Journal of Abnormal Child Psychology, Vol. 26, No. 2, 1998, pp.

83-94

Social Anxiety Amon Adolescents: Linkages with Peer

Relation s and Friendships

Annette M. La Greca1,2 and Nadja Lopez1

Received September 12, 1996; revision received March 21, 1997; accepted March 24, 1997

This study examined the utility of modifying the Social Anxiety Scale for Children—Revised
(SASC-R) for use with adolescents, and examined associations between adolescents' social
anxiety (SA) and their peer relations, friendships, and social functioning. Boys (n = 101)
and girls (n = 149) in the 10th through 12th grades completed the Social Anxiety Scale for
Adolescents (SAS-A) and measures of social support, perceived competence, and number
and quality of their best friendships. Factor analysis of the SAS-A confirmed a three-factor
structure: Fear of Negative Evaluation, Social Avoidance and Distress in General, and Social
Avoidance Specific to New Situations or Unfamiliar Peers. Girls reported more SA than
boys, and SA was more strongly linked to girls' social functioning than boys'. Specifically,
adolescents with higher levels of SA reported poorer social functioning (less support from
classmates, less social acceptance), and girls with higher levels of SA reported fewer friend-
ships, and less intimacy, companionship, and support in their close friendships. These find-
ings extend work on the SASC-R to adolescents, and suggest the importance of SA for
understanding the social functioning and close friendships of adolescents, especially girls.
KEY WORDS: Social anxiety; adolescents; peer relations; friendships; Social Anxiety Scale for chil-
dren—Revised (SASC-R); Social Anxiety Scale for Adolescents (SAS-A); friendships; romantic appeal;
social functioning; anxiety disorders.

Adolescents' relationships with friends and peers adults' interpersonal behavior and psychological
play a critical role in the development of social skills functioning (Leary, 1983), the study of social anxiety
and feelings of personal competence that are essen- among children and adolescents is in its very early
tial for adult functioning (Ingersoll, 1989). Moreover, stages. The present study extended the existing work
peer relationships appear to be instrumental in fa- on social anxiety in several ways.
cilitating adolescents' sense of personal identity and First, in the absence of a well-validated measure
increasing their independence from family influences of social anxiety for adolescents, the utility of the So-
(Dusek, 1991; Ingersoll, 1989). Consequently, factors cial Anxiety Scale for Children—Revised (SASC-R)
that inhibit or impede adolescents' interpersonal for assessing adolescents' social anxiety was evalu-
functioning represent a critical area for clinical and ated (La Greca & Stone, 1993). Measures of social
developmental investigation. In this connection, so- anxiety have been developed and validated primarily
cial anxiety may be an important factor to examine for children (Hymel, & Frank, 1985; La Greca &
among adolescents. Stone, 1993) or adults (Watson & Friend, 1969), but
Although social anxiety has long been recog- not specifically for adolescents. In this regard, it was
nized as an important factor for understanding of interest to determine whether the SASC-R could
be modified for use with adolescents. Published re-
ports using the SASC-R have predominantly focused
1University of Miami, Coral Gables, Florida 33124. on children in Grades 2 through 6 (La Greca &
2Address all corresponsence to Annette M. La Greca, Department Stone, 1993; Silverman, La Greca, & Wasserstein,
of Psychology, P.O. Box 249229, Coral Gables, Florida 33124. 1995), or young adolescents (seventh and eighth

83
0091-0627/98/0400-0083$15.00/0 6 1998 Plenum Publishing Corporation
84 La Greca and Lopez

graders; Vernberg, Abwender, Ewell, & Beery, 1992). 1991) or alcohol abuse (e.g., Schneier, Martin, Lie-
In the present study, the utility of the SASC-R was bowitz, Gorman, & Fyer, 1989). In view of these con-
evaluated for older adolescents (those in Grades 10 cerns, a primary goal of this study was to evaluate
through 12). an adolescent version of the SASC-R (referred to as
Extending the construct of social anxiety to this the Social Anxiety Scale for Adolescents, or SAS-A),
older adolescent age group was important for con- by examining its factor structure and psychometric
ceptual as well as practical reasons. The SASC-R was properties.
initially developed from a model of social anxiety, de- A second goal of this study was to evaluate gen-
rived from studies of adults, that included compo- der differences in adolescents' social anxiety. Among
nents of social evaluative anxiety (fear of negative children (Crick & Ladd, 1993; La Greca & Stone,
evaluation or FNE), and social avoidance and dis- 1993) and early adolescents (Vernberg et al., 1992),
tress (SAD) (Watson & Friend, 1969). Work with ele- girls have reported higher levels of social anxiety
mentary school youths revealed that these two than boys, especially for the social evaluative aspect
components of social anxiety were evident among of social anxiety (i.e., fear of negative evaluation
children as young as 7 years of age (La Greca, Dan- from peers). Similarly, one might expect adolescent
des, Wick, Shaw, & Stone, 1988). For children, how- girls to be more worried than boys about peers' nega-
ever, the SAD component was further differentiated tive evaluations of them. Indeed, surveys have found
by avoidance and distress that was specific to new that adolescent girls are more concerned than boys
situations or unfamiliar peers (SAD-New), and about others' judgments of their appearance and be-
avoidance and distress that was generally experi- havior (Nolen-Hoeksema & Larson, 1992; Rosen &
enced in the company of peers (SAD-General) (La Aneshensel, 1976). Moreover, rates of internalizing
Greca & Stone, 1993). Conceptually, it was of inter- problems are higher among girls than boys (Bern-
est to determine whether the FNE and SAD com- stein, Garfinkel, & Hoberman, 1989; Kashani, Or-
ponents of social anxiety would also be applicable to vaschel, Rosenberg, & Reid, 1989; Quay & La Greca,
adolescents. If so, this would provide some develop- 1986), and adult women are about twice as likely to
mental continuity for these components of social be socially phobic as men (Schneier et al., 1992).
anxiety. Furthermore, from a practical perspective, Thus, adolescent girls may be more vulnerable than
extending the SASC-R to adolescents would provide boys to feelings of social anxiety, and this may have
a means of assessing social anxiety across a broad implications for their social functioning.
age range, thereby facilitating developmental and This leads to the third important goal of this
longitudinal investigations of children's social and af- study, which was to evaluate linkages between ado-
fective functioning. lescents' social anxiety and their interpersonal func-
Clinically, the availability of an instrument such tioning with peers. Two specific aspects of inter-
as the SASC-R may also prove useful for examining personal functioning were of interest: general levels
the onset and course of certain anxiety disorders, of acceptance from peers and close friendships. For
such as social phobia. Social phobia (Diagnostic and many adolescents, the importance of peer group ac-
Statistical Manual of Mental Disorders, 4th ed., or ceptance increases with age, and peaks in mid to late
DSM-IV, American Psychiatric Association, 1994) adolescence (Brennan, 1982). Also during the ado-
has been viewed as an extreme form of social anxiety, lescent years, interest in romantic attachments and
affecting up to 7% to 8% of the adult population opposite-sex relations introduces a new dimension to
(Hazen & Stein, 1995). Although little is known social functioning; namely, the desire to be accepted
about the etiology of social phobia, reports indicate as a romantic partner (Harter, 1988; Kuhlen &
that the disorder begins during childhood or adoles- Houlihan, 1965). One might expect socially anxious
cence, with a mean age of onset between 15 and 16 adolescents to perceive their general social accep-
years of age (Schneier, Johnson, Hornig, Liebowitz, tance and their romantic appeal to be low. Percep-
& Weissman, 1992). Thus, an instrument such as the tions of social exclusion from the peer group may
SASC-R may be useful to investigators interested in directly contribute to feelings of anxiety (Leary,
studying the etiological pathways for social phobia or 1990); in addition, feelings of social anxiety might
for other disorders that are believed to have a sub- limit adolescents' interactions with peers, or inhibit
stantial social component, such as eating disorders their dating and romantic attachments, thereby in-
(e.g., Bulik, Beidel, Duchmann, Weltzin, & Kaye, terfering with their social functioning. Although link-
Adolescents' Social Anxiety 85

ages between social anxiety and adolescents' peer re- ceived support from significant adults (i.e., parents,
lationships have not been studied, previous work has teachers), as well as their self perceptions in nonso-
revealed that anxious children are less well liked by cial areas of competency (i.e., scholastic competence
their peers (Strauss, Frame, & Forehand, 1987), and and behavioral conduct). Social anxiety was not ex-
that children who are rejected or neglected by their pected to be highly related to support from adults,
peers are more socially anxious than their classmates or to adolescents' perceptions of their competencies
(La Greca et al, 1988; La Greca & Stone, 1993). in non-social areas. These associations were explored
Thus, the present study extended this work on link- to provide some support for the discriminant validity
ages between social anxiety and general peer rela- of the social anxiety construct for adolescents.
tions to an adolescent age group. Specifically,
adolescents with high levels of social anxiety were ex-
pected to report less acceptance and support from METHOD
their peer groups (i.e., classmates) and to perceive
themselves as having less romantic appeal. Participants
Aside from peer group acceptance, close friend-
ships also represent a critical aspect of adolescents' The participants were 250 high school students
interpersonal functioning. Close friendships with (101 boys, 149 girls; Grades 10 through 12), who re-
peers take on increasing importance during adoles- sided in a large Southeastern metropolitan area. The
cence, and serve several important functions, such as ethnic composition of the sample was 51.6% white,
providing companionship, emotional support, inti- 31.6% Hispanic, 15.2% African-American, and 1.6%
macy, and a means of expressing emotions and re- Asian. Adolescents ranged in age from 15 to 18 years
solving conflicts (Berndt, 1982). In this connection, (mean = 17.04, SD = .91), and came from predomi-
feelings of social anxiety—particularly generalized nantly middle-class socioeconomic backgrounds. Spe-
social avoidance and distress—could lead to disen- cifically, 72.3% of the adolescents resided in
gagement from peer interactions (La Greca et al., two-parent families; the median family income fell
1988), and interfere with the development of close, in the $40,000 to 60,000 range; and the majority of
supportive ties. Work with early adolescents (Vern- parents had graduated college (53.8%), with an ad-
berg et al., 1992) provides some support for this no- ditional 42.6% having received high school diplomas.
tion. Among seventh and eighth graders, high levels
of generalized social avoidance and distress at the Procedure
beginning of the school year predicted lower levels
of intimacy and companionship in adolescents' close Adolescents were recruited as part of a larger
friendships months later; this was especially true for study of adolescent peer relations. An unselected
girls. The present study extended this earlier work sample of school children, who were initially assessed
by examining linkages between adolescents' social during the spring of 1988, were tracked through the
anxiety and their reports of the number of close county public school database during the spring of
friends, the quality of these friendships, and their 1994. At that time it was determined that 338 stu-
perceptions of competency in their close friendships. dents remained in the local public schools, and were
Specifically, in comparison to less socially anxious potentially available for participation. For these stu-
adolescents, those with higher levels of social anxiety dents, letters were mailed to their parents at their
were expected to report fewer close friendships, and most recently listed addresses, requesting permission
to perceive these relationships as less intimate and for their adolescent's participation. Letters were fol-
supportive. Furthermore, because girls emphasize in- lowed within 2 weeks by phone calls to the adoles-
timacy and emotional support in their friendships to cents' homes. Of the 338 students, 32 (9%) could not
a greater extent than boys (Berndt, 1982; Berndt & be contacted by repeated phone calls or letters. Of
Perry, 1986; Buhrmester & Furman, 1987), problems the 306 students remaining in the pool, 250 (82%)
in close friendships may be more strongly linked to agreed to participate and 56 (18%) declined partici-
feelings of social anxiety for girls than for boys; this pation. Participants did not differ from those who de-
notion was evaluated. clined to participate with respect to ethnicity, age,
Finally, the present study also examined associa- socioeconomic status, or grade. However, a greater
tions between social anxiety and adolescents' per- percentage of the participants were girls (59.6%),
86 La Greca and Lopez

compared to those who refused participation (46.4% to .86 (FNE). Construct validity was supported by pat-
girls) (p < .05). terns of relationships between SASC-R subscales and
Adolescents were interviewed at home by children's self-appraisals, as well as peer-rated so-
trained research assistants. The measures included ciometric status (La Greca & Stone, 1993).
the SAS-A, the Social Support Scale for Children For the present study, the wording of the SASC-R
and Adolescents, the Adolescent Interview, and the was revised to make it more developmentally appro-
Self-Perception Profile for Adolescents. Adolescents priate for adolescents. Thus, the SAS-A was identical
also completed a brief form requesting demographic in format to the elementary school version (i.e., 22
information, such as age, gender, ethnicity, and fam- items; 5-point rating scale), but the item wording was
ily composition. Parents provided background infor- modified slightly for an older age group. Specifically,
m a t i o n r e g a r d i n g ethnicity, education, and items containing the term "other kids" were reworded
employment. Written informed consent was obtained to "peers," "others," or "people," and references to
from adolescents and parents prior to participation. "playing with" others were reworded to "doing things
with" others. For example, Item 5 of the SASC-R, "I
only talk to kids that I know really well," was changed
Measures to, "I only talk to people I know really well." Item 8
of the SASC-R, "I worry about what other kids think
Social Anxiety Scale for Adolescents (SAS-A ). The of me," was modified to, "I worry about what others
SASC-R (La Greca & Stone, 1993) was modified think of me." (See Table I, in the Results section, for
slightly for use with adolescents. The SASC-R was in- a brief listing of each item.)3
itially developed to assess children's subjective expe- Social Support Scale for Children and Adolescents
rience of social anxiety. It contains 18 descriptive (SSSCA; Harter, 1985). This instrument assesses ado-
self-statements and four filler items reflecting activity lescents' perceived social support from four sources:
preferences ("I like to read") or social preferences ("I parents, teachers, classmates, and close friends. Each
like to play with other kids"). Each item is rated on of the 24 items (six per subscale) uses a forced-choice
a 5-point scale according to how much the item "is format, wherein the adolescent first determines which
true for you" (1 = not at all, 5 = all the time). Based of two statements is most like him or her (e.g., "Some
on factor analytic studies, three distinct subscales have kids have parents who treat their child like a person
been identified. The first subscale, Fear of Negative who really matters BUT other kids have parents who
Evaluation (FNE), reflects fears, concerns, or worries don't usually treat their child like a person who really
regarding negative evaluations from peers; it includes matters.") After choosing between the statements, the
eight items (e.g., "I worry about what other kids think adolescent rates how true it is ("Really true for me"
of me"). In addition, there are two subscales for So- or "Sort of true for me"). This yields a score from 1
cial Avoidance and Distress: SAD-New and SAD- to 4 for each item; subscale scores are obtained by
General. SAD-New reflects social avoidance and averaging relevant items.
distress with new social situations or unfamiliar peers; Prior work with the SSSCA has revealed good
it includes six items (e.g., "I get nervous when I meet psychometric properties. Internal consistencies for all
new kids"). SAD-General reflects more generalized four subscales have been considered satisfactory
or pervasive social distress, discomfort, and inhibition; (range of .72 to .88) (Harter, 1985). Harter (1985) re-
it includes four items (e.g., "I feel shy even with kids ported that the Classmate Support subscale correlated
I know well"). Scores are obtained by summing the significantly with perceptions of social acceptance from
ratings for the items comprising each subscale, and peers (.69); the Parent Support subscale correlated
can range from 8 to 40 for FNE, 6 to 30 for SAD- with the degree to which youngsters' values were con-
New, and 4 to 20 for SAD-General; total scores can gruent with their parents' (.48); and the Close Friend
range from 18 to 90. subscale correlated with youngsters' disclosure of per-
Psychometric support for the SASC-R has been sonal thoughts and feelings to friends (.46). Although
satisfactory (La Greca et al., 1988; La Greca & Stone, few gender differences in social support have been ob-
1993). For example, confirmatory factor analysis re- served, girls reported higher levels of support from
vealed a good fit between the three-factor model of close friends than boys (Harter, 1985).
social anxiety and children's responses (La Greca &
Stone, 1993). Internal consistencies for the subscales
3
ranged from .69 (SAD-General) to .78 (SAD-New) Copies of the SAS-A may be obtained from the first author.
Adolescents' Social Anxiety 87

Number and Quality of Friendships. Adolescents RESULTS

were asked to name each of their very best friends.
This served as an index of the number of close Psychometric Analysis of the SAS-A
friends. Next, they completed the Adolescent Inter-
view (AI) to assess the quality of their three closest To examine the factor structure of the SAS-A, a
friendships. The AI was based on the Friendship In- principal-axis factor analysis with varimax rotation
terview developed by Berndt and Perry (1986), as was conducted with the 18 primary items. This yielded
modified by Vernberg (1990). Based on Vernberg's three factors, each with eigenvalues greater than 1.0,
(1990) scoring, the friendship domains of compan- that together accounted for 60% of the variance in
ionship and intimacy were evaluated. Adolescents the SAS-A. The three factors were identical to the
name up to three best friends and then rate how three primary factors of the original SASC-R. (See
frequently specific behaviors occur with each friend. Table I.) Factor I was comprised of the eight items
Each behavior is rated on a 5-point scale for each from the FNE scale, with factor loadings ranging from
.46 to .84. Factor II was comprised of the six items
friend, with 1 = never, and 5 = very often. Four be-
from the SAD-New scale, with loadings ranging from
haviors pertain to intimacy (e.g., telling a secret),
.36 to .83; and Factor III was comprised of the four
and six to companionship (e.g., getting together on
items from the SAD-General scale, with loadings
the weekends). Responses are averaged for all
ranging from .49 to .77. In general, the factor struc-
friends named, yielding scores between 1 and 5. ture was clean, with minimal cross-loading of items.
Vernberg et al. (1992) reported high internal con- Only Item 3 (FNE subscale) had a cross-loading
sistency and moderate stability for the intimacy and greater than .40 on a second factor; a similar pattern
companionship scales (e.g., Cronbach's alpha was was observed for this item on the elementary school
.87 for companionship and .90 for intimacy); also, version of the SASC-R (La Greca & Stone, 1993).
test-retest correlations of .45 for companionship and Thus, factor analysis of the SAS-A yielded a result
.59 for intimacy were obtained over a 6-month time identical to that obtained with younger children.
period. In addition, Confirmatory Factor Analysis (Ben-
Self-Perception Profile for Adolescents (SPPA; tier, 1995) was used to evaluate adolescents' re-
Harter, 1988) This 45-item questionnaire assesses sponses on the SAS-A. Based on the
adolescents' perceptions of competence in eight ar- conceptually-derived aspects of social anxiety, a
eas: social acceptance, romantic appeal, behavioral three-factor model was evaluated (i.e., FNE, SAD-
conduct, close friendship, scholastic competence, ath- New, SAD-General). The specific SAS-A items were
letic competence, physical appearance, and job com- assigned to factors according to the conceptual
petence. Each subscale c o n t a i n s six items; groupings listed in Table I. This analysis revealed a
adolescents first determine which of two statements satisfactory fit for the three-factor model of social
is most like them (e.g., "Some teenagers find it hard anxiety, with a Goodness-of-Fit Index equal to .91,
to make friends, BUT for other teenagers it's pretty an average standardized residual of .048, and x2
easy"), and then rate how true that statement is for (132) = 341.4 (p < .01). Relative to a one-factor
them. Each item receives a score of 1 to 4, and re- model of social anxiety (Goodness-of-Fit Index =
sponses are averaged within each subscale. Harter .77; average standardized residual = .062), the three
(1988) has reported good internal consistency for factor model represented a significant improvement
each of the SPPA subscales (Cronbach's alpha (test of difference between the two models; x2 (1) =
346.1, p < .0001).
ranged from .74 to .93). In the present study, the So-
Following these analyses, scores for FNE, SAD-
cial Acceptance, Romantic Appeal, Close Friendship,
New and SAD-General were computed by summing
Behavioral Conduct, and Scholastic Competence the 5-point ratings for the items comprising each
were included. The first three of these were expected subscale. Internal consistencies (Cronbach's alpha)
to be negatively related to social anxiety, as these were .91 (FNE), .83 (SAD-New), and .76 (SAD-
subscales reflect adolescents' perceptions of social General).4 In addition, interscale correlations re-
competencies. The last two were expected to be mini-
mally related to social anxiety, as they reflect per- 4These internal consistencies were higher than those obtained us-
ceptions of competence in nonsocial areas of ing the SASC-R with children (.86 for FNE, .78 for SAD-New,
functioning. and .69 for SAD-General) (La Greca & Stone, 1993).
88 La Greca and Lopez

Table I. Factor Structure and Factor Loadings for Items of the Social Anxiety Scale for Adolescents (SAS-A)"
Factors
I II III
Fear of Negative Evaluation (FNE)
I worry about what others say about me (12) .84 .24 .17
I worry that others don't like me (14) .83 .25 .23
I'm afraid that others will not like me (9) .80 .24 .19
I worry about what others think of me (8) .77 .32 .11
I feel that others make fun of me (17) .58 .22 .35
I worry about being teased (3) .52 .42 .23
I feel that peers talk about me behind my back (6) .49 .19 .26
If I get into an argument, I worry that the other person will not like me (18) .46 .18 .36
Social Avoidance and Distress— New (SAD-New)
I get nervous when I meet new people (13) .23 ,83 .15
I feel shy around people I don't know (4) .19 .75 .15
I get nervous when I talk to peers I don't know very well (10) .37 .57 .25
I feel nervous when I'm around certain people (20) .29 .55 .28
I only talk to people I known really well (5) .22 .40 .29
I worry about doing something new in front of others (1) .25 .36 .21
Social Avoidance and Distress— General (SAD-O)
It's hard for me to ask others to do things with me (22) .22 .19 .77
I'm afraid to invite others to do things with me because they might say no (19) .26 .09 .70
I am quiet when I'm with a group of people (15) .13 .30 .51
I feel shy even with peers I know very well (21) .12 .19 .49
"Item numbers are in parentheses; SAS-A.

Table II. Means (Standard Deviations in Parentheses) for the Social Anxiety Scale for Adolescents (SAS-A)a
Boys Girls Total F P
n 101 149 250
SAS-A
Fear of Negative Evaluation 15.75 (5.8) 17.52 (6.8) 16.81 (6.4) 4.61 .033
Social Avoidance and Distress— New 14.65 (4.2) 15.86 (4.9) 15.37 (4.7) 4.05 .045
Social Avoidance and Distress— General 6.65 (2.4) 7.08 (3.0) 6.91 (2.8) 1.36 -
Total 37.07 (10.7) 40.46 (12.7) 39.09 (12.0) 4.87 .028
Median 37.0 39.0 38.0
a
SAS-A = Adolescent Version of Social Anxiety Scale for Children— Revised.

vealed that the SAS-A subscales were significantly and distress in new situations (see Table II), although
interrelated, but distinct. These correlations were the magnitude of these differences was relatively
.52 (FNE and SAD-General), .55 (SAD-General small. No grade effects or interactions between grade
and SAD-New), and .67 (between FNE and SAD- and gender were obtained.
New) (all p's < .001). Gender differences were also examined for the
friendship and support variables; boys and girls were
compared using an analysis of variance (ANOVA)
Gender Differences procedures (see Table III). To protect against Type I
error, a Bonferroni correction was applied for each
Mean scores for the three subscales and the to- set of measures. For example, for the analysis of the
tal SAS-A are presented in Table II. A Gender x social support measure (the SPPA), the alpha level
Grade analysis of variance revealed that girls re- was set at .01 (.05/5 subscales = .01). As can be seen
ported more total social anxiety than did boys (p < from Table III, in comparison to boys, girls reported
.013). Follow-up analyses for each subscale indicated more social support from their best friends and from
that adolescent girls reported greater fear of negative their teachers, more intimacy in their close friend-
evaluation from peers, and more social avoidance ships, and perceptions of greater competency in their
Adolescents' Social Anxiety 89

Table III. Means (Standard Deviations in Parentheses) for the Measures of Social Functioninga
Boys Girls Total F P
n 101 149 250
Social Support (SSSCA)
_
Classmates 3.36 (0.5) 3.39 (0.5) 3.38 (0.5) 0.07
Close friends 3.49 (0.6) 3.77 (0.5) 3.66 (0.5) 17.62 .0001
_
Parents 3.35 (0.5) 3.42 (0.6) 3.39 (0.6) 0.83
Teachers 2.81 (0.6) 3.19 (0.6) 3.04 (0.6) 23.99 .0001
Self-Perceptions (SPPA) _
Social acceptance 3.28 (0.5) 3.35 (0.6) 3.32 (0.6) 1.18
Romantic appeal 2.91 (0.6) 2.96 (0.7) 3.66 (0.5) 0.37 -
Close friends 3.33 (0.6) 3.57 (0.6) 3.47 (0.6) 10.85 .001
Scholastic competence 3.02 (0.5) 3.20 (0.6) 3.13 (0.6) 6.09 -
_
Behavioral conduct 2.95 (0.6) 3.12 (0.5) 3.05 (0.6) 5.97
Friendships
_
Number of friends 4.82 (2.3) 4.46 (2.0) 4.60 (2.1) 1.84
Intimacy 3.57 (0.8) 4.16 (0.7) 3.92 (0.8) 37.10 .0001
Companionship 3.43 (0.7) 3.52 (0.7) 3.48 (0.7) 0.98 -
a
SSSCA = Social Support Scale for Children and Adolescents; SPPA = Self-Perception Profile for
Adolescents.

close friendships. Gender differences were not ob- Correlations between the SAS-A and the other
served for support from parents and classmates, for measures were computed separately for girls and
the number of best friends, or for other areas of self- boys, and are presented in Table IV (boys' data are
perceptions. in parentheses). A conservative alpha level (a = .01)
was used to evaluate significance. The median cor-
relation for each group of variables is also indicated
Social Anxiety: Relations to Peer Acceptance, in Table IV
Friendship, and Support Variables
First, in terms of general peer relations, it was
expected that social anxiety would be most closely
Pearson correlations were computed to examine
linked to perceptions of social acceptance and sup-
linkages between adolescents' social anxiety and their
port from the larger peer group (i.e., classmates).
social functioning. However, prior to computing
The pattern of findings is consistent with this notion.
these correlations, the self-perception, social support,
High socially anxious adolescents reported less sup-
and friendship variables were organized into three
port from classmates and lower perceptions of their
conceptual groupings to reflect (a) general peer re-
social acceptance and romantic appeal, than low so-
lations and social acceptance, (b) close friendships
cially anxious adolescents. (See Table IV) Median
and friendship quality, and (c) non-peer-related per-
correlations were the highest for this group of vari-
ceptions of support and competency. These concep-
ables. Although the pattern of results was the same
tual groupings are listed in Table IV, and were
for girls and boys, the correlations were consistently
supported by a factor analysis of these measures.5
higher for girls.6 For example, for girls, higher over-
5
all social anxiety (SAS-A total) was related to lower
A principal-components factor analysis with a varimax rotation
perceived support from classmates (r = -.53), and to
was conducted for the subscales of the SPPA, SSSCA, and the
friendship variables from the AI. The first factor (eigenvalue = lower perceptions of social acceptance (r = -.52) and
4.31; 39% of variance) reflected friendship quality, and consisted romantic appeal (r = -.51). For boys, these correla-
of the following measures (factor loading in parentheses): SPPA- tions were -.25, -.39, and -.30, respectively.
Friendships (.66), SSSCA-Close Friends (.78), AI-Intimacy (.83), Second, adolescents' close friendships appeared
and AI-Companionship (.76). The second factor (eigenvalue = to be more strongly associated with girls' levels of
1.65; 15% of variance) included all the non-peer related meas-
ures: SSSCA-Parents (.61), SSSCA-Teachers (.73), SPPA-Scho- social anxiety than boys'. Specifically, adolescent girls
lastic Competence (.71), and SPPA-Behavioral Conduct (.75). who reported higher levels of social anxiety also re-
The last factor (eigenvalue = 1.01; 9.2% of variance) reflected
general peer relations/social acceptance, and consisted of the fol- 6
For the most part, a difference of .22 or more between the cor-
lowing measures: SPPA-Social Acceptance (.63), SPPA-Roman- relations for girls and boys is significant (based on Fisher's exact
tic Appeal (.77), and SSSCA-Classmates (.81). No significant z test, with p < .05).
 90 La Greca and Lopez

Table IV. Correlations Between SAS-A and Measures of Social Functioning: Girls (Correlations for Boys in
Parentheses)a
SAS-A subscales5
Total FNE SAD-General SAD-New
A. General peer relations/acceptance
SPPA-Social Acceptance -.52C (-.39c) -.42C (-.31C) -.57C (-.33C) -.42C (-.35C)
SPPA-Romantic Appeal -.51r (-.30c) -.48C (-.18) -.39C (-.31C) -.42C (-.33C)
Social Support-Classmates -.53c (-.25b) -.48c (-.24b) -.55c (-.15) -.39c (-.21)
Median correlation -.52 (-.32) -.45 (-.24) -.48 (-.24) -.41 (-.28)
B. Close friendships
Number of best friends -.28c (-.16) -.20* (-.10) -.30c (-.21) -.28C (-.14)
Social Support-Close Friends -.48c (-.12) -.39c (-.05) -.48C (-.28b) -.40c (-.08)
SPPA-Friendships -.41c (-.17) -.29c (-.07) -.52c (-.29c) -34C (-.16)
AI-Intimacy -.33c (-.18) -.24b (-.11) -.36c (-.19) -.25C (-.19)
AI-Companionship -.25c (-.12) -.19b (-.12) -.31C (-.07) -.17 (-.12)
Median correlation -.37 (-.15) -.29 (-.09) -.44 (-.13) -.29 (-.14)
C. Non-peer related
Social Support-Parents -.14 (-.10) -.10 (-.12) -.10 (-.11) -.19 (-.04)
Social Supprt-Teachers -.12 (-.05) -.13 (.02) -.09 (-.11) -.10 (-.10)
SPPA-Scholastic Competence -.21b (-.18) -.14 (-.17) -.26C (-.04) -.18 (-.19)
SPPA-Behavioral Conduct -.17 (-.13) -.08 (-.10) -.24b (.01) -.17 (-.18)
Median correlation -.16 (-.12) -.11 (-.08) -.18 (-.05) -.75 (-.12)
"SAS-A = Adolescent Version of Social Anxiety Scale for Children— Revised; FNE = Fear of Negative Evalu-
ation; SAD-General = Social Avoidance and Distress— General; SAD-New = Social Avoidance and Dis-
tress—New; SPPA = Self-Perception Profile for Adolescents; AI = Adolescent Interview.
bp < .01.
cp < .001 (one-tailed).

ported having fewer best friends, feeling less compe- Finally, to further clarify the linkages between
tent in their friendships, and perceiving their friend- social anxiety and social functioning, hierarchical re-
ships as less supportive, less intimate, and lower in gressions were used to examine the relative strength
companionship (see Table IV). Of the SAS-A of adolescents' friendships and peer acceptance in
subscales, SAD-General was the one most strongly "predicting" social anxiety scores. For each SAS-A
associated with the friendship variables. Although subscale, two regressions were conducted. In the
the overall pattern of results was similar for boys and first, a set of peer acceptance variables7 was entered
girls, social anxiety was generally not significantly re- on the first step, followed by a set of close friendships
lated to boys' friendships. The one exception was that variables8 on the second step; the order of the vari-
boys who reported high levels of SAD-General re- able sets was reversed in the second regression analy-
sis, to evaluate the unique and additive contributions
ported less support from their close friends and per-
of the social/friendship variables to the prediction of
ceived themselves to be less competent in their
social anxiety. Table V summarizes these analyses for
friendships. Overall, the median correlations between
girls and boys, separately.
social anxiety and the friendship variables were lower For girls, both close friendships and peer accep-
than those obtained for the general peer acceptance tance were significant predictors of social anxiety,
variables (described above). and together accounted for 29% to 43% of the vari-
Third, adolescents' reports of social anxiety were ance in SAS-A scores. In addition, girls' peer accep-
not related to their perceptions of support from par-
ents and teachers, and were minimally related to self-
7The set of Peer Acceptance variables included: SPPA-Social Ac-
perceptions in nonsocial areas. For girls only, high ceptance, SPPA-Romantic Appeal, and SSSCA-Classmates.
levels of SAD-General were associated with lower 8he set of close friendship variables included three variables with
perceptions of scholastic competency and behavioral high factor loadings for the "Friendship" construct: SPPA-
conduct. The median correlations for this group of Friendships, SSSCA-Close Friends, and Al-Intimacy. Al-Com-
panionship was not considered in this set as it correlated highly
variables were considerably below those for the more
with AI-Intimacy (rs > .60), and could have led to problems
peer-oriented variables, especially for girls. wiht multicolinearity in the regression analyses.
Adolescents' Social Anxiety 91

Table V. Summary of Hierarchical Regressions Predicting Social Anxiety from Social

Functioning"
Variable set R R2 R2 change F (equation)
Girls (n = 149)
Fear of Negative Evaluation
Step 1 Close friendshipsb .39 .16 8.78f
Step 2 Peer acceptancec .60 .37 .21f
13.47f

Step l Peer acceptance .58 .34 24.68f

Step 2 Close friendships .60 .37 .02 13.47f
SAD-General
Step 1 Close friendships .54 .29 19.77f
f
Step 2 Peer acceptance .66 .43 .14 17.62f

Step 1 Peer acceptance .63 .39 30.74f

Step 2 Close friendships .66 .43 .04d 17.62f
SAD-New
Step 1 Close friendships .41 .17 9.80f
f
Step 2 Peer acceptance .54 .29 .12 9.53f

Step 1 Peer acceptance .51 .26 l6.45 f

Step 2 Close friendships .54 .29 .03 9.53f
Boys (n = 101)
Fear of Negative Evaluation
Step 1 Close friendships .16 .03 .86
Step 2 Peer acceptance .33 .11 .08d 1.90

Step 1 Peer acceptance .31 .10 3.34d

Step 2 Close friendships .33 .11 .01 1.90
SAD-General
Step 1 Close friendships .31 .10 3.44d
Step 2 Peer acceptance .44 .19 .09d 3.59e

Step 1 Peer acceptance .38 .14 5.21e

Step 2 Close friendships .44 .19 .05 3.59e
SAD-New
Step 1 Close friendships .25 .06 2.16
Step 2 Peer acceptance .43 .18 .12e 3.42e

Step 1 Peer acceptance .42 .17 6.55f

Step 2 Close friendships .43 .18 .01 3.42e
a
SAD-General = Social Avoidance and Distress— General; SAD-New = Social Avoidance
and Distress— New.
b
Variables = Adolescent Interview-Intimacy; Social Support-Close Friends; Perceptions-
Friendships.
c
Variables = Social Support-Classmates; Perceptions-Social Acceptance; Perceptions-Ro-
matic Appeal.
d
p < .05.
e
p < .01.
f
p < .001.

tance consistently contributed to the prediction of so- tently associated with social anxiety, predicting 10%
cial anxiety, even after accounting for the contribu- to 17% of the variance in SAS-A scores. In contrast,
tions of close friendships (see Table V). Close close friendships were only predictive of SAD-Gen-
friendships also contributed to the prediction of eral (10% of the variance), and did not add to the
SAD-General, beyond the variance explained by peer prediction of SAD-General when peer acceptance
acceptance. For boys, peer acceptance was consis- was first considered. Across the analyses, betas for
92 La Greca and Lopez

specific variables were negative (i.e., lower peer ac- children with clinically significant anxiety disorders
ceptance and poorer quality friendships were linked are liked less by their peers and tend to be neglected
with greater social anxiety). by their classmates (Strauss, Lease, Kazdin, Dulcan,
& Last, 1989). Further exploration of the potential
role of social anxiety as a mediator between poor
DISCUSSION peer relations and the development of clinically sig-
nificant anxiety disorders in children and adolescents
would be of substantial interest (Ginsburg, La Greca,
The findings of the present investigation extend
& Silverman, 1998; Silverman & Ginsburg, 1995).
previous research on adolescents' interpersonal func-
The availability of an instrument such as the SAS-A
tioning in several important ways. Specifically, the re-
sults provide initial support for the utility of the may facilitate such investigations.
SAS-A for assessing social anxiety among adoles- Perhaps of most importance in the present study
cents. In addition, the findings help to document im- were the findings linking adolescents' social anxiety
portant linkages between adolescents' social anxiety with their general peer acceptance, as well as the ob-
and their friendships and peer relations. The key tained associations between social anxiety and girls'
findings will be discussed in turn. close friendships. First, it was observed that adoles-
A primary goal of this study was to examine the cents who reported higher levels of social anxiety felt
factor structure and psychometric properties of the less accepted and supported by their classmates and
SAS-A. Both exploratory and confirmatory factor less romantically attractive to others. These associa-
analyses revealed essentially the same underlying fac- tions were apparent for both boys and girls, although
tor structure as that obtained with elementary school they were stronger for girls. These findings extend
children (La Greca & Stone, 1993); that is, three fac- previous work with children that revealed linkages
tors emerged: FNE, SAD-New, and SAD-General. between social anxiety and peer acceptance (e.g., La
These findings provide some support for the devel- Greca & Stone, 1993). If socially anxious adolescents
opmental continuity of these components of social perceive their general social acceptance or romantic
anxiety. In addition, satisfactory levels of internal appeal to be low, as the present results indicate, this
consistency were obtained for the SAS-A subscales, may lead them to miss out on important socialization
and moderate interrelationships among the subscales experiences and, over time, may contribute to impair-
were observed. Together, these data provide initial ments in social functioning. Consistent with this per-
psychometric support for the SAS-A. spective, clinical reports suggest that about 70% of
As has been the case with children (Crick & adults with social phobia (an extreme form of social
Ladd, 1993; La Greca & Stone, 1993), gender dif- anxiety) report impairments in their social relation-
ferences were apparent in adolescents' reports of so- ships (e.g., Turner, Beidel, Dancu, & Keys, 1986),
cial anxiety (i.e., FNE, SAD-New). Kashani and and epidemiological studies have shown that adults
colleagues (1989) observed that adolescent girls ap- with social phobia are less likely to marry than nor-
peared to have a greater concerns than boys about mal controls (e.g., Schneier et al., 1992).
feelings of social inadequacy, consistent with our Given the cross-sectional nature of the present
findings of greater social anxiety for adolescent girls. investigation, it was not possible to determine
The availability of an instrument to assess social whether feelings of social anxiety contributed to poor
anxiety that has been validated across a wide age peer relations among adolescents, or whether poor
range (i.e., 7 through 18 years of age) may facilitate peer relations led to feelings of social anxiety. How-
developmental and longitudinal investigations of ever, it is conceivable that being neglected or ex-
children's social and affective functioning. The SAS- cluded from peer interactions represents a significant
A may also prove useful for examining the onset and stressor for adolescents (Frankel, 1990), that leads to
course of social phobia or other anxiety-related dis- feelings of social apprehension, worry, or distress
orders that have substantial social components. To (see Leary, 1990); these subjective feelings, in turn,
date, research into the factors contributing to young- may lead to behavioral avoidance of peers, and con-
sters' anxiety disorders has focused largely on family tribute to missed opportunities for normal socializa-
and biological variables (e.g., Black, 1995; Bruch, tion experiences. In this manner, social anxiety may
1989; Weissman et al., 1984). Peer influences have help to explain the linkages between poor peer rela-
received much less attention, despite findings that tions and the clinical impairment associated with
Adolescents' Social Anxiety 93

youngsters' anxiety disorders (Strauss et al., 1989). ents or teachers (Loeber, Green, & Lahey, 1990).
Further research of a longitudinal nature will be Adolescents are also likely to be the best source of
helpful in elucidating such etiological pathways. information regarding peer relations and friendships.
Also important and unique to the present inves- Parents, for example, may be unaware of their ado-
tigation were the findings linking social anxiety with lescents' friendships. Although adolescents were the
adolescent girls' close friendships. Specifically, socially sole informants, the pattern of obtained findings sug-
anxious girls (especially those with high levels of gen- gests that shared method variance alone could not
eralized social avoidance and distress) reported hav- account for these results. For example, social anxiety
ing fewer best friends and these friendships were was related to adolescents' peer relations and friend-
perceived to be lower in intimacy, companionship, ships but not to social support from adults. Also, dif-
and emotional support than were the friendships of ferent patterns of results were obtained for boys and
less socially anxious girls. In contrast, for boys, social girls. Nevertheless, in future studies, it would be use-
anxiety was not typically related to friendship quali- ful to consider obtaining input from significant others
ties, although generalized social avoidance and dis- (e.g., parents) in the adolescents' lives.
tress was associated with less perceived support and In conclusion, the present results highlight the
competency in boys' close friendships. interconnections between adolescents' social anxiety
These results are troubling, as they suggest that and their interpersonal functioning, and identify gen-
social anxiety may interfere substantially with girls' der differences in adolescents' reports of social anxi-
close interpersonal relationships. Intimacy and emo- ety and close friendships. Kashani et al. (1989)
tional support from close friends are salient features observed that social/interpersonal anxiety appears to
of adolescent girls' friendships, as our findings and be a fact of life for adolescent boys and girls at about
others' have indicated (e.g., Berndt, 1982; Harter, the time when satisfaction with family decreases and
1985); thus factors that interfere with the develop- peers become the focus of attention. Future efforts
ment of close, intimate friendships—such as feelings to study linkages between social anxiety and adoles-
of social anxiety—may have a greater negative im- cents interpersonal functioning seem crucial. In par-
pact on girls than on boys. Alternatively, it may be ticular, efforts to understand why girls appear to be
the case that girls who encounter difficulties in their more vulnerable to feelings of social anxiety than
close friendships are especially vulnerable to feelings boys, as well as efforts to evaluate the specific path-
of anxiety. Our findings are consistent with these no- ways by which social anxiety may lead to social im-
tions, as well as with previous work with early ado- pairment, would be especially illuminating. In
addition, the role of social anxiety in the onset and
lescents (Vernberg et al., 1992), that suggested a
course of anxiety disorder among adolescents ap-
bidirectional relationship between problems in close
pears to be a fruitful avenue for further study.
friendships and feelings of social anxiety for girls.
Despite the important contributions of this
study, limitations should be noted. First, the crosssec-
tional nature of the study precluded an evaluation of ACKNOWLEDGMENTS
reciprocal influences between adolescents' social
Preparation of this manuscript was supported by
anxiety and their interpersonal functioning. Although
a grant from the National Institute of Mental Health
this was a necessary beginning point to explore the
(RO1-MH48028) to the first author. Portions of
construct of social anxiety among adolescents, future
these data were presented at the meeting of the As-
studies might consider using short-term longitudinal
sociation for the Advancement of Behavior Therapy,
methodologies, particularly during important social
November 1995. The authors would like to thank the
transition periods (e.g., school transitions, reloca-
following individuals for their invaluable assistance
tion), to further elucidate the causal pathways be- and input in this project: Nicole Vincent, Mitch Prin-
tween social anxiety and social/emotional stein, and Michael Fetter.
dysfunction.
Another limitation of this study is that the meas-
ures were based exclusively on adolescent reports.
REFERENCES
Adolescents' perspectives were elicited in this study
because adolescents are considered to be the best in- American Psychiatric Association. (1994). Diagnostic and statistical
formants for internalizing difficulties, relative to par- manual of mental disorders. (4th ed.). Washington, DC:
94 La Greca and Lopez

Bentler, P. M. (1995). EQS: A structural equation program (Version La Greca, A. M., & Stone, W. L. (1993). Social Anxiety Scale for
5.1). BMDP Statistical Software. Children—Revised: Factor structure and concurrent validity.
Berndt, T. J. (1982). The features and effects of friendship in early Journal of Clinical Child Psychology, 22, 17-27.
adolescence. Child Development, 53, 1447-1460. Leary, M. R. (1983). Understanding social anxiety: Social, person-
Berndt, T. J., & Perry, T. B. (1986). Children's perceptions of ality, and clinical perspectives. Beverly Hills, CA: Sage.
friendships as supportive relationships. Developmental Psy- Leary, M. R. (1990). Responses to social exclusion: Social anxiety,
chology, 23, 583-592. jealousy, loneliness, depression, and low self-esteem. Journal
Bernstein, G. A., Garfinkel, B. D., & Hoberman, H. M. (1989). of Social and Clinical Psychology, 9, 221-229.
Self-reported anxiety in adolescents. American Journal of Psy- Loeber, R., Green, S. M., & Lahey, B. B. (1990). Mental health
chiatry, 146, 384-386. professionals' perceptions of the utility of children, parents,
Black, B. (1995). Separation anxiety disorder and panic disorder. and teachers as informants on childhood psychopathology.
In J. S. March (Ed.), Anxiety disorders in children and adoles- Journal of Clinical Child Psychology, 19, 136-143.
cents (pp. 212-234). New York: Guilford Press. Nolen-Hoeksema, S., & Larson. J. (1992). The worries of adolescent
Buhrmester, D., & Furman, W. (1987). The development of com- males and females. Unpublished manuscript, Stanford Univer-
panionship and intimacy. Child Development, 58, 1101-1113. sity, Stanford, CA.
Bulik, C. M., Beidel, D. C, Duchmann, E., Weltzin, T. E., & Kaye, Quay, H. C., & La Greca, A. M. (1986). Disorders of anxiety,
withdrawal, and dysphoria. In H. C. Quay & J. S. Werry
W. H. (1991). An analysis of social anxiety in anorexic, bu-
(Eds.), Psychopathological disorders of childhood (3rd ed.).
limic, social phobic and control women. Journal of Psychopa- (pp. 73-110). New York: Wiley.
thology and Behavioral Assessment, 13, 199-211. Rosen, B. C., & Aneshensel, C. S. (1976). The chameleon syn-
Bruch, M. A. (1989). Familial and developmental antecedents of drome: A social psychological dimension of the female sex
social phobia: Issues and findings. Clinical Psychology Review, role. Journal of Marriage and the Family, 38, 605-617.
9, 37-47. Schneier, F. R., Johnson, J., Hornig, C. D., Liebowitz, M. R., &
Crick, N. R., & Ladd, G. W. (1993). Children's perceptions of Weissman, M. M. (1992). Social phobia: Comorbidity in an
their peer experiences: Attributions, loneliness, social anxiety, epidemiological sample. Archives of General Psychiatry, 49,
and social avoidance. Developmental Psychology, 29, 244-254. 282-288.
Dusek, J. (1991). Adolescent development and behavior (3rd ed.). Schneier, F. R., Martin, L. Y., Liebowitz, M. R., Gorman, J. M.,
Palo Alto, CA: Science Research Associates. & Fyer, A. J. (1989). Alcohol abuse in social phobia. Journal
Frankel, K. A. (1990). Girls' perceptions of peer relationship sup- of Anxiety Disorders, 3, 15-23.
port and stress. Journal of Early Adolescence, 10, 69-88. Silverman, W. K., & Ginsburg, G. (1995). Specific phobias and
Ginsburg, G. S., La Greca, A. M., & Silverman, W. S. (1998). generalized anxiety disorder. In J. S. March (Ed.), Anxiety dis-
Social anxiety in children with anxiety disorders: Relation with orders in children and adolescents (pp. 151-180). New York:
social and emotional functioning. Journal of Abnormal Child Guilford Press.
Psychology. Silverman, W. K., La Greca, A. M., & Wasserstein, S. (1995).
Harter, S. (1985). Manual for the Social Support Scale for Children What do children worry about?: Worries and their relation-
and Adolescents. Denver: University of Denver. ship to anxiety. Child Development, 66, 671-686.
Harter, S. (1988). Manual for the Self Perception Profile for Ado- Strauss, C. C., Frame, C. L, & Forehand, R. (1987). Psychosocial
lescents. Denver: University of Denver. impairment associated with anxiety in children. Journal of
Hazen, A. L., & Stein, M. B. (1995). Clinical phenomenology and Clinical Child Psychology, 16, 235-239.
comorbidity. In M. B. Stein (Ed.), Social phobia: Clinical and Strauss, C. C., Lease, C. A., Kazdin, A. E., Dulcan, M. K., & Last,
research perspectives (pp. 3-42). Washington, DC: American C. G. (1989). Multimethod assessment of the social compe-
Psychiatric Press. tence of children with anxiety disorders. Journal of Clinical
Hymel, S., & Frank, S. (1985). Children's peer relations: Assessing Child Psychology, 18, 184-189.
self-perceptions. In B. Schneider, K. Rubin, & J. Ledingham Turner, S. M., Beidel, D. C., Dancu, C. V., & Keys, D. J. (1986).
Psychopathology of social phobia and comparison to avoidant
(Eds.), Children's peer relations: Issues in assessment and in-
personality. Journal of Abnormal Psychology, 95, 389-394.
tervention (pp. 75-92). New York: Springer. Vernberg, E. M. (1990). Psychological adjustment and experiences
Ingersoll, G. M. (1989). Adolescents (2nd ed.). Englewood Cliffs, with peers during early adolescence: Reciprocal, incidental,
NJ: Prentice Hall. or unidirectional relationships? Journal of Abnormal Child
Kashani, J. H., Orvaschel, H., Rosenberg, T. K., & Reid, J. C. Psychology, 18, 187-198.
(1989). Psychopathology in a community sample of children Vernberg, E. M., Abwender, D. A., Ewell, K. K., & Beery, S. H.
and adolescents: A developmental perspective. Journal of the (1992). Social anxiety and peer relationships in early adoles-
American Academy of Child and Adolescent Psychiatry, 28, cence: A prospective analysis. Journal of Clinical Child Psy-
701-706. chology, 21, 189-196.
Kuhlen, R. G., & Houlihan, N. B. (1965). Adolescent heterosexual Watson, D., & Friend, R. (1969). Measurement of social-evalu-
interest in 1942 and 1965. Child Development, 36, 1049-1052. ative anxiety. Journal of Consulting and Clinical Psychology,
La Greca, A. M., Dandes, S. K., Wick, P., Shaw, K., & Stone, W. 33, 448-457.
L. (1988). Development of the Social Anxiety Scale for Chil- Weissman, M. M., Leckman, J. F., Merikangas, K. R., et al. (1984).
dren: Reliability and concurrent validity. Journal of Clinical Depression and anxiety disorders in parents and children. Ar-
Child Psychology, 17, 84-91. chives of General Psychiatry, 41, 845-852.