Psychosocial adjustment of children with craniofacial anomalies


Material Information

Psychosocial adjustment of children with craniofacial anomalies effects of facial appearance, speech quality, and hearing level
Alternate title:
Effects of facial appearance, speech quality, and hearing level
Physical Description:
vii, 128 leaves : ; 29 cm.
Williamson, Linda Marion Graves
Publication Date:


Subjects / Keywords:
Research   ( mesh )
Craniofacial Abnormalities -- Child   ( mesh )
Craniofacial Abnormalities -- Infant   ( mesh )
Craniofacial Abnormalities -- psychology   ( mesh )
Social Adjustment -- Child   ( mesh )
Social Adjustment -- Infant   ( mesh )
Adaptation, Psychological -- Child   ( mesh )
Adaptation, Psychological -- Infant   ( mesh )
Beauty   ( mesh )
Speech   ( mesh )
Hearing   ( mesh )
Psychological Tests   ( mesh )
Data Collection   ( mesh )
Questionnaires   ( mesh )
Department of Clinical and Health Psychology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Health Related Professions -- Department of Clinical and Health Psychology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1996.
Bibliography: leaves 119-127.
Statement of Responsibility:
by Linda Marion Graves Williamson.
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 - 002299963
oclc - 49819879
notis - ALQ3234
System ID:

This item is only available as the following downloads:

Full Text








Perhaps the greatest pleasure in conducting this research was discovering the

kindness of so many people who donated their time and efforts to make it possible. I

cannot adequately thank them all.

I give my thanks first of all to the children and parents who were subjects in this

study and who were the inspiration for it. My appreciation also goes to the staffs of the

craniofacial clinic at the University of Florida and the cleft-palate clinic at All Children's

Hospital, who were so gracious in allowing me to collect data during their clinics.

Thanks are given to Dr. William Williams, Dr. Mary Pavan, and Dr. Barry Pendry for

providing access to subjects. Ms. Virginia Dixon-Wood and Ms. Betty Graves were

exceptionally helpful and kind in many ways. Thanks go to Mrs. Joanne Clarie for her

enthusiastic assistance and to the staff and teachers who welcomed me into their

schools. Dr. Brent Seagle, Dr. Bob Rudman, Dr. Clara Turner, Ms.Virginia Dixon-

Wood, Dr. Anastasia Remer, and Dr. Lynn Maher cheerfully donated their time and

expertise to provide ratings of speech and attractiveness. Martin Ward assisted with

data collection. My appreciation is given to Dr. Michael Robinson, Dr. William

Williams, Dr. Jacquelin Goldman, and Dr. Sheila Eyberg for their guidance and

suggestions. Special thanks go to Dr. Stephen Boggs, who introduced me to pediatric

psychology, has been a role model, and provided many hours of long-distance research

consultation. This project would not have been possible without a tremendous amount

of entirely unselfish help from Ms. Jeanne Herrmann and the late Fritz Herrmann.

My appreciation also goes to Mrs. Barbara Williamson and Ms. Nancy Graves for

watching over my son when research called. My sincerest gratitude is given to my

parents for supporting and encouraging me in every possible way. Highest thanks go

to Dr. David Williamson, for being part of all that I do, and for being my computer

ambassador. Finally, thanks are given to Cameron Williamson and "the kids," for all

your love and patience.



ACKNOWLEDGMENTS......................................... ii

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


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

2 REVIEW OF THE LITERATURE......................... ........ ............ 3

Attractiveness Literature.................... ................................... 3
Speech Impairment Literature.............................. ................ 17
Hearing Impairment Literature......................... ......................20
Craniofacial Anomalies Literature ..........................................28
Overview of Findings ........................................ ......... 50
A M odel of Adjustment............................................. 52

3 OBJECTIVES .............................. ...... ......................... 59

4 HYPOTHESES................... ....... .....................61

5 METHOD ......................... ....................................... 62

Subjects and Settings ...................... .......................62
Measures ... ......................................... 65

6 RESULTS................... ................................ 70

Description of Sample ................................... ...........70
Gender Comparisons ................. ............................. .... 74
Group Comparisons ......................................... 74
Intercorrelations of Variables .............................................79
Multiple Regression Analyses .................................................84

7 DISCUSSION............................ ................................96

Directions for Future Research ..................................... ...... 109
Conclusion ........................................... ............. 112


A SPEECH VISUAL ANALOG SCALE....................................... 115
B ATTRACTIVENESS VISUAL ANALOG SCALE.......................... 116
C MEDICAL EXPERIENCES QUESTIONNAIRE ........................... 117


REFERENCES............................................... .......................... 119

BIOGRAPHICAL SKETCH.............................................................. 128

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



Linda Marion Graves Williamson

May, 1996

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

The relationship of attractiveness, speech quality, and hearing level to

psychosocial adjustment was examined in 63 children with craniofacial anomalies (ages

7 years, 6 months to 13 years, 11 months). These variables were not found to be

related to measures of internalizing (Child Behavior Checklist internalizing scale, Child

Depression Inventory, Revised Children's Manifest Anxiety Scale), externalizing

(Child Behavior Checklist externalizing scale), or self-esteem (Harter Self-Perception

Profile for Children). Furthermore, subjects scored in the normal range on these

measures and were not different from a comparison group of 65 children without

anomalies. These findings are consistent with previous research, which demonstrates

essentially normal adjustment in children with craniofacial anomalies, but are contrary

to expectations of increased adjustment problems based on literature which describes

the relationships of speech, attractiveness, and hearing to social perceptions and

interactions. Possible explanations for the normal adjustment of children with

craniofacial anomalies are described, with suggestions for future research emphasizing

the strengths of these children.


When a child is born with a craniofacial anomaly, parents are frequently

concerned about the effect that the anomaly might have on the child's future

psychological well-being. This concern is often shared by the child's treatment team,

who may base decisions on the timing and extent of surgical corrections on beliefs

about the effects of the child's appearance on his or her psychosocial adjustment

Research literature on appearance suggests that these concerns may be well-founded.

There is abundant evidence that physical appearance has an effect on the way

individuals are perceived and treated by others (e.g. Lefebvre & Ardt, 1988; Eagly,

Ashmore, Makhijani, & Longo, 1991). On the basis of this literature, it might be

expected that children with craniofacial anomalies are at increased risk for poor

psychological adjustment because of their appearance.

Research with children with craniofacial anomalies, however, tends to show

only minimal psychosocial problems, if any (for a review, see Richman & Eliason,

1982). Methodological problems in the studies may account for this pattern of results.

For example, even though unattractive facial appearance is presumed to be a cause of

the children's hypothesized problems, many studies do not include measures of

physical appearance (Bjornsson & Augustsdottir, 1982; Broder & Strauss, 1989;

Kapp, 1979; Kapp-Simon, 1986; Schneiderman, & Auer, 1984). Children with cleft

lip and palate are most often the subjects of such studies. With good surgical repair,

these children are near-normal in appearance; therefore, they probably do not receive

the same social treatment to which more disfigured children might be subjected. In

addition, many children with craniofacial anomalies have associated speech and hearing

impairments. Only a few studies, however (e.g. Richman, 1983; Glass & Starr,

1979), have examined the effects of impaired speech and hearing, despite the fact that

research has shown that speech and hearing impairments impact children's

psychosocial development (Davis, Elfenbein, Schum & Bentler, 1986; Lass, Ruscello,

& Lakawicz, 1988). This paper will review literature on attractiveness and the

psychosocial effects of speech and hearing impairments, as well as summarize existing

research with children with craniofacial anomalies, before describing a research

program which will attempt to more adequately assess the effects of craniofacial

anomalies on children's psychosocial adjustment.


Attractiveness Literature

Influence of Attractiveness on Perception of Others

"Beauty is only skin deep," proclaims the familiar adage, yet research suggests

that we do not really believe this. Since the early 1970s, numerous studies have shown

that people do make character attributions on the basis of physical appearance and that

positive characteristics are more often attributed to attractive than unattractive people.

In one of the earliest studies of the influence of physical attractiveness on

impression formation (Miller, 1970), 720 college students rated yearbook photographs

of other college students on a bipolar adjective checklist. The attractiveness of the

photographs had been previously rated by an independent sample of college students on

a nine-point attractiveness scale. For eight of the 17 items, raters of both sexes rated

attractive photographs of both sexes significantly more positively (p< .01) on the

bipolar adjective scale than they did the less attractive photographs. For seven of the

remaining variables, raters of at least one sex rated more attractive photographs of at

least one sex more positively than less attractive photographs, but the difference was

not significant for all rater-photograph combinations. On two items, differences

between groups were insignificant at the p < .01 level. This study provided some of

the first evidence that, in the absence of other information, attractive individuals are

perceived more positively than unattractive individuals.

In 1972, Dion, Berscheid, and Walster published a study in which they

concluded "what is beautiful is good." Sixty college students rated three photographs

each from a set of 12 college yearbook pictures which had previously been rated as

attractive, average, or unattractive. Subjects rated the photographs on bipolar adjective

scales and ranked them on personality traits by indicating which photographed person

had the least or most of each trait. In a separate task, subjects indicated which

photographed individuals were most and least likely to "ever be divorced," "be a good

parent," "experience deep personal fulfillment." Finally, subjects indicated which

photographed students were most and least likely to obtain employment in thirty jobs

which had been chosen to represent high, average, and low status. Subjects rated the

attractive individuals higher on socially desirable traits than they did unattractive

individuals. They also rated them as more likely to obtain high status employment.

Finally, attractive individuals were rated as less likely "to ever be divorced" and more

likely "to experience deep personal fulfillment" than unattractive individuals. Attractive

individuals were nt rated as more likely "to be a good parent." There were no

differences in any ratings on the basis of sex of either the rater or the photographed

individual. In summary, on every measure except for predicted quality of parenting,

attractive individuals were assumed to have an advantage over their less attractive peers

in terms of both the socially desirable traits they possessed and the personal and

occupational fulfillment they were expected to obtain.

A limitation of the Dion, Berscheid, and Walster (1972) study and the Miller

(1970) study is that they were based on college students' rating of photographs from

college yearbooks. The generalizability of the findings to other age groups or SES

groups is thus unknown. In addition, facial expression, clothing, and hairstyle of the

college photographs may have influenced both the attractiveness ratings and the

attributional ratings. For example, the same person may be perceived as both more

attractive and friendlier when smiling than when frowning. Despite these limitations,

the studies do indicate that college students perceive strangers who present an attractive

appearance as possessing more positive traits than individuals whose appearance is


Gross and Crofton (1977) examined a corollary of "what is beautiful is good"

-- "what is good is beautiful" -- and found it to be true. They presented each of 125

college students with a photograph of a female who had previously been rated as either

high, average or low in attractiveness. The photograph was paired with either a

favorable, average or unfavorable personality description. Attractiveness ratings made

by the subjects were positively related to the favorableness of the personality

description paired with the photo. This effect was particularly pronounced for the

photos which had previously been rated as unattractive. This study demonstrated that,

not only are individuals who are attractive assumed to have other positive attributes, but

individuals known to have positive personality attributes are considered to be more


More recently, Gillen and Sherman (1980) examined the relationship between

attractiveness and gender in the formation of trait attributions. Ten male and ten female

college students rated yearbook photographs of two individuals of each sex at each of

three levels of attractiveness (high, average, low). There were no significant

differences in scores by male versus female raters. Attractive individuals of both sexes

were rated higher on the "good" traits of "friendly," "honest," and "helpful," and were

rated lower on the "bad" of traits "irresponsible," "lazy," and "daydreamer". Attractive

individuals of both sexes were also rated higher on traits that were categorized as either

"masculine-good" or "feminine-good" and lower on traits categorized as "feminine-

bad" than unattractive individuals. The "masculine-bad" traits of "tactless,"

"insincere," and "deceitful" were not significantly related to attractiveness for either

sex. In addition, attractive individuals of both sexes were rated higher on "non-

evaluative masculine" traits, but females were rated equally across attractiveness groups

on "non-evaluative feminine" traits, and attractive males were actually rated lower on

"non-evaluative feminine traits" than unattractive males. On closer inspection, the

"non-evaluative feminine" traits of "emotional," "feelings easily hurt," and "needs

security" may have the potential for more negative connotations than the "non-

evaluative masculine" traits of "dominant," "independent," "direct". It may be that the

"non-evaluative feminine" traits did carry negative connotations, but only when applied

to males, or that they are negative for both sexes, but that the female stereotype is so

strong for these three traits as to override the effects of attractiveness. These results

suggest that, in general, attractive individuals are perceived more positively than their

unattractive counterparts, but that the effect of attractiveness on impression formation is

influenced by the sex of the individual being judged. This is congruent with

Ashmore's (1981, in Eagly et al., 1991) identification of positive intellectual qualities

and negative social qualities that are attributed to males and negative intellectual qualities

and positive social qualities that are attributed to females. Both gender of the subject

being rated and masculine/feminine quality of a trait therefore should be considered

when interpreting associations between attractiveness and trait attributions.

For certain character attributions, attractive people may not have the advantage.

Two-hundred eighteen male freshmen from two universities were given biographical

information forms paired with a picture of either an attractive female or an unattractive

female (Tanke, 1982). Although they rated attractive female students higher than

unattractive females on items measuring social and sexual excitement, unattractive

females were rated higher on items measuring modesty, concern for others, and

interpersonal sensitivity. Thus, for young adult females rated by male peers, the

"beauty is good" stereotype may not necessarily translate to "beauty is kind."

Agnew and Thompson (1994) also demonstrated that being attractive may have

both positive and negative consequences for how one is perceived by others. One

hundred eighty undergraduate students read a fictitious description of a male recent

college graduate who had been diagnosed as HIV-positive. All students read the same

vignette, except that the attractiveness of the man was varied. He was described as

either "very attractive," neither attractive nor unattractive," or "very unattractive."

Students who read the attractive description rated him more positively on a number of

personality dimensions than did those who read the very unattractive description. They

also rated him as more likely to have contracted the virus through multiple heterosexual

contacts, whereas the unattractive man was rated more likely to have contracted the

virus through a single homosexual partner. This may be interpreted as another positive

attribution of attractiveness, since students also rated homosexual contact as being a

more negative means of contracting the virus than heterosexual contact. A negative

effect of attractiveness, however, was demonstrated by the fact that the attractive man

was judged to be more to blame for his condition than the unattractive man. (Ratings

for the neutral attractiveness condition were between the attractive and unattractive

conditions for most items.)

Eagly, Ashmore, Makhijani, and Longo (1991) conducted a meta-analysis of

research on physical attractiveness stereotyping. They selected studies from

psychological, sociological, medical, and educational data bases extending from 1963-

1987. From an initial pool of 600 reports on physical attractiveness, they selected 76

studies which met the following criteria: 1) General physical attractiveness (as opposed

to a specific physical characteristic such as weight or body type) was an independent

variable. 2) Ratings of physical attractiveness were not made by the same individuals

who rated the targets' other characteristics. 3) Personality traits or life outcomes which

could clearly be defined as either positive or negative were used as at least one of the

dependent variables. 4) Subjects rated targets with whom they were not previously

acquainted and were not asked to assume a hypothetical relationship to the target when

making the rating. 5) Targets were not described as having specific characteristics or

experiences (such as mental illness, being sexually active before marriage, having been

raped) which might affect subjects' ratings of them. 6) Subjects and targets were over

the age of 14. 7) Sufficient information was provided to permit computation of at least

one effect size. Effect sizes were computed for eight attribution categories: "social

competence," "adjustment," "potency," "intellectual competence," "integrity," "concern

for others," "general evaluation," and "other" (specific attributes which could be

characterized as positive or negative and did not fit the other categories). Effect sizes

for the variables "integrity" and "concern for others" were very small. The variables

"adjustment," "potency," "intellectual competence," "general evaluation," and "other"

had mean weighted effect sizes (ds) ranging from 0.46-0.57. (Values of d close to

0.50 are classified as "medium" and have been described as being characteristic of

phenomena which would be noticeable in everyday situations (Cohen, 1977, in Eagly,

et al., 1991 ). The variable "social competence" had the greatest d (0.68). (Ad of

0.80 is considered "large" [Cohen, 1977, in Eagly, et al, 1991].) The results of this

analysis indicate that physical attractiveness does indeed exert a noticeable effect on

perceptions of a stranger's personality characteristics. Attractive people are perceived

as being better adjusted, more powerful, smarter, and generally having more positive

characteristics than unattractive people. There is a particularly strong tendency to

perceive them as socially competent. Attributions of integrity and concern for others,

however, are not related to ratings of physical attractiveness. This suggests that, for

adults evaluating non-acquaintances, physical attractiveness is perceived as being

positively related to characteristics that benefit the attractive person (intelligence, social

competence, power), but unrelated to the manner in which the attractive person will

treat others (with or without concern or integrity). The analysis does not address the

effects of attractiveness on the perceptions of individuals who are acquainted with each

other or young children's attractiveness stereotypes.

In general, these studies indicate that attractive individuals are perceived more

positively than unattractive individuals. Important exceptions to this are the findings

that attractive individuals are not perceived as having more integrity and concern for

others than unattractive individuals and that they may be seen as more to blame for

certain of their problems. In addition, when a characteristic is strongly associated with

one gender, there may be an interaction between sex and attractiveness in determining

whether a person will be perceived as possessing that characteristic.

Attractiveness and Social Interaction

Given that attractiveness has been shown to affect the ways in which

individuals are perceived by others, the question remains of whether attractiveness

affects the treatment people receive. Research indicates that it has a pervasive effect on

social interaction. It has been shown to influence preschool children's preferences for

playmates, dating choices of high school and university students, and even whether a

child will be punished for a misdeed or a criminal convicted for a crime (Lefebvre &

Arndt, 1988).

Walster, Aronson, and Abrahams (1966) provided some of the earliest evidence

of the effects of attractiveness on social interaction in their study of dating behavior.

Among 376 randomly computer-matched couples who evaluated their blind dates at a

freshman college dance, attractiveness was by far the strongest predictor of liking (for

male raters, r= .78, for female raters, r= .69). It was also the strongest predictor of

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

(usually male) who was asking for the date. In contrast, intelligence and achievement

were not significantly correlated with liking of the partner, and personality variables

were correlated only weakly with liking.

Goldman and Lewis (1977) examined whether attractive individuals would

maintain their advantage in social interactions when their attractiveness was hidden

from their interaction partner. They had 120 college students rate the social skills and

likability of opposite-sex partners whom they had never seen, after a five minute

telephone conversation. Even when the individual's attractiveness was not visible to

their partner, more attractive individuals received higher ratings of social skills and

likability. Goldman and Lewis offered the explanation that attractive individuals may

be socialized differently in their daily interactions. They may receive more positive

attention, which facilitates the development of social skills that increase the individual's

likeability beyond the effects of attractiveness alone.

Attractiveness has also been shown to affect the process of trial by jury. Efran

(1974) provided 66 college students with a simulated case description of a student who

was being tried by a student-faculty court on the charge of cheating on an examination.

Attached to the description was either a photograph of an individual who had been

previously rated as attractive, a photograph of an unattractive individual, or no

photograph. All descriptions were identical, and all subjects were given photographs

of opposite-sex individuals. Efran found that male subjects indicated less certainty of

guilt, recommended less severe punishment, and rated themselves as more attracted to

the attractive "defendants" than either the unattractive ones or the defendants who were

not pictured. The effects for female "jurors" were nonsignificant, but in the same

direction, for all three ratings. When the "jurors" rated the photographs for

attractiveness, there was less difference between the ratings given by female judges to

attractive versus unattractive photographs than by male judges. It is therefore difficult

to determine whether men are more influenced by attractiveness of opposite-sex

defendants than are women or whether attractiveness was not manipulated effectively

for the female raters. In addition, this study addressed judgments of only one type of

offense. Nonetheless, it provides evidence that the effects of attractiveness on an

individual's treatment by others extends even to judgments of guilt and

recommendation of sentences.

Infants as young as four months of age demonstrate a preference for attractive

faces, suggesting that the effects of attractiveness on social interaction may begin at a

very early age. Samuels, Butterworth, Roberts, Graupner, and Hole (1994) presented

attractive and unattractive photographs of young adult females to 25 infants (ages 4-15

months, mean age 7.94 months). In addition, they presented photographs which had

been altered by computer to eliminate the slight asymmetry which is present in normal

faces. The infants spent more time looking at attractive versus unattractive faces, but

did not demonstrate a preference for either the natural photographs or those which were

altered to produce symmetrical facial images. This suggests that attractiveness affects

the behavior even of very young infants. Furthermore, it appears that infants

perceptions of attractiveness are consistent with those of adults and that this is not

merely the result of a bias towards facial symmetry.

Dion and Berscheid (1974) demonstrated that, by the preschool years,

children's preferences for attractive individuals are evident in their choice of friends.

Peer nominations were used in four nursery school groups of 77 children (ages 4-6) to

rate the children on a number of social variables. Attractiveness of the children was

rated by adult judges. For children over the age of 5 years 4 months, attractive children

of both sexes received higher popularity ratings than unattractive children. For the

younger boys, this relationship was also true, but unattractive girls were more popular

than attractive girls among children aged 4 years 4 months to 5 years 4 months. It is

unclear why unattractive females were more popular among children younger than 5

years 5 months, but it is apparent that by the age of 6, attractive children of both sexes

are more popular than their unattractive peers. In addition, unattractive male children

received more nominations for aggressive behavior and unattractive children of both

sexes received more nominations for being "scary" than attractive children. Attractive

children received more nominations for being independent. This suggests that, in

addition to being related to popularity, attractiveness is associated with specific child

character attributions as early as the preschool years.

Vaughn and Langlois (1983) also examined the relationship between

attractiveness and peer relationships during the preschool years. College students rated

the attractiveness of photographs of fifty-nine preschool children aged 4 years 8 months

through 5 years 7 months. Two sociometric ratings and one measure of attention

received from classmates were obtained for each of the children. Attention received

was not associated with physical attractiveness, but physical attractiveness was

significantly positively correlated (r=.42, Q<.01) with the paired-comparisons

sociometric rating. The association between the picture board sociometric rating and

physical attractiveness showed a similar trend as the paired-comparisons data, but was

nonsignificant. Examining the paired-comparisons data more closely revealed that the

association between physical attractiveness and sociometric rating was due to a strong

association between the two variables for the girls, but that the relationship between the

two variables for boys was nonsignificant. In the Dion and Berscheid (1974) study,

the relationship was positive for males in this age range, but was negative for females.

This may be due to problems in sampling (there were fewer than twenty females in the

young age group in the Dion and Berscheid study) but may also reflect a weaker effect

of attractiveness among children under the age of 5 years 6 months.

There is evidence that the relationship between attractiveness and social

interactions extends beyond the peer group to adult-child interactions. Dion (1972, in

Rich, 1975) found that undergraduate college women rated misbehavior by an attractive

child as less undesirable than the same behavior committed by an unattractive child.

Furthermore, misbehavior by unattractive children was perceived as being part of the

child's personality style, whereas misbehavior by attractive children was seen as an

isolated incident Unattractive children were also rated as more dishonest and

unpleasant than attractive children. Attractiveness rating was not, however, related to

the severity of punishment recommended.

In a study by Rich (1975) attractiveness was related to the severity of

recommended punishment, but there was an unexpected interaction of sex and

attractiveness. Subjects were 144 female teachers who made ratings of blame,

personality, and recommended punishment after reading a description of a hypothetical

third grader who had committed a classroom offense. Either an attractive or

unattractive photograph was attached to each description. Although attractive children

received more positive ratings for personal and academic development, unattractive

girls were perceived as more intelligent than attractive girls, whereas attractive boys

were perceived as more intelligent than unattractive boys and girls of any attractiveness

level. Furthermore, unattractive boys were most often recommended for punishment

and received the most severe punishment recommendations, whereas unattractive girls

were least likely to be recommended for punishment and received the most lenient

punishment recommendations. These results indicate that the sex of the child is an

important variable in determining how the child's attractiveness will effect his or her

interactions with adults. Unfortunately, no male teachers were included in this study,

so it in uncertain whether the same pattern would emerge with male teachers.

Relationship Between Physical Appearance and Psychosocial Adjustment

Because in many circumstances they are perceived and treated less positively

than attractive individuals, it might be predicted that unattractive individuals would have

more psychosocial adjustment problems than attractive individuals. The research which

addresses this hypothesis, however, is not uniformly supportive of it.

Feingold (1992) completed a meta-analysis of 93 published and unpublished

studies which correlated ratings of physical attractiveness with measures of personality,

social behavior, cognitive ability, and sexuality. All subjects were either adolescent or

adult Physical attractiveness was only trivially related to most mental health measures

(average correlations below r = .10), except for loneliness (i =.-15) and lack of social

anxiety with the opposite sex (r =.22). Physical attractiveness was positively related to

social behaviors such as social skills (I = .23) and popularity with the opposite sex

(r= .31). There were only small associations to sexuality measures (average

correlations from r = .01 .18), and cognitive ability was virtually unrelated to

attractiveness (mean = .04, median r = .00). In contrast, in the additional 57 studies

of self-rated attractiveness which Feingold also meta-analyzed, self-rated attractiveness

did have a moderate association to general mental health (_=.24) and self-esteem

(r = .27). This suggests that, among adolescents and adults, self-perception of

attractiveness may be more important to psychological adjustment than physical

appearance as judged by others.

There is some evidence that, for children, self-rated attractiveness may also be

more important than others-rated attractiveness. Kenealy, Frude, and Shaw (1989)

examined the relationship between attractiveness and psychosocial adjustment in 1018

children age 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), and a number of parent and

teacher ratings of behavior were made. In addition, a popularity score was derived

from peer nominations in the child's classroom. There was a strong positive

relationship between child self-ratings of attractiveness and self-esteem scores. Teacher

ratings of attractiveness were positively, but weakly, related to the self-esteem score.

Attractiveness ratings by judges who were unacquainted with the children were

unrelated to self-esteem. Attractiveness ratings by all raters were positively associated

with teacher's ratings of the child's popularity, confidence, leadership, and academic

brightness, as well as with sociometric ratings of the child's popularity. 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 the children's own

perception of their appearance.

In contrast, a study by Cooper (1993) did find an effect of others-rated

physical attractiveness on self-esteem and children. The relationship between

self-esteem and facial attractiveness in 55 learning disabled children ages 8-13 was

examined. Subjects completed the Piers-Harris Children's Self-Concept Inventory and

posed for a facial photograph. Photographs were rated on a nine point Likert scale by

six adult raters and six sixth grade students. Self-esteem was positively related to both

adult and peer ratings of attractiveness in this sample of learning-disabled children. It

should also be noted that 56 percent of the subjects and 50 percent of the raters were

African-American, suggesting that the importance of attractiveness to self-esteem may

generalize across racial lines. Unfortunately, the author did not perform any analyses

for effects of race.

A 1994 study by Appleton et al., found a relationship between self-esteem and

self-perceived attractiveness among individuals with spina bifida. Subjects were 79

children ages 7 years to 18 years, 11 months who had spina bifida. As part of a larger

test battery, subjects were given the Harter Self-Perception Profile for Learning

Disabled Students (HSPPLDS) (Renick & Harter, 1988). The HSPPLDS assesses

nine domains of competence including general intellectual ability; competence in

reading, writing, spelling, and math; social acceptance; athletic competence; behavioral

conduct; and physical appearance. Discrepancy scores for each domain may also be

calculated. These scores take into account both the child's self-perceived competence in

a domain and the importance of that domain to the child. High discrepancy scores

represent low perceived self-competence in areas the child rates as important.

Discrepancy scores for physical appearance had a stronger negative relationship to

global self-worth than any other domain, particularly among females. This study

provides further evidence of the importance of self-perception of physical attractiveness

to general self-esteem.

Two recent studies failed to find enhancement in psychosocial adjustment after

orthodontic treatment to improve dental-facial appearance. Although these may be

interpreted as contradictory evidence against a relationship between attractiveness and

adjustment, the designs of the studies do not necessarily justify such a conclusion.

The first study (Korabik, 1994) did not directly assess attractiveness of the subjects.

Furthermore, as the article notes, the adolescents in the study had higher than average

self-esteem at pre treatment and thus may have failed to demonstrate significant

improvement due to ceiling effects. (The mean score on the Piers-Harris Children's

Self-Concept Scale prior to treatment was at the 79th percentile). Albino, Lawrence,

and Tedesco (1994) compared adolescents who received orthodontic treatment with

waitlist controls and found no group differences in self-esteem or positive social

interactions in spite of higher dental-facial attractiveness ratings (by self, peers, and

parents) obtained by the treatment group. It should be noted, however, that all

subjects were selected on the basis of having been excluded from treatment at the

county clinic due to the fact that their orthodontic conditions were not considered

disabling, and all were judged as having only mild to moderate malocclusions.

Furthermore, their baseline scores on the Rosenberg Self-Image Inventory (Simmons,

Rosenberg, & Rosenberg, 1973) were in the average range. It is possible that changes

in appearance did not lead to enhanced psychosocial adjustment in these children

because they were well-adjusted to begin with and did not have significant physical

appearance concerns. These findings therefore cannot be generalized to children whose

physical appearance is judged to be very unattractive.

In summary, the literature suggests that individuals who are unattractive are

perceived less positively and have less-positive social interactions than their attractive

counterparts. There is contradictory evidence as to whether they may also be less

well-adjusted than attractive peers, and it appears that self-rated attractiveness is more

important than others' judgments of appearance. These findings have important

implications for children with craniofacial anomalies. Physical appearance is not the

only concern of these children, however. Many of them also have speech and hearing

impairments. This review therefore will turn to the psychosocial concomitants of

speech and hearing impairments before examining the literature dealing specifically with

children with craniofacial anomalies.

Perception and Social Interaction of Individuals with Speech Impairments

Like unattractive children, children whose speech differs from that of peers may

be perceived negatively by others. A study by Lass, Ruscello, and Lakawicz (1988)

demonstrated that dysarthric speakers (individuals who have problems with articulation

of speech) are perceived more negatively than children with normal speech on a broad

range of personality characteristics. Eight children (ages 6 to 11) with dysarthric

speech and eight children with normal speech were recorded speaking six three-word

phrases. The tapes were played to 25 college students who had normal speech and

hearing. The students rated each of the speakers on 24 bipolar semantic differential

scales. On 22 of the scales, dysarthric children were rated significantly more negatively

than normal speakers. This suggests that, when little or no information other than

speech quality is available, children with poor speech are evaluated negatively by

others. This was also demonstrated in a later study with identical design which used

13-year-old students as raters (Lass, Ruscello, Harkins & Blankenship, 1993). The 19

adolescent raters in that study rated dysarthric students significantly more negatively

than normal speakers on all of 22 semantic differential scales.

Negative responses to speech impairment are evident even in very young

children. Blood and Hyman (1977) played audiotapes of four girls reading a short

passage to 120 children in kindergarten through second grade. The readers varied in

speech quality from normal to severely hypemasal. After each tape, the children were

asked whether they liked the reader, liked the way the reader talked, thought the reader

had trouble talking, would like to talk to the reader, and thought the reader needed help

with talking. The percentage of negative responses increased with the severity of the

hypernasality. None of the raters indicated that they liked the severely hypemasal

reader. This indicates that hypemasality could have a significant impact on peer

acceptance of an elementary-school child.

Silverman and Paulus (1989) found that high-school aged children who

substitute "w" for "r" are also perceived negatively by their peers. In this study, 26

high school sophomores completed a semantic differential scale based on the following

scenario "A classmate who says /w/ when he means to say /r/. For example, he or she

might say 'wed' when he means to say 'red'." These ratings were compared with

ratings of a "typical classmate" by 22 sophomores from the same school. Speakers

who substitute "w" for "r" were rated "more tense, nervous, afraid, handicapped,

isolated, and uncomfortable, and less employable, friendly, sane, educated, and

confident" than the typical classmate. Although this study is quite unsophisticated, the

results do indicate that students judge peers more negatively when they have even a

minor speech impairment such as substitution of one sound for another.

A weakness shared by all of the preceding studies is the lack of any

information about the speakers other than speech quality. This is an artificial situation

which forces the subject to use the quality of each child's speech as the basis of

judgments about the child. In a naturalistic setting, where other information is

available, the importance of the children's speech quality in determining how they are

perceived by others may decline.

Perrin (1954) found a negative relationship between speech impairment and

popularity of elementary school children. Positive peer nominations were obtained

from 444 children in grades one through six. Thirty-seven of the subjects were speech

impaired. There was a significantly higher percentage of "social isolates" (children

receiving fewer than two nominations) in the speech impairment group than in the

control group (21.6% vs. 13.5%). Furthermore, there were no "stars" (children

receiving 20 or more votes) in the speech impaired group, whereas 6.5% of the control

group were "stars." These results indicate that speech impairment is associated with

decreased popularity among elementary school children. The more recent finding of

Rice, Sell, and Hadley (1991) that preschool children with normal speech and language

receive more peer-initiated interactions than those with speech or language problems is

also consistent with the hypothesis that clear communication skills are important for the

development of positive social interactions. Unfortunately, the 1991 study did not

separate the effects of speech versus language disorder, and the sample size was very

small for the speech impaired group (n=3).

Adjustment of Children with Speech Impairment

Negative peer reactions and social isolation might be expected to produce social

inhibition in a child. Lindholm and Touliatos's (1979) study of 3097 children in

kindergarten through sixth grade provides some support for that expectation, but is not

conclusive. Teachers completed the Quay Behavior Problem Checklist on 2991 control

children and 106 children who had been placed in special education classes for speech

therapy. The speech disordered children scored significantly higher on the "personality

problems" (anxiety and withdrawal) scale of the checklist than children with normal

speech. The difference, however, while statistically significant, may not have been

clinically meaningful (control M = 1.25, SD.= 2.16; speech impaired M = 2.25,

SD= 2.86).

Baker, Cantwell, and Mattison (1980) also examined behavior problems in

children with speech impairments. They examined parent and teacher ratings for

speech impaired children on the Conners' Parent and Teacher Questionnaire (Conners,

1973) and the Rutter Parent and Teacher Questionnaire (Rutter, Graham, & Yule,

1970). The most frequently reported problems were oversensitivity, frustration,

shyness, short attention span, and restlessness. Unfortunately, no normal control

group was used and mean scores were not compared to scale norms, so it uncertain

whether speech-disordered children display a higher incidence of these problems than

the general population. Clearly, more research is needed on the adjustment of children

with speech disorders.

Hearing Impairment Literature

Hearing Impairment: Social Effects

Many children with craniofacial anomalies have related hearing impairments

which may also influence their psychosocial well-being. A study by Silverman and

Klees (1989) indicates that the presence of a visible hearing aid has an impact on peer

perceptions. Forty eleventh and twelfth graders used a semantic differential scale to

rate a photograph of a male peer. Half of the subjects were given a photograph of the

peer wearing a visible hearing aid; the other 20 subjects were given a photograph of the

peer without a visible hearing aid. There were significant differences on 19 of the

scales. In general, the individual with the hearing aid was described as being more

introverted, anxious and depressed, and as having lower self-esteem than the same

individual without the hearing aid. The only positive effect was that the hearing-aid

photograph received higher ratings on maturity. This demonstrates that knowledge of

an adolescent male's hearing impairment is sufficient to negatively influence peer

perceptions of him; it is apparently not necessary for high-school aged peers to directly

experience difficulty communicating with the hearing impaired individual for negative

attributions to be made. Dengerink and Porter (1984), using similar methodology as

Silverman and Klees (1989), found that negative peer perceptions of males with

hearing aids begins at least as early as fifth grade.

When teachers rate female children, a different picture emerges (Cox, Cooper,

& Mc Dade, 1989). Sixty teachers gave lower achievement ratings to girls (ages 10

through 14) who were pictured wearing hearing aids than to girls pictured without

aids. However, they also rated girls pictured with hearing aids as being more attractive

and assertive and as having better personalities than girls without hearing aids.

It is uncertain whether the difference in findings of these two studies is due to

the sex of the child or to the fact that the ratings were made by teachers instead of peers.

Teachers may have consciously avoided giving low ratings to deaf children except in

the area of achievement (where it is known that deaf children have problems) because

they were aware of the dangers of stereotyping. On the other hand, the discrepant

results may reflect a true difference in the effect of hearing impairment on the

personality attributions of male versus female children.

Mother-Child Interaction of Hearing-Impaired Children

It has been hypothesized that psychosocial difficulties of a hearing impaired

child may begin with impaired parent-child interactions due to communication difficulty

between a hearing parent and hearing-impaired child. Wedell-Monnig and Lumley

(1980) analyzed mother-child interaction patterns in six deaf child/hearing mother

dyads and six hearing child/hearing mother dyads. ("Deaf' children had severe to

profound hearing loss.) Children were between 13 and 29 months old. They found

that hearing mothers of hearing impaired children were more dominant (made more

spontaneous attempts to interact ) than mothers of hearing children and that

hearing-impaired children were more passive (inactive) than hearing children when not

being stimulated by their mothers. Furthermore, they noted that mothers of hearing

children interacted more with older children, whereas mothers of hearing-impaired

children interacted less with older children.

Meadow-Orlans and Steinberg (1993) observed hearing mothers' interactions

with their 18-month-old infants. Twenty infants had significant hearing loss and 20

were normal-hearing. Mothers of hearing-impaired infants scored higher on

intrusiveness and lower on flexibility, consistency, and use of positive touch than

mothers of hearing infants. This was true of mothers who reported low levels of social

support, but not of those who reported high social support. Social support was not

related to mothers' interactions with hearing infants. Social support for the mothers of

deaf children was more likely to include help from professionals and the community

than was the case for mothers of hearing children. This suggests that support from

individuals who have experience with the hearing-impaired may be very useful to

hearing parents in learning to communicate with their deaf children. On the other hand,

the deaf children's contributions to the interactions were not rated significantly less

positively than the hearing children's. The researchers offer several explanations for

this. It may be that the hearing mothers were adapting their communication in a manner

which was simply not measured by the researchers but was nonetheless effective, or

that the children may have benefitted from outside influences. (All of the children were

enrolled in early intervention programs).

Henggeler and Cooper (1983), using a sample of older children (3.42- 6 years,

hearing loss of 28 to 80 dB with aids), found that mothers of hearing-impaired children

interacted less with their children than mothers of hearing children, used more direct

commands and were less responsive to their children's questions. Hearing impaired

children were less compliant with direct and indirect commands, less responsive to

questions, and asked fewer questions than hearing children.

Jamieson (1994) compared communication patterns in dyads of either a hearing

mother-hearing child, hearing mother-deaf child, or deaf mother-deaf child (ages 4

years 9 months to 5 years 5 months). Mothers were asked to teach their child how to

build a pyramid of blocks and were observed during this instruction. The deaf children

with hearing mothers made the lowest number of child-initiated interactions.

Furthermore, it was noted that hearing mothers of deaf children tended to repeat verbal

instructions that the child did not understand rather than attempting to shift the

instruction to a visual modality. There were only three dyads in each group, so it is not

known whether these results generalize to the majority of the deaf population.

However, if these findings are replicated with larger groups, they would indicate that

the 90 percent of deaf children who have hearing parents may face problems in

communicating in their own families due to the difficulty of the hearing parent in

adjusting to the child's need for visual input.

It is hypothesized that mothers initially give their hearing-impaired children

extra stimulation to compensate for the hearing loss, but progressively withdraw as an

age-appropriate method of communication fails to develop. The parents are forced to

rely on simple communication, which tends to take the form of commands and

attention-getting, and the child becomes passive under this direction. The children may

also decrease efforts to communicate as their needs and ideas become too complex to

express with the communication system they shares with their parents (Greenberg,

Calderon, & Kusche, 1984). There is some evidence, however, that early intervention

and high levels of maternal social support may alter these interactions in a positive

direction (Meadow-Orlans & Steinberg, 1993).

The pattern described above may be at the root of Stinson's (1978) finding that

mothers of hearing impaired children do not expect hearing impaired children to reach

certain social / emotional milestones as rapidly as other children. Mothers of

profoundly hearing impaired (n = 31) and normal (n = 33) boys ages 8 to 12 completed

a questionnaire on which they indicated the appropriate ages at which mothers should

expect their children to reach milestones in verbal skills acquisition, independence, and

social / emotional development Not surprisingly, they predicted that hearing-impaired

children would develop language at a later age than normal peers. More interestingly,

mothers of hearing-impaired children also recommended that a mother of a

hearing-impaired child "teach her child that crying is not the way to get what he wants,"

"teach her child not to use his fingers when eating," and "teach her child not to cry

every time he gets hurt" at later ages than mothers of normal children recommended for

a normal child. These lowered expectancies for psychosocial development may be due

in part to the difficulty of communicating these type of ideas to a deaf child and to the

parent's need to focus their finite energies on communication of more basic ideas such

as health and safety issues.

Adjustment of Hearing-Impaired Children

A myriad of psychosocial and educational problems have been linked to hearing

disability. The educational achievements and earning potential of deaf individuals are

well below average (Compton, 1993; Northern & Downs, 1991). In addition, some

authors have stated that deaf individuals are more likely to be perceived as immature,

egocentric, impulsive, and as displaying an above-average level of conduct problems

than individuals with normal hearing (Feinstein & Lytle, 1987; Schlesinger & Meadow,

1972; Meadow, in Stein, Mindel, & Jabaley, 1981). Most of these difficulties are

believed to be the result of the deaf individual's limited ability to communicate with the

hearing world. Without hearing, speech development is delayed or nonexistent, so that

the child is not only unable to understand the verbal communication of others, but is

also unable to effectively express himself. Compton and Niemeyer's (1994) review

indicates that hearing-impaired children's exchange of affectionate behavior is also

hindered by their inability to perceive auditory cues of affection (such as tone of voice)

and that their initiations of affection are often rejected by peers with normal hearing.

"Acting-out" immaturely or aggressively may thus be the child's only recourse

(Schlesinger & Meadow, 1972). Poor self-esteem may result when these interactional

difficulties lead to isolation from or rejection by peers and even family (Feinstein &

Lytle, 1987). Furthermore, research indicates that hearing loss well below the level of

profound deafness may have negative developmental consequences. Hearing losses as

low as 15 dB have been associated with language delay (Northern & Downs, 1991),

and hearing losses of less than 40 dB have been associated with psychological

adjustment difficulties (Davis, Elfenbein, Schum, & Bender, 1986).

In an effort to test the assumption that children with disabilities experience greater

psychological and emotional difficulties, Tavormina, Kastner, Slater, and Watt

(1976) gave a battery of psychological tests to 144 children with physical problems,

including 16 hearing impaired children. Subjects were described as having "mild

disability" decibels hearing loss was not reported. The hearing impaired children

scored significantly below scale norms and significantly below the other children in the

study on the Piers-Harris Self-Concept Scale. In addition, they scored as significantly

more external on the Nowicki-Strickland Locus of Control Scale (Nowicki &

Strickland, 1973) and scored higher on aggression and immaturity than scale norms on

the Missouri Children's Picture Series (Sines, Pauker, & Sines, 1971). This

suggests that hearing-impaired children may experience psychosocial adjustment

difficulties related to their condition. Due to the small sample size of 16, however,

these results should be interpreted cautiously.

Raymond and Matson (1989), using a sample of approximately 50 hearing

impaired children (84% were severely or profoundly hearing impaired), found that

hearing impaired children scored within the normal range on the Teacher Report Form

of the Child Behavior Checklist (Edelbrock & Achenbach, 1984), Aggression,

Withdrawal, and Unpopular subscales and on the Inappropriate

Assertiveness/Impulsiveness subscale of the Matson Evaluation of Social Skills with

Youngsters, Teacher Report Form (MESSY) (Matson, Macklin, & Helsel, 1984).

However, a greater than expected number scored below normal limits on the

appropriate Social Skills Subscale of the MESSY. This suggests that hearing impaired

children may be deficient in social skills (or may use skills not measured by the

MESSY), but that they are not grossly impaired. A large proportion of children in the

Raymond study were residents at a school for the deaf, which may have eased their

adjustment by providing them with an environment tailored to their disability and in

which they were not different from other children. This makes it difficult to compare

these results those of Tavormina et al. (1976) or to generalize the Raymond and Matson

findings to children who live at home. In addition, the lower scores on the MESSY

may be attributable to differences in social interactions between deaf versus hearing

children and may be representative of normal functioning within the child's usual social

context Whether mainstreamed deaf children demonstrate appropriate social skills in

interactions with hearing peers is not addressed by this study.

Bat-Chava's (1993) meta-analysis of 42 studies of self-esteem in deaf people

revealed a pattern of lower self-esteem scores for deaf versus hearing individuals.

Bat-Chava notes, however, that in those studies in which measures were administered

in sign language or were modified to be more easily understood by the hearing

impaired, deaf individuals did not receive lower self-esteem ratings. Furthermore,

self-esteem was positively related to use of sign language in the home and to presence

of other deaf people in the child's environment. This underscores the importance of

the social context to the hearing-impaired child's adjustment. Isolation from other deaf

individuals and failure of others to adjust to the child's communication needs may be

partially responsible for frequent descriptions of maladjustment in deafness literature.

Cole and Edelmann (1991) administered the Porteous Checklist (Porteous,

1985) to 51 deaf adolescents (ages 14-16). Subjects scored within the normal range on

seven of nine problems areas addressed by the scale. Only in the area of employment

did the deaf students report significantly more problems than the hearing norms,

although problems in boy/girl relationships also approached significance. The scale

was administered in sign language, again suggesting that the strengths of deaf

individuals may be more apparent when they are observed functioning in situations

which are adjusted to their needs rather than when they are forced to accommodate to

hearing people.

Adverse effects of mild hearing loss on academic as well as psychosocial

development were found in a study of 40 children of normal intelligence between the

ages of 5 and 18 (Davis, Elfenbein, Shum, & Bentler, 1986). The children were

grouped according to pure tone average hearing loss. There were no significant group

differences in IQ, but the combined sample mean verbal IQ was 9 points below

WISC-R population norms, whereas the performance IQ was 5 points above average.

Furthermore, the average verbal-performance difference was significant at the p < .001

level. On the Peabody Picture Vocabulary Test, Revised (Dunn & Dunn, 1981),

children with moderate impairment scored an average of one year below age level,

while severely impaired children scored three years below age level. Reading

achievement was also significantly below normal, although math achievement was not.

Clearly, verbal ability was adversely affected by hearing loss in even mild and

moderately impaired children. Personality measures were also administered. There

were no differences between hearing-impairment groups on any of these measures. On

the Missouri Children's Picture Series, the hearing-impaired mean aggression and

somatizations scores were significantly above the test population mean. On the

Achenbach Child Behavior Checklist, hearing impaired children scored significantly

below the test mean on the social interaction (T = 40.1) and school scales (I = 37.1),

indicating that they had difficulty establishing friendships and interacting with peers,

parents and siblings, and that their academic achievement was below average. They

also scored significantly above the test mean for Externalizing, although the group

mean Externalizing score was not clinically elevated (I = 54.2). These results indicate

that children with even mild-to-moderate hearing impairment are likely to exhibit school

performance problems, difficulty in social interaction, and slightly increased incidence

of externalizing behaviors.

Simon, Larson, and Lehrer (1988) also found a relationship between hearing

loss and social and educational development. Two hundred ninety-one preschool

children ages 3 to 5 were administered brief screening measures for hearing and

cognitive, linguistic, and social development. Hearing loss (classified as present or

absent, with a 25 dB loss cut-point) was associated with decreased cognitive and social

development. The association between hearing loss and both cognitive and social

problems was largely accounted for by the language delay in the hearing-impaired

children. In addition, the relationship between lower social development scores and

hearing impairment was mediated by an association between hearing loss and

socioeconomic status (greater incidence of hearing impairment at lower socioeconomic

status levels).

In summary, hearing-impairment has been associated with decreased verbal

skills and problems in psychosocial adjustment, particularly the development of

excessive externalizing behavior. Literature suggests that individuals with impaired

speech or unattractive appearance may also be at risk for certain psychosocial problems.

Because craniofacial anomalies may be associated with unattractive appearance, hearing

impairment, speech impairment, or any combination of the three, children with these

anomalies may also be at high risk for psychosocial adjustment difficulties. Following

is a review of the literature that focuses specifically on individuals with craniofacial


Craniofacial Anomalies Literature

Perception of Individuals with Facial Anomalies

Research suggests that individuals with facial anomalies are perceived by others

as having more negative personality characteristics than individuals with normal faces.

For example, Tobiasen (1987) presented 317 children with color photographs of

children with clefts of the lip and palate. Copies of the eleven photographs were altered

so that all evidence of the cleft was removed. Subjects were shown pictures of normal

children, plus either corrected or uncorrected versions of the children with clefts.

Children rated each of the photographs on friendliness, popularity, intelligence,

attractiveness, and desirability as a friend. Uncorrected photographs received

significantly lower ratings on all scales than corrected photographs. Furthermore, the

differences between ratings of attractiveness and friendliness, popularity, and

desirability as a friend, for corrected versus uncorrected photographs, were twice as

large for female faces than for males faces. These results suggest that the negative

social impact of facial deformity may be more severe for females aged 8 through 16

years than for males of the same age, but that craniofacially anomalous children of both

sexes are perceived more negatively by peers than children with normal faces.

Two studies have examined the development of attractiveness stereotypes by

comparing ratings of photographs of individuals taken before and after oral surgery to

correct malocclusions. Elliot, Bull, James, and Lansdown (1986) compared ratings of

photographs taken before and after oral surgery to correct malocclusion. Four age

groups (6-7 year olds, 9-11 year olds, 13-15 year olds, 25-35 year olds) rated

pre-surgery and post-surgery photographs on attractiveness, friendliness, happiness,

intelligence, and social skill. For the two youngest age groups there were no

significant differences between ratings of pre- and post-surgery pictures, but the older

groups rated the post-surgery pictures as more attractive, intelligent, and happy than the

pre-surgery photographs. Rumsey, Bull, and Gahagan (1986) also compared pre- and

post-surgery ratings, using 5 through 11 year old raters. They found that 11 year olds

rated the post-surgery photographs as more attractive and gave them more positive

personality ratings than the pre-surgery photographs. Children under 11 did not rate

photographs differentially on the basis of surgery status. Among children ages 7 and

older, however, personality ratings were related to the child's rating of the

photograph's attractiveness. These results suggest that, for relatively minor anomalies

such as malocclusions, children may not begin to equate abnormality with

unattractiveness until about the age of eleven. However, children as young as 7 may

begin to form personality stereotypes on the basis of their own idiosyncratic judgments

of attractiveness.

Schneiderman and Harding (1984) used a bi-polar adjective checklist to

compare elementary school children's perceptions of children with repaired cleft lips

versus children without clefts. Color slides of children (ages 4 through 9 years) with

repaired bilateral clefts, repaired unilateral clefts, and without clefts were shown to 78

children ages 7 to 10 years. Each child rated all of the photographs on the adjective

checklist. Children with clefts were rated significantly more negatively on each of the

scales than were children without clefts. Furthermore, children with bilateral clefts

were rated more negatively on seven of the items than were children with unilateral


It should be noted that children under the age of 11 did not stereotype

consistently by facial abnormality in the Rumsey et al. (1986) and Elliot et al. (1986)

studies described above. Those over the age of six did, however, make personality

attributions on the basis of their own attractiveness judgments. Taken together, these

studies indicate that children as young as seven do make negative character attributions

on the basis of appearance, but that there is less agreement among young children than

adults regarding standards of attractiveness. The lack of agreement may reflect the fact

that the cognitive development of very young children is such that it is difficult to obtain

reliable ratings from them. It may also be that children agree on the unattractiveness of

clefts at an earlier age than they agree on the unattractiveness of malocclusions. An

alternative explanation is that the clefts were more severe than the malocclusions and

that it is was the degree rather than the type of abnormality to which child raters in the

Schneiderman and Harding (1984) study were responding.

Social Interactions of Individuals with Craniofacial Anomalies

If individuals with craniofacial anomaly are perceived differently than others,

are their interactions with others also affected? This question has been the basis of a

group of studies on the social interactions of individuals with craniofacial anomalies.

This research deals with the very earliest social relationship--the parent/child dyad.

Mothers are typically used in this research because they tend to be more available to the

researchers and have traditionally assumed the bulk of child-care responsibilities.

Barden, Ford, Jensen, Rogers-Salyer, & Salyer (1989) compared mother-child

interactions of five, four-month old infants with craniofacial anomalies to interactions

of five infants without anomalies. The anomaly group included infants with the

following diagnoses: Crouzon's syndrome, Aperts syndrome, hemifacial microsomia,

unilateral cleft lip and palate, and bilateral cleft. None of the infants had neurological or

mental disorders or had physical problems in addition to those associated with the

diagnosis. Infants and their mothers were observed in three interaction situations.

They were coded on six specific mother behaviors and six specific child behaviors, as

well as general ratings of gratification, responsiveness, and affective tone for each

member of the dyad. In addition, mothers completed the Life Experience Survey (LES)

(Sarason, Johnson, & Siegel, 1978), the Satisfaction with Parenting Scale (SWPS)

(Ragozin, Basham, Crnic, Greenberg, & Robinson, 1982) and a social support

questionnaire. Mothers of the infants with anomalies reported more satisfaction with

parenting and more positive life experiences following pregnancy then mothers of

normal children. There were no differences in self-reported general life satisfaction or

social support. Observational measures demonstrated significant differences between

the groups. Mothers of infants with anomalies touched their baby with affection less

often, engaged their infants in less tactile-kinesthetic stimulation, were less responsive

to the infant, and held the infant face-to-face less often than control mothers. The

infants' contributions to the interactions were also different between the groups.

Infants with anomalies were less responsive to their mothers, touched their mothers

less often, turned away from their mothers more frequently, and smiled and laughed

less often than their normal counterparts. This suggests that, despite the positive

self-report of mothers of children with craniofacial anomalies, presence of craniofacial

anomaly appears to be associated with disturbance in the mother-child interaction in the

direction of less responsive, physically stimulating and affectionate, face-to-face

interaction between mother and child. The results of this study must be viewed with

caution, however, due to the extremely small sample size.

Speltz, Armsden, and Clarren (1990) also compared infants with craniofacial

anomalies and normal controls on both observational measures of mother-child

interaction and parent self-report measures. In the anomaly group were 12 infants with

cleft palate, 11 infants with cleft lip and palate, and 10 infants with sagittal synostosis.

The control group consisted of 22 normal infants. The average age of both groups was

20 months. In contrast to the findings of Barden et al. (1989), there were no

differences between groups on the observational measures. There were, however,

significant differences on parent report measures. Specifically, mothers of infants with

craniofacial anomalies had higher PSI Parent Domain and Parental Competence scores

(greater stress related to being a parent and lower feeling of parental competence) lower

General Well-being Schedule (Dupuy, 1979) scores (lower psychological and

emotional well-being), and lower Locke-Wallace Marital Adjustment Scale (Locke &

Wallace, 1959) scores (less marital satisfaction) than mothers of normal infants. These

results indicate that mothers of children with craniofacial anomalies experience

increased stress related to parenting and decreased general and marital

well-being when compared to mothers of normal children.

That Barden et al. (1989) failed to find differences between groups on

self-report measures may have been due to the extremely small sample size of that

study. Barden et al did, however, detect differences in mother-child interaction which

were not apparent in the Speltz et al.(1990) study, despite the larger sample size. This

difference in findings may be due to the fact that the Speltz et al. study focused

primarily on verbal interaction measures, whereas Barden et al found differences in

nonverbal interactions such as touching the infant and sitting face-to-face with it.

Children in the Speltz et al. study were also approximately 16 months older than

children in the Barden et al study, so differences may reflect the course of the families'

adaptation to the presence of a child with a craniofacial anomaly. The combined results

of both studies indicate that parenting an infant with a craniofacial anomaly is stressful

and may be associated with less positive non-verbal mother-child interaction patterns.

Clifford (1969) found evidence that the effects of different types of anomalies

on the social development of the child between birth and two years of age may be

additive. Thirty pairs of parents of children with clefts of the lip and palate and 30

pairs of parents of children with clefts of the lip Dr palate were interviewed within the

first two years of their child's life. Parents of children with clefts of the lip and palate

described their children as having more difficult temperaments and as being less likable

and cute than did parents of children with clefts of either the lip or palate. Children

with clefts of the palate may be more irritable as a result of feeding difficulties and

discomfort related to ear infections. Children with cleft lips may be perceived as less

cute and likable than children without visible defects because of their appearance. In

addition, both feeding difficulties and appearance may impact on parental treatment of

the child, which could itself affect the child's behavior. Clifford's results suggest that

the problems of both types of defects are additive in their influence of the parent's

perception of the child, which may or may not be reflective of the child's actual


Speltz, Goodell, Endriga, and Clarren (1994) compared three-month-old infants

with either cleft palate (CP) (n = 19), cleft lip and palate (CLP) (n = 15) or no cleft

(NC) (n = 17) on measures of feeding interaction. They found that infants with clefts
gave less clear signals of readiness to feed and smiled less during feeding than infants

without clefts. Mothers of infants with cleft lip and palate also were rated as less

sensitive to infant feeding cues than mothers of children with cleft palate only or infants

without clefts.

The authors suggest that the maternal insensitivity to infant feeding cues in the

CLP group may be the result of difficulty with the physical aspects of nursing a CLP

child. CLP mothers have to be particularly diligent about ensuring that their children

receive adequate nourishment, which may cause them to focus more on the goal of

feeding a certain amount of formula than on the interactive qualities of the nursing

experience. These results suggest that, in the earliest important social interaction

(feeding), children with clefts may be at a disadvantage because of the physical

difficulties involved in nursing.

At the age of three, differences remain in the mother-child interaction patterns of

children with craniofacial and other physical anomalies as compared to peers without

anomalies. Alien, Wasserman, and Seidman (1990) observed 37 children with a wide

range of anomalies (all of whom were mobile and able to manipulate objects

independently) and 44 normal children interacting with their mothers during four

situations. Stanford-Binet (Form L-M, 1972 norms) (Terman & Merrill, 1973) and

Preschool Language Scale (Zimmerman, Steiner, & Pond, 1979) scores were also

obtained. Presence of anomaly did not account for variations in performance on

intellectual measures. During the mother-child interaction period, children with

anomalies were more compliant and passive, and their mothers were more active and

controlling than normal mother-child dyads. Children with anomalies also made fewer

demands on their mothers during free play. It appears that a pattern of passivity and

low demandingness of adult attention is already in place during the preschool years. The

continuation of this pattern during the school years will be discussed in the following


Adjustment of Children with Craniofacial Anomalies

Richman and Eliason (1982) reviewed the early literature on the intelligence,

achievement, and psychological adjustment of children with clefts of the lip and / or

palate. They reported that children with clefts score slightly lower on tests of verbal

intelligence than children without clefts, but do not differ on tests of non-verbal

intelligence and generally fall within the normal range of intellectual functioning. They

noted that children with clefts who also have additional physical anomalies tend to score

lower than children without additional anomalies, and that speech and hearing problems

may be responsible for language delays in some children with clefts, accounting for the

slightly lower verbal intelligence scores of children with clefts as compared to controls.

Despite nearly comparable intelligence scores, children with clefts score

significantly lower than controls on tests of academic achievement. Richman and

Eliason (1982) speculated that several factors may be responsible for the reduced

achievement of children with clefts, including language deficits, decreased expectations

of teachers and parents, and behavioral inhibition of children with clefts in the

classroom. On the Behavior Problem Checklist (Quay & Peterson, 1967), teachers

rated children with clefts as more inhibited than other children in their classrooms

(Richman, 1978, in Richman & Eliason, 1982). Richman and Eliason hypothesized

that children with clefts develop an inhibited interaction style to avoid teasing about

their speech or appearance, and that this inhibition interferes with their classroom

participation, thereby lowering their achievement. In reviewing research on the

personality and behavioral characteristics of children with clefts, Richman and Eliason

found additional support for the idea that children with clefts are more inhibited than

their peers without clefts. They noted that there was no evidence of severe

psychopathology in children with clefts, but that several studies did find evidence of

increased inhibition, using such measures as the Minnesota Multiphasic Personality

Inventory (Hathaway & McKinley, 1943) (Harper & Richman, 1978, in Richman &

Eliason, 1982) and the Missouri Children's Picture Series (Sines, Pauker, & Sines,

1971) (Richman & Harper, 1979, in Richman & Eliason, 1982).

Since the Richman and Eliason review, a number of additional studies of the

adjustment of individuals with craniofacial anomalies have been added to the literature.

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

Problem Checklist 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

categorized as having "conduct problems." Unfortunately, it is difficult to determine

how this pattern compares with the adjustment of children without clefts, because a

control group was not used. The authors did note, however, that teacher ratings of

conduct problems were "higher" (statistical comparison was not performed) than the

normative data of the Behavior Problem Checklist. It is difficult to evaluate the

adjustment of children with clefts as compared to normals on the basis of this study.

The study does, however, highlight the need to consider the effects of age and sex

when studying the psychosocial adjustment of children with craniofacial anomalies.

Tobiasen and Hiebert (1984) examined the behavior of 41 male children (aged

2-12 years) with cleft lip and palate, using the Eyberg Child Behavior Inventory

(ECBI) (Robinson, Eyberg, & Ross, 1980). The ECBI provides two scores, an

Intensity Score which is the sum of the frequency of disruptive behaviors reported by

the parent, and a Problem Score which is the number of behaviors the parent considers

to be problematic. Comparing each of the 36 items separately, Tobiasen and Heibert

found no differences between the group with clefts and the ECBI norm group on

frequency of the problematic behaviors, except that the norm group scored higher on

the item "Is overactive or restless." They did not report the total Intensity score for the

group with clefts or statistically compare the total Intensity score to the norm group

score. For the Problem score, they found that parents of children with clefts reported

an average of 4.7 problem behaviors, as compared to an average of 7.1 for the

normative group. Although they did not report a statistical comparison of the two

groups on the Problem Score, they stated that "parents of children with clefts were

tolerant of a greater number of behavior problems than parents of children without

clefts." Because they had not found a difference between groups on the intensity score,

they concluded that parents of children with clefts are "more tolerant of misbehavior"

than other parents, a conclusion which has widely been quoted. Considering that

no statistical comparison of the Problem score was reported, and that the parents of

children with clefts reported only 2.4 fewer problems than the normative group, it is

surprising that this conclusion has been so uncritically accepted. Furthermore, for

every behavior which a significantly smaller percentage of parents of children with

clefts endorsed as problematic than the norm group, the mean frequency (intensity)

score was also lower, although this reached statistical significance for only one item. It

should not be concluded on the basis of this study (as it often has) that parents of

children with clefts of the lip and palate are more lenient than parents of normal

children. What can be concluded is that male children with clefts ages 2-12 years

display no more (and possibly fewer) disruptive behaviors in the home than do their

peers without clefts.

Palkes, Marsh, and Talent (1986) found that six of ten children (ages five

through fifteen years) with craniofacial anomalies were rated as having significant

problems in family interactions, school productivity, and peer relationships as

measured by the Missouri Behavior Checklist (Sines, Pauker, & Sines, 1971). Due to

the extremely small sample size of this study, any interpretations must be made with

caution. A strength of the study, however, was that it departed from the usual

sampling of children with cleft lip and palate to include children with craniosynostoses,

craniofacial microsomia, craniofacial dystoses, and Hallermann-Streiff syndrome, in

addition to those with major facial clefts. Although the difference in findings may have

been attributable to the small sample size of this study, it may also be that children with

other types of anomaly differ in adjustment from children with clefts.

Kapp-Simon (1986) compared scores on the Primary Self-Concept Inventory

(PSCI) (Muller & Leonetti, 1974) of 50 children with cleft lip and / or palate (ages five

through nine years) with those of 172 control children. She found that a significantly

higher percentage of children with clefts than control children fell into the "at risk"

category for Total Self-Concept and the Social-Self Domain. Kapp-Simon interprets

this as indicating that children with clefts are at risk of developing poor self-concepts,

particularly in the social domain. It should be noted however, that "at risk" was

defined as "falling below the 50th percentile of the original sample for Total Self-

Concept," and that only 41.9% of the control group scored in this range. This suggests

that the control group may have been unusually well-adjusted and therefore not an

appropriate comparison group. In actuality, only 54% of the group with clefts scored

in the "at risk" range, a percentage which would probably not have been significantly

different from a more "average" comparison group. Fully 74% of the group with clefts

did score in the "at risk" range on the social self domain, suggesting that children with

clefts may have circumscribed self-esteem problems in the area of social interaction.

An earlier study by Kapp (1979), using the Piers-Harris Children's

Self-Concept Scale, failed to find differences in the social domain of self-concept for

children with clefts versus controls. This difference in findings may be due to age

differences in the groups studied (the 1979 study examined 11 through 13-year-olds) or

in the measures utilized. The social domain of the Piers-Harris examines children's

feelings about their relative popularity, whereas the Primary Self-Concept Inventory

social domain encompasses both peer acceptance and helpfulness. Children with clefts

scored lower on the helpfulness items, but not on the peer acceptance items of PSCI.

Children with clefts may have more concerns about the social roles they play (help-

recipient versus helper) than about whether they will have friends at all. The 1979

study did not find differences in Global Self-Concept between children with clefts and

controls, but children with clefts did report significantly less satisfaction with physical

appearance. In addition, girls with clefts (but not boys) reported being more anxious,

less happy and satisfied, and less confident about school performance than girls

without clefts. This pattern of findings suggests that concerns with physical

appearance and social roles may be a problem for some children with clefts and that

adolescent girls may be more at risk for poor self-esteem than other children with


Broder and Strauss (1989) also examined the self-concept of children with

clefts, using the Primary Self-Concept Inventory. Their subjects were 20 children

between the ages of 6 and 9 years who were of normal intelligence and did not have

additional associated congenital anomalies. Subjects were divided into three groups,

cleft lip only, cleft palate only, cleft lip and palate. All groups with clefts combined

scored in the at risk range and significantly below the controls (18 first graders without

clefts) on the total self-concept score, both factors of the social domain, and the

physical factor of the personal self domain. This is consistent with previous findings

of low self-esteem in elementary school-aged children with clefts, particularly in regard

to social self-concept and physical self-concept. In addition, children in the group with

clefts of the lip and palate scored significantly lower than the groups with clefts of the

lip or of the palate only on the intellectual, social and total self-concept scales. Broder

and Strauss proposed that "the visible (cleft lip) and invisible (cleft palate) defects have

a compounding negative effect on self-concept." It may be that the added negative

effect of the cleft palate is associated with problems in speech. Unfortunately, no

speech ratings were reported.

Studies with Measures of Physical Appearance

Although physical appearance is presumed to be responsible for adjustment

difficulties of children with craniofacial anomalies, only a few studies actually employ

measures of physical appearance. A rating of facial disfigurement was included in

Richman's (1976) study of children with cleft palates who were ages ages nine to

fourteen years. The facial disfigurement rating was not significantly correlated with

measures of achievement or behavioral ratings. Unfortunately, neither the method of

rating nor the distribution of disfigurement scores was reported, so it is difficult to

determine whether the lack of association was due to poor measurement, low

variability, or a true independence of attractiveness and adjustment. Richman did find

differences between the group with clefts and controls. Children in the group with

clefts (n=44) scored higher on the personality disorder scale (a measure of

internalizing) of the Behavior Problem Checklist and lower on the Iowa Tests of Basic

Skills (Hoover, Hieronymus, Frisbie, & Dunbar, 1993) (a measure of academic

achievement) than did controls (n=44). Boys with clefts performed an average of one

year below the mean of boys without clefts on the Iowa Tests of Basic Skills, while

girls with clefts performed one half year below the mean of girls without clefts, despite

nearly identical mean intelligence quotients for the two groups (cleft mean IQ =104,

non-cleft mean IQ = 103.8). This study suggests that children ages 9-14 years with

clefts of the palate or lip and palate are inhibited and underachieving when compared

with their peers without clefts. It is difficult to determine whether these characteristics

are related to appearance, due to the lack of information on the facial disfigurement

measure used in the study.

Starr (1980) examined the relationships between physical attractiveness,

self-esteem, ratings on the Missouri Children's Behavioral Checklist, and an attitudes

toward clefting scale. Subjects were 49 clinic patients ages ten years and older. Starr

examined the relationship between the psychosocial adjustment variables and

attractiveness in two ways. First, he selected 14 patients with scars on their upper lips,

as rated by the clinic's Social Service staff. These patients were compared with 14

patients without scars, who were matched on age, sex, and cleft type. There were no

significant differences between groups on any of the three adjustment measures. For

the next investigation, full-face photographs of the patients were given attractiveness

ratings by five college students, using a five-point scale. Patients receiving a sum score

of 11-15 (n = 19) were placed in the attractive group, while patients with sum scores of

nine and below (n= 19) were placed in the unattractive group. Eleven patients whose

sum score was 10 were omitted from the analyses. No differences were found between

groups on any of the measures. Unfortunately, the distribution of attractiveness scores

was not reported, so it is difficult to determine whether the groups' differences in

attractiveness were large enough to be meaningful. None of the patients in the

"attractive" group received a rating higher than 15, which means that the average score

of the most attractive person was three, exactly in the middle of the attractiveness rating

scale. Larger group differences in attractiveness may be necessary to adequately test

the relationship between attractiveness and adjustment.

Lefebvre and Barclay (1982) used the Hay's Standardized Rating Scale of

Appearance (Hay & Heather, 1973) to examine the relationship between appearance

and psychosocial adjustment in 175 patients undergoing craniofacial surgery. Patients

presented with a wide range of facially disfiguring conditions. Patients and their

parents rated the child's appearance before and after reconstructive surgery. In

addition, 76 patients completed the Piers-Harris Self-Esteem Inventory (Piers, 1969)

and all patients completed pre and post-operative semistructured interviews addressing

social adjustment, school performance, peer reactions to appearance, and future plans.

Children under the age of 13 rated their appearance significantly more positively than

did their parents. Adolescents' appearance self-ratings were not significantly different

from parental ratings. No analyses of the relationships between improvement in

appearance and the adjustment measures were reported; however, a large percentage of

patients reported better psychosocial adjustment after surgery to improve appearance,

suggesting that improved appearance may have had a beneficial effect on psychosocial

adjustment. On the Piers-Harris Self-esteem Inventory, 24 percent of patients scored in

the below average range pre-operatively, compared with 14 percent post-operatively,

but the difference was not significant. Only 35 post-operative ratings were available,

and these were compared with the entire pre-operative group rather than matched to the

individual patient's pre-operative rating, so the lack of a significant difference is not

conclusive. In summary, both subjective and objective improvements in psychosocial

adjustment were noted in patients following craniofacial surgery to improve


Amdt, Travis, Lefebvre, Niec, and Munro (1986) also examined patients'

adjustment both pre and post-operatively. Subjects were 22 children between the ages

of 8 and 17 years, with a variety of diagnoses involving facial disfigurement. Subjects

were divided into two groups, severely affected (n=14) and moderately affected

(malocclusions only, n=8). Patients completed the Piers-Harris Self-Concept scale and

made self-ratings of disfigurement, using the Hays Rating Scale. Patients were also

rated by their parents and a panel of 15 judges, using the Hays Rating Scale. Prior to

surgery, patients' and parents' disfigurement ratings were equal or higher than those of

judges for both groups. Post-surgery, appearance ratings by judges, parents, and

patients improved significantly for the mildly affected children. For the severely

affected patients, only ratings by parents and patients improved significantly. Judge

ratings indicated that the severely affected children continued to appear quite disfigured

to outside observers, yet there were significant increases in self-esteem for both

groups. These results suggest that it is the patients' belief about his or her appearance

which affects self-esteem, not the "reality" which others see. This echoes the findings

of Kenealy, Frude, and Shaw (1989) with a normal population. In that study,

children's self-ratings of attractiveness correlated highly with self-reported self-esteem,

but unacquainted judges' ratings of the children's appearance were not related to the

children's self-esteem ratings.

Richmond, Holmes, and Eliason (1985) examined the relationships between

self-ratings of disfigurement, teacher ratings of disfigurement, and parent and child

ratings of adjustment as measured by the Behavior Problem Checklist (BC).

Thirty-six adolescents ages 14 through 17 years were divided into two groups based on

parent ratings on the BC. Adolescents with both conduct problem and personality

problem scores within one standard deviation of the normative means on the BC were

placed in the "well-adjusted" group, while those with scores more than one standard

deviation above the normative mean on either dimension were considered "poorly

adjusted." In the well-adjusted group, there were no significant differences between

parent and child ratings on the BC or between teacher and child ratings of appearance.

Adolescents in the poorly adjusted group agreed with parental ratings on the conduct

problems dimension of the BC, but rated themselves as having significantly fewer

personality problems (internalizing) than did parents and as significantly more attractive

than did teachers. Adolescents' self-ratings on all three measures were not significantly

different between the two adjustment groups. The authors hypothesized that some

adolescents use denial of their facial disfigurement as a defense mechanism and that

their social withdrawal is a consequence of this use of denial. Thus, children who are

rated by parents as being the most withdrawn may have higher self ratings as a result of

their use of denial. Closer inspection of the data provides a clue as to which children

may be most likely to use denial. Children in the poorly adjusted groups were rated by

teachers as much more disfigured than children in the well-adjusted group. It might be

that more severe disfigurement is associated with the use of denial and, consequently,

with increased introversion. Although severely affected children in the Arndt, et al.

(1986) study did not rate their appearance unrealistically (use denial), they were

anticipating surgery and may have had less of a need to deny the extent of their

unattractiveness because of the hope that surgery would change their appearance.

Broder and Strauss (1989) did not directly measure physical attractiveness, but

did compare invisible (cleft palate only) versus visible (cleft lip or cleft lip and palate)

defects. As described above, children ages 6-9 years with either a visible or an

invisible defect scored lower than normals on a measure of self-esteem, but children

with both visible and invisible defects scored significantly lower than either group with

only one type of defect. This suggests that the appearance of the child made an

independent contribution to the decrease in self-esteem, apart from the effects having a

defect or of having problems specifically related to clefts of the palate (such as speech


In addition to appearing different from their peers, many children with

craniofacial anomalies also have to contend with speech and hearing impairments.

Several studies therefore have focused on speech and hearing impairments in children

with craniofacial anomalies.

Influence of Speech Impairment on Perception of Others

In a study by Glass and Starr (1979), color photographs and slides of 50 adults

were rated on a seven point attractiveness scale by adult judges. Judges also made

speech ratings, based on two-minute recordings of patients reading a standard

paragraph. Judges were presented with both a picture and the audio tape, and were

required to rate either appearance or the acceptability of the speech. There was a trend

for appearance of the photograph to affect speech ratings, while acceptability of the

speech significantly affected appearance ratings. Judges' ratings of appearance of the

photographs decreased as the photographs were paired with decreasingly acceptable

speech recordings. This suggests that individuals whose speech is negatively affected

by the presence of a craniofacial anomaly (such as cleft palate) may be viewed less

positively than others because of the quality of their speech.

Sinko and Hedrick (1982) also examined the effects of speech on ratings of

appearance and appearance on ratings of speech, using similar methodology as in the

Glass and Starr (1979) study, with the exception that the faces were presented via

videotape rather than still photography, and faces were paired only with the individual's

own speech. They found that, not only did speech acceptability influence ratings of

facial attractiveness, but facial attractiveness also influenced ratings of speech

acceptability. That Sinko and Hedrick found a significant influence of appearance on

speech acceptability while the effect was only a trend in the Glass and Starr study is

probably due to the greater variability in attractiveness ratings in the Sinko and Hedrick

study versus the Glass and Starr study. The results of these studies suggest that both

children's appearance and speech should be accounted for when examining the effects

of either.

The adverse effects of speech impairment may extend to potential employers'

impressions of an individual. Scheuerle, Guilford, and Garcia (1982) videotaped a

24 year old male with a cleft palate reading 2 minute passages with and without an

obturator (an appliance which is used to seal palatal clefts). Twenty-six adult male

judges rated the speaker after each passage on bi-polar adjective scales assessing the

dimensions of competence, agreeableness, and self-assurance. These dimensions

were chosen as "three characteristics of a desirable employee" on the basis of previous

research. Subjects rated the speaker as less competent, agreeable, and self-assured

after viewing segments during which the obturator was removed than after segments

during which the obturator was in place. Ratings of the competence dimension were

particularly affected. This indicates that speech impairment may negatively affect

potential employers' perceptions of an applicant's competence.

Studies Which Address Speech: Adjustment

Tobiasen, Levy, Carpenter, and Hiebert (1987) compared children (ages 2-12

years) with clefts only with children who had clefts and associated congenital

malformations, on measures of conduct and school problems. Although children in the

associated anomalies group received significantly higher frequency scores on eight

disruptive behavior items of the Eyberg Child Behavior Checklist (Robinson, Eyberg,

& Ross, 1980) than did children without associated anomalies, there were no group

differences in number of problems behaviors, and none of the groups received

abnormally high problem behavior ratings. Parents were significantly more likely to

rate children in the associated anomalies group as having the following school

problems: following instructions, slow learner, short attention span, and not finishing

work. When a post hoc analysis of speech deficits was also conducted, children with

associated congenital anomalies were found to have been given significantly poorer

speech ratings by speech language pathologists during routine clinic evaluations than

children without associated congenital anomalies. Interpretation of these findings is

limited by the small sample size (as few as six subjects in a group), but they do suggest

that there is an association between speech deficits and school problems in children

with clefts. The speech problems of children with associated anomalies may have a

causative role in school problems. Alternatively, the association may due to a common

variable underlying the associated anomalies, the speech deficits, and the school


Richman (1983) examined the relationship between self-reported speech

concerns, appearance concerns, elevated MMPI scales (2 depression,

7 psychaesthenia, and 0 introversion), satisfaction with social interactions, and

satisfaction with school performance among adolescents (15-18 years; 16 male,

14 female) with cleft lip and palate. Social and educational dissatisfaction were

positively associated with elevated MMPI scales. MMPI elevations were positively

related to level of facial appearance concern, but not to to level of speech concern.

These results suggest that, during adolescence, appearance assumes a greater role in the

child's adjustment than does speech quality.

McWilliams and Musgrave (1972) compared children (ages 3-16 years) with

normal speech (n = 32), children with normal voice quality who misarticulated

(n = 77), and children whose speech was hypernasal and who misarticulated

(1 = 61). All children had clefts of the palate or of the lip and palate. On intelligence

tests, the children with the poorest speech scored significantly lower than the other

groups, although their mean IQ (96) was still well within the normal range. Mothers

also completed a behavior problem inventory for each of the children. Both groups

with speech impairments were reported to have significantly more behavior problems

than the normal speech group. The group of children with normal voice quality and

misarticulation had a higher incidence of bad temper, fearfulness, bed wetting and

preferring to be alone than either of the other groups. Children with poor voice quality

and articulation errors were not distinguishable from the other groups in frequency of

individual behavior problems. The results of this study suggest that children with

speech deficits manifest more behavioral problems than their peers with normal speech.

A possible confound is the fact that there was a higher percentage of children with clefts

of the lip and palate (vs. palate only) in the speech disordered group than in the normal

group. It may be that differences in adjustment were due to differences in appearance

rather than speech.

In brief, the literature implies that there is some connection between speech

impairment and adjustment difficulties. The evidence is far from conclusive, however,

and no consistent pattern of behavioral disorder emerges.

Studies Which Address Speech and Hearing: Achievement

Hubbard, Paradise, McWilliams, Elster, and Taylor (1985) found no

differences between scores of children with cleft palates and test population norms on

the WISC-R (Wechsler, 1974) (verbal and performance subscales), the Vineland Social

Maturity Scale (Doll, 1965), the Coopersmith Self-Esteem Inventory (Coopersmith,

1959), and the Child Behavior Checklist (Achenbach, 1979; Achenbach & Edelbrock,

1979). The 48 children (ages 5 through 11 years) in this study were enrolled in two

cleft palate treatment centers which provided management of otitis media. Perhaps as a

result of this treatment, their average hearing loss was below 13 dB. The lack of

intellectual and social difficulties in this sample may have been due to minimization of

hearing loss due to good medical management

Kommers and Sullivan (1979) examined written language skills of 17 children

(10 male, 7 female; ages 8 to 13 years) with cleft palates, using the Myklebust Picture

Story Language Test (PSLT) (Myklebust, 1965). Sixteen of the children had histories

of hearing loss, tube placement, and/or ear infections. Twelve of the subjects had nasal

speech or difficulty pronouncing sibilants. All subjects had WISC performance IQ

scores of at least 85 and were above average intelligence as a group. Mean scores on

the PSLT were below average, with the exception of abstract-concrete relationships.

Performance IQ was not significantly related to PSLT scores. Kommers and Sullivan

suggest that the poor written language performance relative to performance IQ scores

may be related to speech impairments. Specifically, they hypothesize that children with

clefts palates simplify their speech in order to accommodate difficulties in production,

and that this simplification is carried-over into their written communication. It is also

possible that hearing loss during early childhood may have had a negative impact on

language acquisition in this sample.

Estes and Morris (1970) examined the relationship between hearing sensitivity,

speech proficiency, and intellectual functioning in a sample of 167 children under the

age of 15 years with clefts of the lip and / or palate. Children with cleft palate only

scored lower on intelligence measures than either children with cleft lips or cleft lips

and palates. Verbal IQ was particularly depressed. There was only a weak relationship

between communication defectiveness and hearing loss. For all groups, speech

proficiency was related to WISC performance and full-scale IQ, but not to verbal IQ,

Binet IQ or to ratings on the Vineland Social Maturity Scale. The authors note that the

lack of relationship between WISC verbal IQ and speech proficiency suggests that the

decreased verbal IQ of children with cleft palates only is due to underlying factors

independent of speech or hearing related communication deficits. This is in line with

findings by Richmond and Eliason (1984) of general language deficiency in children

with cleft palates only.

Richman (1980) identified two types of verbal disability in children with clefts,

depressed verbal IQ with intact symbolic mediation skills and depressed verbal IQ with

deficits in basic symbolic language. She hypothesized that the first type of verbal

deficit was related to speech and hearing impairments, whereas the second type may be

due to an underlying general language disability. There was a higher proportion of

children with cleft palate only in the group with symbolic language deficits. In light of

this, the effects of underlying general language deficiency must be considered when

children with cleft palate only are included in studies of the relationship between speech

or hearing deficits and intellectual achievement

Need for a Model

Studies of children with craniofacial anomalies suggest that poor self-esteem,

social inhibition and decreased school achievement may be problems for subgroups of

these children, as would be predicted from literature on attractiveness, speech

impairments, and hearing loss. Findings are inconsistent, however, and little attempt

has been made to reconcile differences between studies. Furthermore, few studies

examine more than one or two variables at a time, and there currently exists no unifying

model which accounts for attractiveness, hearing impairment, speech impairment, age,

and sex in predicting psychosocial adjustment of children with craniofacial anomalies.

Such a model is necessary to provide structure for future research and to make sense of

what is currently known.

Overview of Findings

Development of a model of psychosocial adjustment of children with

craniofacial anomalies may best begin with an examination of what is already known.

Due to the high prevalence of clefts of the lip and palate relative to other craniofacial

anomalies, most of this literature focuses on children with clefts. The major findings of

the craniofacial literature may be summarized as follows.

By age seven, children perceive children with craniofacial anomalies less

positively than children with normal faces (Schneiderman & Harding, 1984; Tobiasen,

1987; Elliot, et al., 1986). The difference between ratings of children with clefts

versus normal children is greater for females than for males (Tobiasen, 1987), and for

children with bilateral clefts than those with unilateral clefts (Schneiderman & Harding,

1984). Speech quality of an individual with a cleft may also have an effect on the

characteristics attributed to that person by others (Scheurle, et al., 1982). Perception of

an individual's attractiveness is affected by speech quality, and acceptability of speech

is affected by attractiveness (Glass & Starr, 1979; Sinko & Hedrick, 1982).

There is some evidence that mother-infant interaction at 4-5 months is different

for children with craniofacial anomalies. Less face-to-face holding, decreased

touching, and less responsiveness of mother and infant toward each other than in

normal dyads has been observed (Barden et al., 1989). In addition, mothers of

toddlers report increased parenting stress, lower feelings of competence, and decreased

well-being and marital satisfaction (Speltz et al., 1990). Between birth and age two,

children with clefts of the lip and palate are perceived as having more difficult

temperaments than children with clefts of the lip or palate alone, suggesting an additive

effect of the problems unique to each type of cleft (Clifford, 1969). By age three,

mothers of children with craniofacial anomalies are more active and controlling of

interactions than mothers of normal children, and the children with craniofacial

anomalies are more passive and compliant than their peers (Allen, et al., 1990).

Children with clefts generally score in the average range of intellectual

functioning, although their verbal IQ is typically slightly depressed (Richman &

Eliason, 1982; Estes & Morris, 1970). Despite this, the achievement of children with

clefts is typically below average by about 1 year for males and 1/2 year for females

(Richman & Eliason, 1982). Speech and hearing problems are thought to account for

the low written language achievement of some children with clefts (Kommers &

Sullivan, 1979; Richman, 1980; Richman & Eliason, 1984; Estes & Morris, 1970). In

addition, it has been suggested that a subgroup of children with cleft palate also have a

general language learning disability (Richman, 1980; Richman & Eliason, 1984).

Children whose craniofacial anomalies are associated with genetic syndromes may also

have learning disabilities or retardation, but these are not the focus of this review.

Interestingly, self-reported concerns about appearance have also been associated with

self-reported dissatisfaction with education among adolescents (Richman, 1983).

In general, children with clefts have been described as inhibited and as having a

relatively high incidence of internalizing problems (Richman & Eliason, 1982). By

elementary and junior high school, however, boys with clefts have also been reported

to display increasing incidence of conduct problems. Adolescent girls do not show this

tendency (Schneiderman & Auer, 1984). They are more likely to be anxious and

dissatisfied with school problems than their male peers with clefts, despite better school

performance (Kapp, 1979). More disruptive behavior problems have been noted in

children with associated anomalies and those with greater speech problems than in other

children with clefts (Tobiasen, 1987; McWilliams & Musgrave, 1972).

The self-esteem of both males and females with clefts is lower than average at

all ages, but most studies show only a mild effect (Kapp-Simon, 1986; Broder &

Strauss, 1989; Kapp, 1979). Clefts of the lip and palate are associated with greater

decreases in self-esteem than either cleft lip or cleft palate alone, possibly reflecting

additive effects of speech impairment and appearance (Broder & Strauss, 1989).

Increases in self-esteem, subjective measures of well-being, and objective

measures of well-being have been noted following surgery to correct the appearance of

individuals with craniofacial anomalies (Amt et al, 1986; Le Febvre & Barclay 1982).

Concerns about appearance are related to increased introversion in late adolescence

(Richman, 1983). Among midadolescents, deformity ratings made by others are

positively associated with parent ratings of internalizing problems and conduct

problems, but not with adolescent ratings. In addition, severely deformed adolescents

rate their appearance less negatively than others, whereas less severely deformed

adolescents' ratings of both appearance and internalizing are close to ratings made by

others (Richman, Holmes, & Eliason, 1985). Elementary school children with

craniofacial anomalies have also been observed to rate their own appearance more

highly than others (Lefebvre & Barclay, 1982).

A Model of Adjustment

Based on the previously listed observations, the following model is proposed.

It is presented not as a final model, but as a framework to be tested and modified, and

as a guide for future research. For clarity of reading, variable relationships will be

presented below in statement form. but they represent hypotheses only.

Figure 1 illustrates variables affecting the child's development from birth

through the first year of life. "Child characteristics" include the infant's attractiveness,

temperament, responsiveness, and ease of feeding. "Parent characteristics" include the

parents' feelings of responsibility for the child's condition, sense of parental

competence, personality adjustment prior to the child's birth, and general well-being.

"Family stresses and resources" include child hospitalizations, medical costs, financial

resources, social supports, and career demands.

According to this model, the child is directly affected by family stressors and

resources, such as the ability of the parents to provide adequate food and medical care.

Child characteristics, in turn, influence family resources (Benson, Gross, Messer,

Kellum, & Passmore, 1991). For example, friends and relatives may be reluctant to

baby-sit for an infant that is difficult to feed or is extremely disfigured. Parental

personality characteristics also play a role in determining family resources, both in

terms of job opportunities and the formation of social support networks. At the same

time, parental well-being is affected by family stresses and resources. The parent may

be fatigued and worried by frequent child hospitalizations and concerned about financial

costs of the infant's medical care. On the other hand, adequate insurance coverage, a

career situation which allows absences for the child's medical care, and a network of

relatives and friends who are willing to help care for the child increase parental


Characteristics of both the parent and the child contribute to the quality and

quantity of their interactions with each other. Child characteristics of attractiveness,

temperament, and responsiveness influence the parent's response to infant The

parent's contribution to the interaction is affected by the parent's physical and mental

well being and by the parent's pre-existing personality traits. In addition,

characteristics of both the parent and child are influenced by their interactions with each

other. Responsiveness and mood of the infant are affected by the parents' interactions

with it, and the parents' sense of competence and well-being are affected by their

interactions with the child.

Family Stresses / Resources

Child C aracteristics (~ Parental Characteristics

Figure 1
Factors Affecting Adjustment: First Year of Life

Figure 2 continues to follow the child's development from 12 months through

the beginning of school. The variable relationships outlined in the previous model

remain. At this point, it also becomes useful to distinguish between child

characteristics which are a direct result of the craniofacial anomaly and variables which

are representative of the psychosocial development of the child. "Craniofacial

variables" include physical attractiveness, speech quality, and hearing ability. "Child

psychosocial variables" include internalizing behavior problems, externalizing behavior

problems, social skills, and self-concept

At this stage, acquisition of language becomes an important influence on the

child's psychosocial development. Language acquisition is affected by the child's

hearing ability and physical speech apparatus. Development of speech affects the

parent-child relationship through its effect on the child's ability to verbally respond to

the parent. The child's responsiveness to the parent is also affected by the child's

ability to hear the parent's speech. The child's responsiveness to the parent in turn

influences the parent's manner of interacting with the child. In particular, parental

activity may increase in order to elicit responses from a passive or speech-delayed

child. Completing the cycle, degree of parental control and activity influences the

child's activity and responsiveness, the child becoming more passive in response to an

active, controlling parent.

F mily Stresses / Resources

Child Psychosocial Variables

Craniofacial Variables(-- )Parental Characteristics

Figure 2
Factors Affecting Adjustment: One Year School Entry

The child's developing self-concept is also affected by the cycle described

above. Opportunities for mastery experiences may be determined by the child's

activity level and the parent's degree of control. The child who is active and whose

parent permits exploration has increased opportunities for esteem-enhancing mastery


The child's self-concept is also influenced by attractiveness, as the child

becomes aware of the social meanings of attractiveness and of her or his own

attractiveness level. This is moderated by parental attitudes towards the child's

appearance. Accepting parents may decrease the negative impact of unattractiveness by

emphasizing their child's positive attributes.

Figure 3 illustrates the child's development in the preschool and early

elementary school years. Academic performance is added to the child psychosocial

development variables, and a new factor, peer relationships, is also included. At this

time, peer interactions become very important. Both attractiveness and speech quality

are associated with varying amounts of teasing by peers. Although parental attitudes

continue to modify the direct influence of appearance on the child's self-esteem, both

appearance and speech indirectly affect self-concept through teasing by peers. The

degree to which a child is behaviorally inhibited and experiences internalizing problems

(anxiety, depression) is also affected by teasing, which may cause the child to

withdraw or become anxious and depressed.

Speech and hearing quality impact on the child's school performance due to

their effects on language acquisition and on the child's ability to hear the teacher.

Hearing loss is also related to degree of conduct problems, probably as a result of

frustration with communication limitations.

Family Stresses / Resourc


Child Psychosocial ,
/ Parental Characteristics

Peer Interactions

Figure 3
Factors Affectine Adjustment: Preschool and Early Elementary School

Within the child psychosocial category, variables may be mutually influential.

For example, child self-concept and degree of internalizing may affect school

performance by influencing the degree to which the child participates in classroom

discussions and activities. Conversely, school performance impacts on the child's

self-esteem and degree of internalizing problems.

Figure 4 demonstrates that, by late elementary school, gender differences

begin to modify the effects of some of the variables. Conduct problems (included in the

psychosocial development category) begin to develop in more male than female children

with craniofacial anomalies. Conduct problems develop in response to teasing

(fighting), as an expression of frustration with communication limits imposed by

hearing difficulties, as a response to frustration or boredom with school, and as a

manifestation of poor self-concept Development of conduct problems is also

moderated by parenting style. School performance is affected by conduct problems and

by gender. The school performance of females exceeds that of males at this age. The

rest of the model is unchanged from Figure 3.

Family Stresses / Resources

\ Child Psychosocial
\'. Parental Characteristics

S gender

Peer Interactions'^

Figure 4
Factors Affecting Adjustment: Late Elementary School

Figure 5 illustrates the variables influencing the psychosocial adjustment of the

adolescent. The basic structure of Figure 4 remains. Females exhibit more

internalizing and self-concept problems than males at this age. Therefore, gender now

moderates all effects in the model which are related to internalizing problems, in

addition to those related to conduct problems. In addition, the effect of attractiveness

on self-concept is stronger for females than for males. A new variable is also added at

this age-"denial." During adolescence, appearance is extremely important to the

individual's social functioning and self-concept. Denial of severity of disfigurement

develops as a defense mechanism to preserve self-esteem. Use of denial is influenced

by appearance; the most severely disfigured are the most likely to use denial.
Withdrawal may be necessary to maintain denial by limiting social feedback,

consequently, inhibition is increased with increased use of denial. Other elements of

the model are unchanged.

Family Stresses / Resources

denial -

\ Child Psychosocial

/\ gender
/ llarental Characteristics

/ r
Peer Interactions

Figure 5
Factors Affecting Adjustment: Adolescence


The model presented above is complicated by the wide age range and large

number of variables it encompasses. Many of the hypothesized relationships are based

on sparse empirical evidence or attempts to fill gaps in current knowledge with logical

speculation. The model is therefore proposed as a framework for a program of

research, by the end of which it may be altered appreciably. In the process, much may

be learned about the problems faced by children with craniofacial anomalies and their

families throughout the developmental stages of the child. It is hoped that this

knowledge may eventually be used to identify children who are particularly at risk for

psychological maladjustment, as well as factors which are conducive to good

adjustment. This knowledge may ultimately be useful not only to psychologists

working with this population, but to plastic surgeons, orthodontists, speech therapists,

and other professionals who must make treatment decisions by balancing the time,

discomfort, and financial costs of their interventions with potential benefits to the child.

Knowledge of which variables are most salient to a child's adjustment at each stage of

his or her development would clearly be helpful in making such decisions. To this end,

the study outlined below represents the first step of a research program designed to

identify risk and resiliency factors which are important to the psychosocial adjustment

of male and female children with craniofacial anomalies from birth through


The objective of this study is to examine the impact of facial attractiveness,

speech disorders and hearing impairment on the psychosocial adjustment of children

with craniofacial anomalies during the ages of eight through 13 years. These variables

were chosen both because craniofacial anomalies are frequently associated with

disfigurement, speech deficits, and hearing loss, and because all three variables have

been associated with psychosocial adjustment difficulties in individuals without

craniofacial anomalies. The inclusion of all three of these variables represents an

improvement over much of the current literature, which typically focuses on one

variable at a time, without taking into account the effects of the other variables on

children's adjustment Effects of one variable may therefore be masked due to failure

to control for variations in another variable, and interactions between variables cannot

be assessed.

The age range of the subjects was determined by a logical partitioning of the

model into those ages at which changes in variable effects are noted in the literature.

Specifically, middle childhood is an age range during which sex differences begin to

become more apparent than previously, but the social pressures of adolescence are not

yet in full effect. It is important that future studies also be derived from models which

account for age and sex effects. If they fail to due so, effects may be obscured due to

their differential impact on children of different ages or sex. In addition, use of such a

model for structural research will allow the results to be more effectively integrated

across studies. Present studies of the adjustment of children with craniofacial

anomalies tend to stand in isolation. They provide snapshots of children at diverse

ages, but there has been little attempt to unite their findings in a manner which would

portray the story of the child's adjustment from birth through adulthood. It was the

purpose of the investigator to use the results of this study to verify or modify the

proposed model, which could then be used as a basis for further studies of the

psychosocial adjustment of children with craniofacial anomalies.


The hypotheses of this study were as follows:

1) It was hypothesized that there would be a negative relationship between
internalizing behavior problems and the child characteristics of physical
attractiveness, quality of speech, and hearing level.

2) It was hypothesized that the relationship between internalizing behavior
problems and the child characteristics of physical attractiveness, quality of
speech, and hearing level would be mediated by the quality of peer interactions.

3) It was hypothesized that there would be a higher level of externalizing
problems among males than females.

4) It was hypothesized that physical attractiveness, hearing level, and
speech quality will be would be negatively related to externalizing behavior

5) It was hypothesized that the relationship between externalizing behavior
and the child characteristics of physical attractiveness, quality of speech, and
hearing level would be mediated by the quality of peer interactions.


Subjects and Settings

Sixty-one children with craniofacial anomalies (31 male, 30 female) and 65

children without anomalies (32 male, 33 female) were subjects in this study.

Subjects with craniofacial anomalies were recruited from two outpatient

craniofacial clinics in two Florida cities. Subjects were screened for mental and

physical disability through information obtained from their medical records. Children

who were enrolled in educable or trainable mentally handicapped classes or who had

significant physical disabilities such as blindness or inability to walk were not asked to

participate in the study. Of those approached, eight declined and two were not tested

because the child could not read. In addition, data for two individuals were discarded

because the children were later determined to have significant physical disabilities. Of

those who declined, three parents had concerns about their child being photographed,

two parents reported previous negative experiences with research, one child reported

feeling ill, and two children stated that they did not want to participate. Data from six

children were discarded because of equipment problems and from two children because

hearing data were unavailable. Subjects ranged in age from 7 years, 6 months to

13 years, 11 months. Mean age was 10 years, 9 months (SD = 21 months). Annual

household income ranged from 2892 to 200,000 dollars (M= 34630, SD = 34,331,

n = 38). Caucasians accounted for 90.16 percent of clinic subjects, 6.56 percent were

African-American, and 3.28 percent were of other racial groups. Twenty-three clinic

subjects (37.71%) were enrolled in special classes for speech, learning problems, or

behavior problems (see Table 1). Twenty-six children (43%) presented with cleft lip

and palate, 5 (8 percent) had cleft lip only, 16 (26%) had cleft palate only, and 14

(23%) had other diagnoses such as Crouzon's syndrome or facial asymmetries.

Table 2 gives a complete list of the conditions represented. Clinic subjects reported a

mean of 4.9 hospitalizations (n=57, SD =6.552).
Children without craniofacial anomalies were recruited from two private schools

in the same city as one of the clinics. Of 158 parents contacted by letter and phone

follow-up, 74 (47%) gave written consent for their child to participate. One child was

not tested due to scheduling difficulties. Data for eight children were discarded because

parents did not return measures to the researcher. Non-clinical subjects ranged in age

from 8 years 0 months to 13 years, 11 months. Mean age was 10 years, 9 months

(SD = 16 months). Annual household income ranged from 6150 to 200,000 dollars

(M = 50516, SD = 31450, n=50). Caucasians accounted for 90.77 percent of

non-clinic subjects, 7.69 percent were African-American, and 1.54 percent were of

other racial groups. One subject was enrolled in a special class for reading problems.

Non-clinic subjects reported a mean of .5 hospitalizations (a_= 64, SD=.96).

The craniofacial and control groups did not differ in age (1 (124) = .24, p = .82)

or racial composition (Mann Whitney U= 1997.50, p = .886). The control group

reported significantly higher annual household income than the group with craniofacial

anomalies (1 (78) = -2.35, p = .02).

Table 1
Percentage of Children in Special Classes Craniofacial Group (n = 61)

Type of Class Percntage

Speech 16.39
Learning Problems 6.56
Waiting Placement (Learning) 1.64
Speech and Learning Problems 8.20
Behavioral 1.64
Unspecified 3.28
Total 37.8

Table 2
Craniofacial Diagnoses

Diagnosis Number Percent

Cleft Lip and Palate 26 42.62

Cleft Palate 16 26.23

Cleft Lip 5 8.20

Crouzon's Syndrome 3 4.92

Ectodermal Dysplasia and CP or CLP 2 3.28

Pierre-Robin Syndrome 2 3.28

Crouzon's with Apert's Components 1 1.64

Rhomberg's Disease 1 1.64

Maxillary Hypoplasia 1 1.64

Facial Asymmetry and Mosaic Trisomy 1 1.64

Hemifacial Microsomia and CLP 1 1.64

Macrostomia with Dental and Tear Duct Anomalies 1 1.64

Bi-coronal Synostosis 1 1.64


Achenbach Child Behavior Checklist (CBCL)

The CBCL (Achenbach, 1991) is a parent report measure of child internalizing

behavior problems, externalizing behavior problems, school performance, and social

competence. The CBCL was normed on a sample of 1300 children and has good

psychometric properties. Test-retest reliability and inter-parent agreement coefficients

have exceeded r= .90. The social competence score of the CBCL was used to assess

positive aspects of peer interactions. It has correlated positively with the positive

interactions factor ( = .47, p <.01) and the extensiveness of peer interactions factor

( = .48, p <.01) of the Friendship Questionnaire, which is described below (Bierman

& Mc Cauley, 1987).

Harter Self-Perception Profile for Children (Harter)

The Harter (Harter, 1982) provides six factor scores assessing the child's

perception of academic competence, social competence, athletic competence, physical

appearance, conduct, and general self-worth. It was normed on total of 1543 children

in grades three through eight Chronbach's alpha for each of the subscales has ranged

from .71 to .86 (Harter, 1982). An advantage of the Harter over other measures of

child self-esteem is its inclusion of a separate global self-worth scale, rather than

summing all items to derive a total self-worth score. On this measure, children who

know they perform well in school and sports may still report being generally

dissatisfied with themselves, and children who admit to being unattractive or poor at

sports may still report feeling generally pleased with themselves.

Children's Depression Inventory (CDI)

The CDI (Kovacs, 1985) is a 27 item child-report measure designed to detect

depression in children ages 8-17 years. Internal consistency based on a sample of 870

children was r = .87, and nine-week test-retest reliability on a sample of 90 children

was I = .84 (Kramer & Conoley, 1990, p. 47). One-week test-retest reliability was r

= .87 for a group of 28 emotionally disturbed children and I =.38 for a group of 69

normal fifth and sixth graders (Saylor, Finch, Spirito & Bennett, cited in Kramer &

Conoly, 1990). High scores on the CDI have been found to be related to independent

diagnoses of clinical depression in 102 children undergoing psychiatric treatment

(Cantwell & Carson, cited in Kramer & Conoly, 1990).

Revised Children's Manifest Anxiety Scale (RCMAS)

The RCMAS (Reynolds & Richmond, 1985) is a 37 item self-report measure of

anxiety for children ages 6-19. It was standardized on 4972 children from 13 states.

Internal reliability estimates range from r = .79 to .85. Three-week test-reset reliability

is reported to be r = .98. Nine month test-reset reliability based on a sample of 534

children was = .68. Convergent validity was demonstrated by a r= .85 correlation

of the RCMAS with the trait scale of the State-Trait Anxiety Inventory for Children

(Conoly & Kramer, 1989).

The Friendship Ouestionnaire

The Friendship Questionnaire (Bierman & Mc Cauley, 1987) is a self-report

measure of children's interactions with peers. The negative interactions factor of the

scale consists of 16 items describing negative peer interactions in the home and school

settings. Item examples are "Is there someone who teases you or makes fun of you -

how often?;" "Is there someone you get mad at how often?." The Friendship

Questionnaire also contains a positive peer interaction factor and an extensiveness of

peer interactions factor. Only the negative interactions factor was used because it

correlates significantly with sociometric ratings for children in grades three through

six, whereas the other factors do not correlate significantly with peer nominations for

children below the fifth grade. In a sample of 176 school children ages 8 through

13 years, the negative interactions factor of the Friendship Questionnaire was

significantly related to both negative peer nominations (r = .24, p <.01) and score on

the Roster and Ratings Scale (Roistacher, 1974), a measure on which classmates rate

each other on a five point scale ranging from 1 "not friends" to 5 "best friends"

(I= .25, p <.01) (Bierman & Mc Cauley, 1987). Higher scores on the negative
interactions factor indicate more negative peer interactions.

Hearing Measure

For the group with craniofacial anomalies only, each subject's speech reception

threshold (SRT) was obtained from the child's medical chart as a measure of the

child's hearing loss. The SRT is the decibel level at which an individual is able to

correctly repeat 50 percent of stimulus words spoken to him. Spondee words (words

with two equally stressed syllables) were used as the stimuli for measurement of the

SRT. GSI 10 and Belltone 2000 audiometers were used to regulate the decibel level of

the spondee words during clinical administration of the assessment, which was

performed by certified audiologists with previous experience assessing children with

clefts. Normal hearing is indicated by an SRT of 0-20 dB; 20-40 dB is indicative of

mild hearing loss; 40-60 dB is classified as moderate hearing loss; 60 to 70 dB falls in

the moderate-severe range; 70-90 dB is categorized as severe hearing loss; and SRT

greater than 90 dB is indicative of profound hearing loss. The SRT in the child's better

ear was used as the measure of hearing level, as is customary in the hearing loss

literature. All children in the non-clinical sample had normal hearing by parent report.

Speech Measure

Speech samples were recorded during the subject's clinic visit or at the child's

school. Subjects read a brief paragraph and seventeen sentences which had been

chosen by consultation with two speech pathologists to provide a sample of speech

varied enough to reveal most problems of articulation and hypemasality which may be

associated with craniofacial anomalies. All speech samples were rated on a visual

analog scale by three speech pathologists (Appendix A). The scale was weighted at

each end with the descriptors "very unacceptable" and "very acceptable." The average

of the speech pathologists' ratings were used as the child's speech score. Glass and

Starr (1979) reported intrerrater reliability estimates for this method ranging from

r = .81-.86, and Sinko and Hedrick (1982) reported interrater reliability ofr = .98

using this method. High reliability for this method of speech rating has also been

reported using as few as two speech pathologists as raters (Estes & Morris, 1970). In

that study, the two raters assigned identical scores (on a seven-point scale) to seven of

eleven children and assigned ratings of no ore than one point difference to the

remaining four children. Speech intelligibility was rated at the same time as speech

acceptability, using a visual analog scale anchored with the descriptors "very

intelligible" and "very unintelligible."

Physical Attractiveness Measure

The subject's appearance was rated by a plastic surgeon, an orthodontist,and

an oral and maxillofacial surgeon. Subjects were photographed from the shoulders up,

wearing a drape over their clothing and stocking cap over their hair, without jewelry.

All photographs were taken against the same background, and subjects assumed a

neutral facial expression. The photographs were rated on a visual analog scale,

weighted at each end with the descriptors "very unattractive" and "very attractive"

(Appendix B). The average of the professionals' ratings was used as the child's

attractiveness score. Adequate reliability has been reported for this method in

numerous studies (e.g., Dion & Berscheid, 1974; Farina, Bums, Austad, Bugglin &

Fischer, 1986; Reis et al., 1982; Rich, 1975; Richman, Holmes & Eliason, 1985).

Medical Experiences Questionnaire

The Medical Experiences Questionnaire (see appendix C) is a six-item

questionnaire on which the parent listed the number and recency of child

hospitalizations, child reaction to past dental and medical experiences, medical and

dental procedures planned for the near future, anticipated child reactions to upcoming

procedures, and the child's current anxiety about upcoming procedures. It was used as

a control measure for effects of medical and dental experiences. Three of the items

were taken from a measure by Johnson and Baldwin (1969). Similar parent

questionnaires have been found to predict behavioral observations of children's distress

during medical and dental procedures (Dahliquist et al., 1986; Johnson & Balwin,

1969) and post-surgery scores on the Behavior Problem Checklist (Melamed & Siegel,

1975), suggesting that they tap child anxiety and distress related to medical and dental



Subjects with craniofacial anomalies were asked to participate while waiting for

appointments at the craniofacial clinics. Questionnaires were completed and speech

and appearance samples were obtained in the child's outpatient clinic room while the

child was waiting to be seen by various members of the multidisciplinary craniofacial

teams. Child-report measures and speech and appearance samples for children in the

non-clinical group were completed in small groups (3-5) at the children's schools.

Parent-report measures for the non-clinical group were sent home with the children and

mailed to the researcher by the parents. Data for six non-clinical subjects were collected

at the children's homes.


The first section of this chapter contains descriptive statistics, which are

summarized in Table 3. These are followed by gender and group comparisons,

summarized in Tables 4-7. Intercorrelations of variables are presented next (Tables

8 -11), and tests of the hypotheses using multiple regression complete the chapter

(Tables 12-15).

Description of Sample

Distribution of Variables

Means and standard deviations of interval scale variables for each group are

presented in Table 3. Scores were not normally distributed for the following variables:

speech acceptability, CDI, hearing, number of hospitalizations, past reactions to

medical treatment, anticipated reactions to future treatment, anxiety over upcoming

procedures, and raw scores of the CBCL internalizing and externalizing scales. The

remainder of the variables were normally distributed. Speech acceptability was

transformed to a normal distribution using the algorithm: transformed speech =

log (160-speech acceptability). The remainder of the non-normally distributed scales

could not be transformed to normal distributions.

-n o r-;S c^ no4 Oo-
S mou^ ado0<'

os I \Q0o
o^ odm"

23l/ S $00

\o- oo \oc

t 00 0 '*T 9 n
, "0 'o -'R

U, ~mh
r- 'n cr o
c \oOo

W m \o-It -
W 00 0qCO

0Ci e U,'C'C C 0r-n 0
00CS D 0 (00\1C o'
n oioo 0ooUno~os
C-*< 'C'^ s- ^

l G (t- G-'c' G--- G--


to C
5 d U
> S U

a L)U u l ^
u u
|~ ~ g I g1 Kg

jgl ~ ss Bas^|i

esa S~aQ|.SS ~cm
8 ZS^Sgu ('~ qua u uu u
*a E l 3 3 ag a
: & uu i5 a
< ~ r S

el I


Mean age in years was 10.75 for the craniofacial sample (SD = 1.75) and

10.77 for the control sample (SE = 1.37). Age range was 7 years, 6 months to

13 years, 11 months.

Medical Experiences

Fifty-four percent of parents of subjects with craniofacial anomalies reported

child hospitalizations within the last two years, 41 percent reported hospitalizations

between birth and age two, and 5 percent reported no hospitalizations. Eleven percent

of parents of control subjects reported child hospitalizations within the last two years,

22 percent reported hospitalizations between birth and age two, 65 percent reported no

hospitalizations, and 2 percent did not respond. Ninety-five percent of parents of

subjects with craniofacial anomalies and 57 percent of parents of control subjects

reported that they expected their child to have medical or dental procedures in the near

future. Past reaction to medical procedures, anticipated reaction to future medical

procedures, and anxiety over upcoming procedures were rated on four point scales,

with higher scores indicating better reactions and less anxiety. The average score for all

three of these variables was above 3 for both groups (see Table 3), indicating that

parents rated the subjects as having moderately good to good reactions and moderately

low to low anxiety about upcoming procedures.

Social Characteristics

Both samples reported lower mean rates of negative peer interactions than did

the groups on which the Friendship Questionnaire was developed (see Table 3). (Third

and fourth grade norm group M = 37.41, SD = 9.56 ,n = 94; fifth and sixth grade

norm group M = 32.88, SD =8.53, n= 82.) Similarly, both groups scored in the

normal range of the CBCL social competence scale.

Speech. Attractiveness and Hearing Ratings

The average interrater correlation for speech acceptability ratings for the total

sample (craniofacial and non-clinical) was r =.63. The average intercorrelation for the

craniofacial sample was r =.65, and it was =.34 for the control group. The lower

average intercorrelation for the control group was probably due to the limited variability

of the scores for that group when compared with the craniofacial group. Possible

scores on this measure ranged from 0-153. Speech acceptability ratings ranged from

16.0 to 143.67 for the group with craniofacial anomalies (M-98.88, SD=30.31).

Speech acceptability ratings for the non-clinical group ranged from 77.67 to 146.33

(M= 125.59, SM= 14.79).

The mean interrater correlation for attractiveness for the total sample

(craniofacial and non-clinical) was = .61. This compares favorably with the weighted

mean interrater correlation for attractiveness of r =.54 which was reported in a recent

meta-analysis of attractiveness studies (Feingold, 1992). The average intercorrelation

for the craniofacial sample was r =.67, and it was =.37 for the control group. In the

attractiveness as well as the speech ratings, the lower average intercorrelation for the

control group was probably due to the limited variability of the scores for that group

when compared with the craniofacial group. Possible scores on this measure ranged

from 0-153. Attractiveness ratings ranged from 24.33 to 118.0 for the group with

craniofacial anomalies (M=60.95, S2=17.68). Attractiveness ratings for the non-

clinical group ranged from 49.67 to 125.33 (M= 81.92, SD= 15.75). Hearing level

for the group with craniofacial anomalies ranged from SRT= 0 to SRT = 65. The

mean SRT was 10.97 (SE = 13.64), which is in the normal range of hearing, and

80.33% of subjects were classified as normal hearing. All non-clinical subjects were

reported to have normal hearing.

Child Adjustment Measures

Mean child-reported levels of depression, anxiety, and self-esteem were in the

normal range for both samples (see Table 3). Mean parent-reported levels of

internalizing and externalizing behavior problems were also within the normal range.

Gender Comparisons

Within each sample, gender comparisons were made for child characteristics,

medical experiences, social characteristics, and child adjustment variables. For

variables with normal distributions, t-tests were performed. Mann-Whitney U tests

were performed on variables with non-normal distributions. Results are presented in

Tables 4 and 5. Males in the control group scored significantly higher than females on

the Negative Interactions Factor of the Friendship Questionnaire, indicating that they

reported a higher level of negative peer interactions. No other gender differences were

significant. The hypothesis that males would have higher externalizing behavior

problems than females was not supported. Because no gender differences were

detected in the main variables of interest (child characteristic and child adjustment

variables), data from male and female subjects were combined in all subsequent


Group Comparisons

Between-group comparisons were made for the child characteristics, medical

experiences, social characteristics, and child adjustment variables. For variables with

normal distributions, t-tests were performed. Mann-Whitney U tests were performed

on variables with non-normal distributions. Results are presented in Tables 6 and 7.

The control group was rated as significantly more attractive and as having more

Soo m
cd os aoomr

c r 0r r

m3 m aomm n
rN-- mC m-mn

c cU mmm
CL. coo oOrs s
oI I u

u w z ur-

C C4
5 0 s-a s^

| U^az gf asz


| 888.
r k r-ir 2

00 N000

C! en W)l\0ri "
~1 ocr-m

S ''| ooo o'/^o
00 t- m

en (n el M^ mw

C, C

U~ U
1.3 ii ii

I >3 uuu5 uu

u u

** *

Fjl vi 00 Cl 0
l m 000v

-c C4 too
tn in m o* VII t


I >nO "O C A o
1 0 80

o Qo
Q| p oui

2 ---

I I ao

^ gv



acceptable speech than the group with craniofacial anomalies. The group with

craniofacial anomalies reported significantly more hospitalizations than the control

group. Additionally, the control group scored significantly higher on the social

competence scale of the CBCL. There were no group differences in scores on the child

adjustment variables.
Intercorrelations of Variables

Pearson correlations were computed for normally distributed variables (Tables 8

and 10), and Spearman correlations were computed for non-normally distributed

variables (Tables 9 and 11). Variables with relatively strong associations are reported

below. A large number of correlations were computed; therefore, many of these

associations may be due to chance. They are presented as an aid in interpreting the

regression analyses which follow.

Craniofacial Group

In the group with craniofacial anomalies, more acceptable speech was

associated with higher parent ratings of social competence on the CBCL (r = -.31.)

The negative r value represents a positive relationship between speech quality and

social competence, because the speech scores were log transformations. More attractive

children rated themselves as less anxious on the RCMAS (r =-.27.) In addition,

increased hearing loss was associated with lower ratings of attractiveness (r = -.49.)
Negative peer interactions (Negative Interactions Factor of the Friendship

Questionnaire) were positively correlated with self-rated anxiety (RCMAS) ( = .39)

and depression (DC) (r = .42) and negatively correlated with self-esteem (Harter) ( =

.37). Social competence (the Social Competence subscale of the CBCL) was

negatively related to self-reported anxiety (RCMAS) ( = -.32) and depression (DC) (r

= -.58) and to parent-reported externalizing behavior problems (Externalizing subscale

of the CBCL) (r = -.42). Additionally, social competence was positively associated

with self-esteem (Harter) ( = .42). Self-reported anxiety (RCMAS) was negatively

related to self-esteem (Harter) (r = -.44) and positively associated with child-reported

depression (DC) ( = .61). In addition, the child-reported anxiety (RCMAS) was

negatively related to parent-reports of how well children reacted to past medical

procedures ( = -.36) and how well the parent thought the child would react to future

procedures (r = -.41) (Medical Experiences Questionnaire).Parent-reported child

internalizing behavior problems (CBCL Internalizing Scale raw score) and externalizing

behavior problems (CBCL Externalizing Scale raw score) were positively associated

(r= .62).

On the Medical Experiences Questionnaire, parent-reported child reactions to

past procedures and anticipated reactions to future procedures were positively related

(t = .69). On the same measure, anxiety over upcoming procedures and anticipated
reactions to future procedures were also positively associated (C = .53). In addition,

reactions to past procedures and anxiety over upcoming procedures (r = .40) were

positively associated with each other (higher scores indicated better reactions and less

anxiety). Child-reported depression (CDI) was negatively associated with how well the

parent thought the child would react to future medical or dental procedures (Medical

Experiences Questionnaire) (i = .40), and recency of the latest hospitalization was

positively associated with the total number of child hospitalizations (r = .42). The

negative r value indicates a positive relationship between number of hospitalizations and

recency of last hospitalization because a lower number indicates a more recent


Control Group

In the control group, less acceptable speech was associated with higher anxiety

ratings on the RCMAS (I = .27) and poorer anticipated reactions to upcoming medical

procedures (r = -.36). Contrary to predictions, there was a strong negative association

between attractiveness and social competence as measured by the CBCL (r = -.42). In


I 8



I 0
0 -

l all 8 s %N

q q ,



z 8


o~ ~ ~ S-;o 0 t

- 0 -

00 N -^
- 0 M -i
* Ns

-i N ?;

N N N -?

0 0 0,
Qj N. 8 00 2 *

^ 8p' i,
i a
^~4 .2 wxoo e

r^ Z C~i '< 33 U QS Q



0, S

N -
-3 N

addition, older age was associated with lower ratings of physical attractiveness in this

group ( = -.281 and with higher parent-rated social competence (Social Competence

subscale of the CBCL) (r = .33). Self-esteem (Harter) was positively related to social

competence (r = .29) and negatively related to child-reported anxiety (RCMAS) (=

-.50) and depression (CDI) ( = -.69). Child-reported anxiety (RCMAS) and

depression (CDI) were positively correlated (r = .71). Parent-reported child

internalizing behavior problems (CBCL Internalizing Scale raw score) and externalizing

behavior problems (CBCL Externalizing Scale raw score) were positively correlated

(I= .56).

On the Medical Experiences Questionnaire, parent-reported child reactions to

past procedures and anticipated reactions to future procedures (I = .63) were positively

associated. Anxiety over upcoming procedures and anticipated reactions to future

procedures (r = .51) were also positively associated. In addition, reactions to past

procedures and anxiety over upcoming procedures (r= .50) were positively correlated

with each other (higher scores indicated better reactions and less anxiety), and recency

of the latest hospitalization was positively associated with the total number of child

hospitalizations (I = -.64). The negative r value indicates a positive relationship

between number of hospitalizations and recency of last hospitalization because a lower

number indicates a more recent hospitalization.

Multiple Regression Analyses

Craniofacial Group

Four stepwise multiple regressions were conducted to test the hypothesis that

there would be a negative relationship between internalizing behavior problems and the

child characteristics of physical attractiveness, quality of speech, and hearing level.

The CDI, RCMAS, Harter General Self-Concept Scale, and raw score of the


.~ .~ ~

-~" 'U ~


0 a '

- 0 a

B 0

N N~




~ r- N


'3% -- '

'8 .3 *:

!a .

- 0
- N -


0 0

rN N 0

3 o o

1 0 IC
5 PC
M .

t] ]

Internalizing Scale of the CBCL were regressed on physical attractiveness, hearing

loss, and the transformed speech ratings (Table 12). The transformed speech ratings

entered into the equations predicting child-reported depression (CDI) and self-esteem

(Harter), and physical attractiveness entered into the equation predicting child-reported

anxiety (RCMAS) and self-esteem, but they did not account for a significant proportion

of the variance in these child adjustment measures. No other child characteristics

entered any of the regressions to predict child internalizing behavior problems in the

group with craniofacial anomalies. The hypothesis that there would be a negative

relationship between internalizing behavior problems and the child characteristics of

physical attractiveness, quality of speech, and hearing level was not supported in this


The hypothesis that the relationship between internalizing behavior problems

and the child characteristics of physical attractiveness, quality of speech and hearing

level would be mediated by the quality of peer interactions could not be tested because

the child characteristics did not predict child internalizing behavior problems in this

group. Because the social variables may also be conceptualized as outcome variables

in a model of child psychosocial adjustment, the Negative Peer Interactions Scale of the

Friendship Questionnaire and the Social Competence Scale of the CBCL (raw score)

were regressed on the child characteristics of physical attractiveness, quality of speech,

and hearing level (Table 13). Transformed speech accounted for 11.6% of the

variance in social competence (o < .05). The negative beta weight indicates a positive

relationship between speech quality and parent-reported social competence because the

speech score was a log linear transformation. Hearing loss predicted child-reported

negative peer interactions, accounting for 8.1% of the variance (, < .05).

In order to test the hypothesis that the child characteristics of physical

attractiveness, hearing level, and speech quality would be negatively related to

externalizing behavior problems, the Externalizing Scale of the CBCL was regressed on






U3 9

| s


I '" I a *

N U U1
C, 78 8<> r



J-" 8 i

a ? M 00

.0 o


P c-
uc (
IS a >


S < a *

physical attractiveness, hearing level, and the transformed speech scores (Table 12).

None of the child characteristics entered into the equation, so the hypothesis was

unsupported. The hypothesis that the relationship between externalizing behavior

problems and the child characteristics of physical attractiveness, quality of speech, and

hearing level would be mediated by the quality of peer interactions could not be tested

because the child characteristics did not predict child externalizing behavior problems in

this group.

Control Group

The interrater reliability of the speech and attractiveness ratings was low for the

control group. The results of these analyses should therefore be viewed with caution,

but are presented for the reader's consideration and as possible material for replication


Four stepwise multiple regressions were conducted to test the hypothesis that there

would be a negative relationship between internalizing behavior problems and the child

characteristics of physical attractiveness, quality of speech, and hearing level. The

CDI, RCMAS, Harter General Self-Concept Scale, and raw score of the Internalizing

Scale of the CBCL were regressed on physical attractiveness, hearing loss, and the

transformed speech ratings (Table 14). The transformed speech ratings predicted

child-reported anxiety (RCMAS), accounting for 7.1 percent of the variance in scores

on this measure. This relationship was significant (p <.05). The positive beta weight

for the transformed speech scores represents a negative relationship between speech

quality and anxiety, because the speech scores were log transformations. Physical

attractiveness entered into the equations predicting child-reported anxiety (RCMAS)

and depression (CDI), but did not account for a significant proportion of the variance

in either of these child adjustment measures. No other child characteristics entered any

of the regressions to predict child internalizing behavior problems.





8 I

N a
o C

- 0S




h I

Baron and Kenny (1986) state that, in order to prove that one variable mediates

the effects of an independent variable on the outcome variable, three regressions should

be conducted: "first, regressing the mediator on the independent variable; second,

regressing the dependent variable on the independent variable; and third, regression of

the dependent variable on both the independent variable and on the mediator" (p. 1177).

Furthermore, three conditions must be met:

First, the independent variable must affect the mediator in the first

equation; second, the independent variable must be shown to affect

the dependent variable in the second equation; third, the mediator

must affect the dependent variable in the third equation (p. 1177).

Therefore, to test the hypothesis that the relationship between child-reported

anxiety (RCMAS) and quality of speech was mediated by peer interactions, the

Negative Peer Interactions Scale of the Friendship Questionnaire, and the Social

Competence Scale of the CBCL (raw score) were regressed on the transformed speech

ratings (Table 15). The transformed speech ratings did not enter into either regression

predicting negative peer interactions or social competence. The hypothesis that the

relationship between anxiety (RCMAS) and the child characteristic of quality of speech

would be mediated by the quality of peer interactions was therefore unsupported.

Because none of the other measures of internalizing behavior problems were

predicted by the child characteristics, social characteristics could not be tested as

mediators between child characteristics, and scores on those measures but were instead

treated as outcome measures and regressed on the child characteristics of speech quality

and physical attractiveness. Neither child characteristic entered into the equation

predicting negative peer interactions. In a stepwise regression to predict social

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
INGEST IEID E6EMT6TNH_CZ854T INGEST_TIME 2012-09-24T12:51:23Z PACKAGE AA00011814_00001