were conducted on

independent variable means to determine if there were significant within group

differences that would impact subsequent analyses and interpretation of results. For

variables with normal distributions, either t-tests or ANOVAs were performed. Mann-

Whitney U and Kruskal-Wallis independent sample tests were performed on nonnormally

distributed variables. With the exception of SES, no significant differences were found

on any of the variables used in tests of the hypotheses. Thus, the data were pooled across

these distinctions. As noted previously with regard to SES, there was a significant

difference in SES between those adolescents from UF and those from USF. Hollingshead

(1975) mean scores were 44.57 (SD = 12.07) for UF families, 34.66 (SD = 11.29) for

USF families, and 38.56 (SD = 9.24) for families from Nemours. Furthermore, the only

other significant institution, diagnostic, or gender group difference found among all the

measures employed was on the SSRS-Parent Total Social Skills Scale, where

standardized scores for males (M = 101.23, SD = 13.27) were significantly higher than

scores for females (M = 93.86; SD = 16.79), t (80) = 2.18, E = .03).

Additionally, scores on independent and dependent variables were compared on

the basis of participant's ethnicity and whether they had ever been diagnosed with a

learning disability. The former was done to evaluate possible confounds due to racial or

ethnically based experiences of prejudice and stigmatization. Due to the relatively small

percentage (15%) of ethnic minority adolescents, participants were divided into only two

groups Majority group adolescents (Caucasian) and Minority group adolescents

(African-American, Hispanic-American, and Other). Using these two groups, there were

no statistically significant differences on age, number of functional impairments, SES,

perceptions of stigmatization, attributional style, or any other independent or dependent

variables with the exception of the PSSQ Externalizing Behavior Problem Scale and the

related PSSQ Total Behavior Problem Scale. On the PSSQ Externalizing Behavior

Problem Scale, Majority group adolescents scored significantly higher than Minority

group adolescents (Majority M = 4.76; Minority M = 2.7; Externalizing Mann-Whitney

U: z = -2.91, p=.004). Majority group adolescents' Total Behavior Problem Standard

Scores (M = 104.44, SD = 14.8) were also significantly higher than Minority group

adolescents' standard scores (M = 95.8; SD = 8.7), t(79) = 2.99, 2=.005. Of interest,

given the negative correlation between SES and externalizing behavior problems in the

sample, is the fact that families of Minority adolescents had a higher average SES than

Majority adolescents, though the difference was not significant. Adolescent rated self-

concept was also higher among Minority than Majority participants while the reverse was

true for self-ratings of appearance.

The posthoc decision to examine the relationship between learning disabilities

(LD) and outcome variables was based on growing evidence for unusually high rates of

learning disabilities among individuals with craniofacial conditions and a growing

awareness of the need to explore the possible impact of learning disabilities on

psychosocial adjustment in this population. There were no significant differences among

this sample's adolescents with and without LD in terms of age, SES, overall attributional

style, perceptions of stigmatization, and self-ratings of appearance and self-concept.

However, significant differences between participants with and without learning

disabilities were observed on ratings of social skills, internalizing behavior problems, and

attributional style for negative events. Adolescents with learning disabilities rated

themselves as having fewer social skills than did adolescents without learning disabilities,

t(80) = -3.39, p=.001. Parents also rated their adolescents with LD as having fewer

social skills, t (1,80) = -3.02, p=.003, and as having more internalizing problems than

their nonleaming disabled peers (Mann-Whitney U z=-1.93; p=.05). In addition, the

adolescents with LD had significantly higher CASQ scores for negative events than did

those without learning disabilities, indicating that they have a greater tendency to

attribute bad outcomes to internal, stable, and global causes (Mann-Whitney U z = -2.60;

S= .009).

Intercorrelations of Variables

A correlation matrix including each of the dependent variables, demographic

characteristics, and other selected variables of interest is presented in Table 7. Pearson

correlations were computed for normally distributed variables and Spearman correlations

were computed for nonnormally distributed variables. As a large number of correlations

were computed, many of these associations may be due to chance. However, they are

presented as an aid in interpreting the regression analyses and understanding patterns of

association among variables.

There were few significant associations between adolescent participant

characteristics and other variables. Among the variables of age, gender and SES, only

the correlation between gender and SES was significant (r = -.22, p = .05). Males in the

sample were from families with higher mean SES levels than were females. Higher SES

was also associated with fewer number of functional impairments (r = -.29, p = .009),

lower adolescent perceptions of stigmatization (r = -.34, p = .002), and reduced

externalizing behavior problems (r = -.26, p = .016) as well as with higher levels of

parent reported social skills (r = .30, p =.007). Increased number of functional

impairments was negatively correlated with adolescent-rated social skills (Q = -.26, 1 =

.02). However, and contrary to expectations from previous literature, there were no

significant associations between number of functional impairments and adolescent

perceptions of stigmatization (r = .00, p = .99), adolescent self-reported physical

appearance (r = -.08, p = .55) and self-concept (r = -.09, p = .52), or parent reported

internalizing (r= .21, = .06), externalizing (r = .16, g = .14), or total behavior problems

(r= .20, = .08).

In contrast, adolescent perceptions of stigmatization related to their facial

difference were significantly and highly correlated with a large number of variables.

Higher ratings of perceived stigmatization were associated with higher levels of

externalizing (r = .28, p =.01) and internalizing (r = .34, p = .002) behavior problems and



J0 I
g o .


0 O
lit 0
0 CO
S g *s 8r o o *&

i3 8E te

0 0
o en &, -CO

| g

0 0 0 0 o
0 0

Ci)| ; o 0 0 0

rrB U Sfo v v

C I v
*P 0

E l i i |01
I .0 1
0CC 0 .s s c
0a C l
-~ q

CC 0 L a ~ 4
0CC ~ a 0 --
~ o l 0

o *.t 0 -
o 0 -

3 uO

m- 0


S ^M



0O .



10 0


o '3
- 0

*fl 00

" m

4. *o
0' '0 C


4. 4r 4

'0 -

i o m r-
* 0

00 00
m 3

* '

i -

N d

2 S

'Y N '0
00~ ~ 0 F le 0

00 00 00i0 rr

** *
- 00
w in I

o >4>

0 m

sj- ^

'4- *'4

0 *

'4" &

|,a~ ; a |
M~~a t "4

I 1 ^
o N 00
4.u 4. '4 !
-) h*- = M. 4


c a e
I. 0,

ra *

r a
c 0 .a~_ *
00) 0U 0 T ^ ^ e
S 00 C *f 4*40

>' j ^a c* .o ,. *

g SS c ^ < r ^ *-
0 34 S0

rt~ ~ Q^rQ c ~ ^ Mg M& -s^ of B+ oi
004' CC N/ S' V) WQi fl |
00 u0 a ~"
a s 4)

6's' 4.l 5'l c
'4 m Im 4*a 4* ~

lower ratings of physical appearance (r = -.63, p =.000), self-concept (r = -.50, 2 =.000),

and social skills (r = -.34, p = .002 and -.40, p = .000, for adolescents and parents,

respectively). Greater perceived stigmatization was also significantly related to a less

optimistic attributional style (r = -.47, p = .000), as higher scores on the CASQ Overall

scale reflect a more positive or optimistic approach. The association between parent and

adolescent perceptions of stigmatization was also highly significant r = .53, P = .000).

Furthermore, correlations between parent reports of perceived stigmatization and other

variables are similar to adolescent perceptions.

To ascertain those psychological and interpersonal variables that are related to

attributional style, correlations were calculated using not only the composite attributional

style score but also the attributional style scores for both positive and negative events. To

control for the family-wise error rate, a Bonferroni adjustment was made (.05/13=.004);

correlations were considered statistically significant only ifp < .004. These results are

given in Table 8.

None of the demographic or condition related variables were significantly

correlated with the adolescents' attributional style for either positive or negative events,

or with their composite attributional style. However, as with adolescent perceptions of

stigmatization, attributional style was significantly and highly correlated with a number

ofpsychosocial variables. Higher scores on both the positive events and composite

scales are generally associated with more adaptive attributional styles. Thus, more

positive optimistic attributions were positively correlated with self-reported physical

appearance, self-concept, and social skills, and negatively correlated with perceptions of

stigmatization and parent-reported internalizing behavior problems. A more pessimistic

Table 8

Correlations of CASO Scales and Variables

# Functional Impairments
# Hospitalizations
# Surgeries
Social Perceptions
SDQ Physical Appearance
SDQ General Self
CSSQ Social Skills
PSSQ Externalizing
PSSQ Internalizing
PIR Peer Network
Spearman's correlations.
** L .004



CASQ Attributional Style


attributional style was significantly associated with lower ratings of physical appearance,

self-concept, and social skills as well as higher levels of internalizing behaviors.

Correlations between the three attributional style variables and those of adolescent

externalizing behavior problems and peer network were not significant.

As with perceptions of stigmatization, composite adolescent attributional style

was significantly and positively correlated with parental composite attributional style ( =

.27; p< .05). However, in contrast to parent perceptions of stigmatization, none of the

other adolescent psychosocial variables were significantly correlated with parental

attributional style.

The correlation between adolescent self-reported physical appearance and general

self concept was highly significant (r = .62, =.000). Higher self-rated physical




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd