Journal of Child Psychology and Psychiatry 45:8 (2004), pp 13631372

doi: 10.1111/j.1469-7610.2004.00327.x

Behavioral and emotional adjustment, family

functioning, academic performance, and social
relationships in children with selective mutism
Charles E. Cunningham,1,2 Angela McHolm,1 Michael H. Boyle,2 and Sejal Patel2
1

McMaster Childrens Hospital, Hamilton, Ontario, Canada; 2McMaster University, Hamilton, Ontario, Canada

This study addressed four questions which parents of children with selective mutism (SM) frequently
ask: (1) Is SM associated with anxiety or oppositional behavior? (2) Is SM associated with parenting and
family dysfunction? (3) Will my child fail at school? and (4) Will my child make friends or be teased and
bullied? In comparison to a sample of 52 community controls, 52 children with SM were more anxious,
obsessive, and prone to somatic complaints. In contrast, children with SM were less oppositional and
evidenced fewer attentional difficulties at school. We found no group differences in family structure,
economic resources, family functioning, maternal mood difficulties, recreational activities, or social
networks. While parents reported no differences in parenting strategies, children with SM were described as less cooperative in disciplinary situations. The academic (e.g., reading and math) and
classroom cooperative skills of children with SM did not differ from controls. Parents and teachers
reported that children with SM had significant deficits in social skills. Though teachers and parents
rated children with SM as less socially assertive, neither teachers nor parents reported that children
with SM were victimized more frequently by peers. Keywords: Selective mutism, anxiety, academic
performance, peer victimization.

Selective Mutism (SM), or Elective Mutism according

to the ICD-10 (World Health Organization, 1992), is a
disorder in which children fail to talk in specific situations while speaking in others (Diagnostic and Statistical Manual 4th edn, 1994). While many children
with SM do not speak to teachers or classmates, they
converse normally at home with parents and siblings. Early epidemiological studies suggested that
SM was rare, with prevalence figures ranging from
.038% to .069% of elementary school samples
(Brown & Lloyd, 1975; Fundudis, Kolvin, & Garside,
1979). Recent studies suggest SM is more common
(Kopp & Gillberg, 1997; Bergman, Piacentini, &
McCracken, 2002), with some reporting figures as
high as 2% of 2nd graders (Kumpulainen, Rasanen,
Raaska, & Somppi, 1998). While some studies report
a decline in the prevalence of SM in the first several
years at school (Fundudis et al., 1979), others suggest that SM is a more persistent disorder (Bergman
et al., 2002).
Although descriptive studies have contributed
much to our understanding of SM (Black & Uhde,
1995; Dummit et al., 1997; Ford, Sladeczek, Carlson, & Kratochwill, 1998; Steinhausen & Juzi, 1996),
several issues emphasize the need for additional research. First, the absence of control groups in these
studies makes it difficult to determine whether the
prevalence of problems in children with SM differs
from children without SM. The present study contributes to a small number of existing studies that
have included controls (Kristensen, 2001; Bergman
et al., 2002) by comparing 52 children with SM to a
community sample of 52 children without SM.

Second, while SM often emerges in the preschool

years (Steinhausen & Juzi, 1996), the largest controlled study available enrolled children averaging
more than 9 years of age (Kristensen, 2001). Because children with SM may begin speaking during
the first several years of school (Fundudis et al.,
1979), older samples including a disproportionate
number of children with persistent mutism may
overestimate problems among children with SM. The
present study, therefore, enrolled children with SM
several years earlier than most previous reports.
Third, with recent studies conducted in the United
States (Bergman et al., 2002; Ford et al., 1998),
Finland (Kumpulainen et al., 1998), Sweden (Kopp &
Gillberg, 1997), Norway (Kristensen, 2001), Switzerland, and Germany (Steinhausen & Juzi, 1996),
national factors such as immigration rates (Bradley
& Sloman, 1975; Brown & Lloyd, 1975), access to
early childhood education, the age at which formal
schooling begins, and the availability of effective
treatments might influence the onset, correlates,
and course of SM. Distinguishing common characteristics from unique local factors requires systematic replication in different countries. The
present study, therefore, addressed four questions in
a large Canadian sample.

Question 1: Is SM associated with anxiety

or oppositional behavior?
Both descriptive studies (Black & Uhde, 1995;
Dummit et al., 1997; Ford et al., 1998; Steinhausen
& Juzi, 1996) and comparisons including controls

Association for Child Psychology and Psychiatry, 2004.

Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA

1364

Charles E. Cunningham et al.

(Bergman et al., 2002; Kristensen, 2001) report a

high prevalence of anxiety-related disorders among
children with SM. Indeed, some authors suggest that
SM should be reclassified as a social phobia (Black &
Uhde, 1995; Dummit et al., 1997) or anxiety disorder (Anstendig, 1999). The DSM-IV (American Psychiatric Association, 1994) and ICD-10 (World
Health Organization, 1992) clinical descriptions,
descriptive studies (Kolvin & Fundudis, 1981;
Kumpulainen et al., 1998; Krohn, Weckstein, &
Wright, 1992; Steinhausen & Juzi, 1996) and controlled comparisons suggest that oppositional behavior is also common among children with SM.
Kristensen (2001), for example, found parentreported externalizing problems twice as frequent
among children with SM as controls. Kristensen
(2000) cautioned, however, that the SM childs
avoidant behavior may be misinterpreted as controlling or manipulative rather than symptomatic of
shyness or anxiety. To explore this issue, we compared internalizing (anxiety, depression, and obsessive behavior) and externalizing (ADHD, oppositional
disorder, and conduct disorder) symptoms in children with SM and community controls. While parents typically observe children with SM speaking
comfortably at home, teachers observe children with
SM in situations where a failure to respond to speech
demands might be misinterpreted as oppositional
behavior. Because the childs behavioral and emotional responses may be different at home and
school, parents and teachers completed independent
ratings. We predicted that parents and teachers
would rate children with SM as more anxious and
oppositional than controls.

Question 2: Is SM linked to parenting

and family dysfunction?
Although case reports have suggested a link between
family dysfunction and SM (Brown, Fuller, & Gericke, 1975; Goll, 1979; Pustrom & Speers, 1964;
Rosenberg & Linblad, 1978; Wright, 1968), this has
not been confirmed in larger studies (Brown & Lloyd,
1975). Other investigators have linked SM to
stressful events. Steinhausen & Juzi (1996) reported
that 31% of a sample of children with SM had
experienced a stressful life event prior to the onset of
SM. Several studies reported a higher prevalence of
SM among families who have immigrated (Bradley &
Sloman, 1975; Brown & Lloyd, 1975). Finally, recent
studies find parental internalizing disorders to be an
important correlate (Black & Uhde, 1995; Kristensen
& Torgersen, 2001) and outcome predictor (Kolvin &
Fundudis, 1981) among children with SM.
The present study compared the structure and
economic resources of families of children with SM
and controls. Next, we examined parenting strategies, parental adjustment, and general family dysfunction. Finally, we explored whether immigration,
support networks, or less frequent participation in

extracurricular activities were associated with SM.

Given previous studies, we predicted that parents of
children with SM would report greater family dysfunction, less effective social networks, and poorer
parental adjustment than those of children without
SM.

Question 3: Will my child fail at school?

Although some authors suggest that children with
SM do well academically (Cline & Baldwin, 1994),
empirical studies have yielded conflicting results.
While Kumpulainen et al. (1998) reported that 32%
of children with SM were performing below grade
level, the absence of either objective measurement or
a control group makes this finding difficult to interpret. Moreover, although teachers in studies with
control groups rate the academic performance of
children with SM significantly lower than peers
(Bergman et al., 2002), teachers may underestimate
the skills of children who do not speak in the classroom. The present study, therefore, employed both
teacher ratings and objective measures of reading
and mathematical skills. Teachers rated cooperative
participation in classroom activities and parents
reported home activities that might influence academic progress at school. We predicted that children
with SM would score lower on academic tests and
receive lower teacher ratings of academic performance than controls.

Question 4: How will SM affect my childs peer

relationships?
Children with SM are often shy and inhibited (Black
& Uhde, 1995; Kristensen & Torgersen, 2001),
characteristics linked to victimization by peers (Boivin, Hymel, & Bukowski, 1995; Olweus, 1994;
Schwartz, Dodge, & Coie, 1993). While Kumpulaninen et al. (1998) reported that 16% of children with
SM were rejected and 5% were bullied by peers,
higher levels of victimization have been reported in
community samples without SM (Olweus, 1994;
Wolke, Woods, Bloofield, & Karstadt, 2000). This
study, therefore, examined social skills and victimization by peers. We predicted that children with SM
would be rated less socially competent by parents
and teachers and be victimized by peers more
frequently than controls.

Method
Participants
A sample of 104 children, 52 with SM and 52 community controls, participated in this study. Children with
SM were recruited over a period of 9 years from a regional service providing child assessments, school
consultations, and workshops for parents and teachers.
Using interviews and parent ratings, children with SM

Selective mutism

were diagnosed according to DSM-IV criteria as: (1)

failing to speak in selected situations, (2) speaking in
other situations, and (3) mutism persisting for at least
one month. Although children with mutism judged to be
secondary to speech and language disorders or unfamiliarity with English would have been excluded, no
cases met these criteria. In the month prior to our home
visit, parents reported that 100% of children with SM
spoke in a normal (i.e., non-whisper) tone of voice to
their mothers and siblings and 98% spoke in a normal
tone to their fathers. At school, in contrast, only 11.5%
had ever spoken in a normal voice to their teachers in
class and only 8.2% had spoken in a normal voice in
front of the class. Parents reported that the duration of
SM ranged from 37 to 151 months (average
6.9 years), with age of onset ranging from 2 to 5 years
(average 3.3 years). As in previous studies (Dummit
et al., 1997; Ford et al., 1998; Kumpulainen et al.,
1998; Steinhausen & Juzi, 1996), our sample included
a greater percentage of girls with SM (59.6%). Parents of
children with SM reported SM in 15.4% of their families
and a history of anxiety, panic or related disorders in
23.5% of their families. Controls were blocked on age
and sex and randomly selected from a stratified random
sample of 2439 children (89.4% of those approached)
from 61 schools in 11 boards of education selected by
the Ministry of Education for the Tri Ministry Project
(Boyle et al., 1999). To ensure a non-problem control
group, we excluded children with parent and teacher
internalizing scores or parent and teacher externalizing
scores greater than 2 standard deviations above the
mean for age (612) and sex on the Revised Ontario
Child Health Study (OCHS-R) scales (Boyle et al.,
1993). The SM and control groups did not differ on child
gender or age, family status, principal caregiver
employment, family income, the receipt of economic
assistance, or child immigrant status (Table 1).

Procedure
During a home visit for both the control and SM groups,
parents provided written informed consent, completed a
battery of questionnaires, and were given a structured
interview (Boyle et al., 1999). Children completed brief
reading and arithmetic tests. A parallel set of questionnaires was mailed to teachers (Boyle et al., 1993;

Table 1 Demographic characteristics of children with SM and

controls
Selective
mute
Mean

Sample size
% Girls
Child age in months
% Single parent
Principal caregiver
employed
Family income level
% Economic assistance
% Child immigrated to
Canada

52
52
59.6
57.7
.84 .50
85.9 (22.5) 83.4 (22.0) .58 .56
11.5 11.5
1.00
.62
69.2 74.5
.55
.35
6.2
14.3
2.1

Mean

6.0
7.7
.35

SD

X2/t

Measure

5.8

SD

Control

.40 .18
.29 .23
.61

1365

Boyle et al., 1999). The return rate for teacher questionnaires was 96%. If subscale items were missing but
75% of the questions on that subscale were completed,
we prorated the subscale upward using the average
value of the completed items.

Dependent measures
Parent and teacher reports: internalizing and
externalizing problems. Parents completed the
separation anxiety disorder, overanxious disorder,
depression, attention deficit disorder, oppositional
defiant disorder, and conduct disorder subscales of the
Revised Ontario Child Health Study (OCHS-R) scales
(Boyle et al., 1993), a revision of the Ontario Child
Health Studys Survey Diagnostic Instrument (Boyle
et al., 1987) designed to measure DSM-III-R symptoms
in population studies. Teachers completed the overanxious disorder, depression, attention deficit disorder,
oppositional defiant disorder, and conduct disorder
subscales of the OCHS-R (Boyle et al., 1993). Items are
rated on a 3-point Likert scale (0 never or not true,
1 sometimes true, and 2 often or very true). The
number of items per subscale, internal consistency,
and testretest reliability of the OCHS-R parent and
teacher scales are summarized in Table 2.

Family structure, resources, and functioning.

Parents reported marital status, income, economic
assistance, and immigration and completed a 17-item
four-point Likert scale (Fast Track Project, 2002) describing five permissive discipline practices (a .65)
(e.g., If you ask ___ to do something and he/she does not
do it, how often do you give up trying?), four coercive
discipline practices (alpha .60) (e.g., When ____ has
done something wrong, how often do you lose your temper toward him/her?) and eight uncooperative/non
compliant child responses (a .81) (e.g., How often do
you have to threaten ____ with punishment in order to get
him/her to do something?). The 12-item Family
Assessment Device (FAD) General Functioning Scale
(Epstein, Baldwin, & Bishop, 1983) measured family
relationships (e.g., we dont get along well together),
communication (e.g., we avoid discussing our fears and
concerns) and problem solving (e.g., we are able to make
decisions about how to solve problems) on a four-point
scale (strongly disagree to strongly agree). The FAD has
adequate internal consistency (a .86) and validity
(Byles, Byrne, Boyle, & Offord, 1988). The Social Provision Scales 24-item, four-point Likert scale divided
social support into six four-item subscales (Cutrona &
Russell, 1987). Reliability coefficients in the
present study were .79 for social integration, .76 for attachment, .79 for reliable attachment, .82 for parental
guidance, .72 for recognized assets, and .68 for parent
nurturing.

Parental depression. Parents completed the 20-item

Center for Epidemiological Studies Depression Scale
(CES-D). Items are rated on a four-point Likert scale
ranging from 0 none of the time or rarely (less than
1 day) to 3 most or all of the time (57 days). The
CESD has adequate internal consistency (a .85) and
1 month testretest reliability (r .54) (Radloff, 1977).

1366

Charles E. Cunningham et al.

Table 2 Number of items, internal consistency, and testretest reliability for Revised Ontario Child Health Study (OCHS-R) scales
Questions/symptoms
scored

Internal consistency

Testretest reliability

Measure

Parent

Teacher

Parent

Teacher

Parent

Teacher

Attention Deficit Disorder

Oppositional Defiant Disorder
Conduct Disorder
Separation Anxiety Disorder
Anxiety Disorder
Depressive Disorder

14/14
9/9
12/12
9/9
10/7
17/9

14/14
9/9
9/9

6/6
14/8

.90
.85
.68
.73
.68
.72

.95
.93
.68

.75
.76

.85
.87
.71
.65
.65
.65

.93
.89
.87

.84
.88

Dependent measures: academic performance and

social competence.
Academic performance. Teachers rated reading, math,
and general performance on a scale ranging from 1 (far
below grade level) to 5 (far above grade level). At home
visits, interviewers administered the Revised Wide
Range Achievement Tests (WRAT-R) reading and
arithmetic subtests (Jastak & Wilkinson, 1984). In the
SM group, 37% would not speak to interviewers. When
children would not complete academic tasks verbally,
the interviewer left the room and children read words or
answered math questions for parents. Parents noted
whether they read to their child, listened to their child
read, or their child read for fun in the last 7 days.
Social competence. Parents completed a 38-item version of the Social Skills Rating System (SSRS) (Gresham
& Elliott, 1990) from the Tri-Ministry Project (Boyle
et al., 1999). We computed four item total scores: (1)
social assertion (e.g., inviting others over or starting
conversations), (2) social control (e.g., controlling temper or ending disagreements), (3) social responsibility
(e.g., answering the phone or asking permission), and
(4) social cooperation (e.g., putting toys away or helping
without being asked). The internal consistency of the
SSRS parent scales ranges from .87 to .90 (American
Guidance Service, 2003). Teachers completed a 30-item
SSRS questionnaire. We computed three item total
scores: (1) social cooperation (e.g., following directions
or finishing class work on time), (2) social control (e.g.,
cooperating with peers or controlling temper), and (3)
social assertion (e.g., initiating conversations with
peers or inviting others to join activities). The internal
consistency of the SSRS teacher scales ranges from .93
to .94 (American Guidance Service, 2003). Parents
noted whether children participated in one or more of
ten sports, eight clubs or recreational activities, and two
art or musical activities in the last 7 days.
Submissiveness and victimization. Parents rated six
submissiveness items (e.g., gives in to peers or is defenseless) (a .71) and five victimization items (e.g.,
picked on, teased, or bullied) on a 3-point OCHS-R
scale (a .73) (Cunningham, Boyle, Offord, Racine, &
Hundert, 1999). Teachers completed identical submissiveness (a .83) and victimization (a .87) scales.

Question 1: Is SM associated with anxiety

or oppositional behavior?
Parent- and teacher-reported OCHS-R scores are
summarized in Table 3. A MANOVA using internalizing scores (Separation Anxiety, Anxiety, Depression, and OCD) yielded a significant group effect,
F(4, 99) 4.93, p .001. Parents gave children with
SM higher anxiety and OCD scores than controls.
The Depression and Separation Anxiety scores of
children with SM did not differ from controls. While
the OCHS-Rs anxiety scale included somatic complaints, the DSM-IV considers somatic complaints a
separate disorder. When somatic complaints and
anxiety items were separated, children with SM had
higher anxiety scores and more somatic complaints
than controls. A MANOVA including parent-reported
externalizing problems (ADHD, ODD, and Conduct
Disorder) was not significant, F(3, 100) .52,
p .669.
A MANOVA across teacher-reported internalizing
scores (Anxiety, Depression, and OCD) yielded a
significant group effect, F(3, 95) 3.02, p .033.
While teachers rated children with SM more anxious,
the depression scores of children with SM did not
differ from controls (Table 3). A MANOVA using
teacher-reported externalizing scores (ADHD, ODD,
and CD) yielded significant group effect, F(3,
95) 2.79, p .045. Teachers gave children with
SM lower ADHD and ODD ratings than controls
(Table 3). Scores on the CD scale did not differ.
In the SM group, there were significant correlations between parent and teacher reports of ADHD,
r(48) .51, p < .001, and anxiety, r(48) .47,
p .001, symptoms. In the control group, parents
and teachers ratings of ADHD, r(52) .58, p < .001,
ODD, r(52) .59, p < .001, and depression symptoms, r(52) .36, p .009, were significantly
correlated.

Question 2: Is SM linked to parenting

and family dysfunction?

Results
We computed one-way (SM vs. control) MANOVAs on
groups of related dependent measures. When the
main effect for group was significant, we computed
univariate ANOVAs.

A MANOVA across the Parenting Practice Scales

subscales was significant, F(3, 96) 3.84, p .012.
While parents of children with SM did not report
more coercive or permissive parenting, children with
SM were less cooperative than controls (Table 4).

Selective mutism

1367

Table 3 Question 1: Is SM associated with externalizing or internalizing problems? Parent- and teacher-reported OCHS-R mean
total scores and standard deviations for SM and control groups
Selective Mute
Subscale
Parent report
ADHD
ODD
Conduct Disorder
Separation Anxiety
Anxiety
Anxiety (DSM-IV)
Somatic Complaints
OCD
Depression
Teacher report
ADHD
ODD
Conduct Disorder
Anxiety
OCD
Depression

Control

(Range)

Mean

SD

Mean

SD

(028)
(018)
(024)
(018)
(014)
(012)
(08)
(08)
(018)

6.5
3.3
1.0
3.2
4.0
3.5
.9
1.6
3.1

(4.6)
(2.5)
(1.3)
(3.0)
(2.8)
(2.5)
(1.1)
(1.7)
(2.6)

6.3
3.6
.9
2.4
2.2
2.0
.3
.7
2.4

(4.8)
(2.6)
(1.4)
(2.2)
(1.4)
(1.4)
(.6)
(.9)
(1.9)

.05
.48
.09
2.15
16.88
13.68
11.85
11.97
2.80

.82
.49
.77
.15
<.001
<.001
.001
.001
.10

(028)
(018)
(018)
(012)
(08)
(016)

3.3
.9
.3
3.0
.8
2.9

(3.9)
(1.6)
(.6)
(1.8)
(1.3)
(2.7)

6.4
2.4
.6
2.0
.6
2.1

(7.4)
(3.6)
(1.4)
(2.1)
(1.0)
(2.2)

6.36
7.03
1.66
6.46
.74
2.39

.01
.01
.20
.01
.39
.13

Table 4 Question 2: Is SM associated with parenting and family dysfunction? Means and standard deviations for selectively mute
and control groups
Selective mute
1

Measure

Mean

Parenting Practices Inventory1

Coercive parenting
5.5
Permissive parenting
7.1
Child noncompliance
7.2
Family dysfunction (FAD)1
19.4
Parent depression (CES-D)1
5.4
Social Provisions Scale Effectiveness of Social Networks2
Attachment
13.9
Social integration
13.5
Social guidance
13.7
Nurturance
13.4
Reliable attachment
14.2
Recognized assets
13.0
Total score
81.7
1
2

Control
SD

Mean

SD

(1.8)
(2.6)
(4.2)
(4.9)
(4.9)

6.0
6.5
5.3
19.7
7.0

(1.8)
(2.6)
(3.5)
(5.4)
(8.5)

1.97
.92
5.94
.07
1.41

.21
.29
.02
.79
.24

(2.1)
(1.8)
(2.4)
(2.2)
(1.7)
(1.8)
(9.7)

14.0
13.5
14.3
13.7
14.5
13.4
83.4

(1.9)
(1.6)
(2.0)
(2.0)
(1.9)
(1.9)
(8.7)

.11
.00
1.59
.56
.59
.95
.80

.74
.98
.21
.46
.44
.33
.37

Higher scores reflect poorer functioning.

Higher scores reflect better social networks.

Family functioning and parental depression in the

SM and control groups did not differ (Table 4). A
MANOVA across the Social Provision Scales subscales was not significant, F(6, 92) 1.06, p .393
(Table 4).

Question 3: Will my child fail at school?

Academic performance is summarized in Table 5.
The arithmetic and reading test scores of children
with SM did not differ from controls. Teacher ratings
of math, reading, and overall academic and school
performance did not differ. There were no differences
in the proportion of SM and control group parents
who reported academic activities at home such as
reading to the child, child reading to the parent,
child reading for fun, or the use of computers.

Question 4: Will my child develop friends

or be bullied?
In Table 6, the percentage of SM and control group
children playing with peers outside of school, or
enrolled in sports, clubs, or art activities did not
differ. Teacher- and parent-reported social skills are
summarized in Table 7. A MANOVA across teacherreported SSRS scores yielded a significant group
effect, F(3, 92) 477.01, p < .001. While teachers
found children with SM less socially assertive, social
cooperation and social control did not differ significantly.
A MANOVA across parent-reported SSRS subscale
scores yielded a significant group effect, F(4,
95) 23.47, p < .001. Parents rated children with
SM lower than controls on social assertion, social

1368

Charles E. Cunningham et al.

Table 5 Question 3: Will children with SM fail academically?

Academic performance means and standard deviations for
selective mute and control groups
Selective
mute
Statement

Mean

SD

Control

Mean

SD

2.7
(.9) 3.1 (1.0) 3.86 .052
Reading rating by
teacher1
Math rating by teacher1 3.0
(.9) 3.0
(.9) .10 .92
Overall academic rating 2.8
(.8) 3.0
(.9) 1.35 .25
by teacher1
General performance
2.6 (1.0) 2.7 (1.1) .01 .92
at school2
WRAT-R Reading
95.0 (21.2) 96.1 (12.1) .09 .76
WRAT-R Arithmetic
90.2 (23.9) 91.6 (18.3) .11 .74
1
Reading, math, and overall academic performance ratings
ranged from 1 (far below grade level) to 5 (far above grade level).
2
Ranged from 1 (very well, excellent student) to 5 (not well at
all, poor student).

Table 6 Percentages and X2 analyses of home academic and

recreational activities in selectively mute and control groups
Measure

Selective mute

Academic activities at home

Parent reads to child
88.5
Child reads to parent
73.1
Child reads for fun
90.4
Used computers
51.9
Extracurricular recreational activities
Played with children
78.8
outside of school
Percent in sports
61.6
Percent in clubs
56.9
Percent in art or music
34.6

Control

X2

87.5
81.3
89.6
34.0

.02
.94
.02
.10

.56
.23
.58
.10

89.6

.18

.12

64.6
40.4
21.3

.84
.11
.18

.46
.08
.11

Table 7 Question 4: Will children have friends or be teased

and bullied? Means and standard deviations from parent and
teacher reports of social skills in selective mute and control
groups

Measure

Selective
Control
mute
Mean SD Mean SD

Social Skill Rating System (SSRS): Teacher reports

Social Cooperation
15.8 (4.2) 15.4 (4.9)
.21 .65
Social Assertion
4.7 (3.6) 12.1 (4.0) 91.48 <.001
Social Control
15.0 (6.0) 15.1 (4.2)
.01 .92
Social Skill Rating System (SSRS): Parent reports
Social Responsibility 9.5 (2.4) 13.8 (3.0) 65.87 <.001
Social Assertion
12.4 (2.7) 16.8 (2.8) 64.95 <.001
Social Control
11.7 (3.3) 13.0 (2.8) 4.80 .03
Social Cooperation
11.0 (3.5) 13.2 (3.5) 10.97 .001
Submissiveness and victimization: Teacher report
Submissive to Peers
3.1 (2.0) 2.5 (1.9) 2.70 .10
Victimized by Peers
2.2 (2.2) 2.5 (1.8)
.59 .45
Submissiveness and victimization: Parent report
Submissive to Peers
2.5 (2.4) 1.9 (2.1) 2.10 .15
Victimized by Peers
1.0 (1.5) 1.3 (1.9)
.78 .38

responsibility, social cooperation, and social control

subscales. While correlations between parent and
teacher SSRS ratings in the SM group were not

significant, among controls, parent and teacher ratings of social assertion, r(52) .32, p .023, social
cooperation, r(52) .31, p .027, and social control, r(52) .44, p .001 were significantly correlated.
Although a MANOVA with teacher and parent reports of victimization and submissiveness (Table 6)
was significant, F(4, 93) 2.50, p .048, univariate
comparisons were not significant.

Discussion
Question 1: Is selective mutism associated with
anxious or oppositional behavior?
Parents and teachers agreed that children with SM
were more anxious than controls. This is consistent
with previous descriptive (Black & Uhde, 1995;
Dummit et al., 1997; Steinhausen & Juzi, 1996) and
controlled studies (Bergman et al., 2002; Kristensen,
2001). In addition, parents reported more obsessivecompulsive symptoms and more frequent somatic
complaints among children with SM. Previous studies have also found OCD symptoms and somatic
complaints more common in children with SM than
controls (Kristensen, 2001).
The absence of group differences in depressive
symptoms is consistent with descriptive studies
using structured diagnostic interviews (Dummit
et al., 1997). Although we did not detect significant
differences (p .145), separation anxiety disorders
have been reported in 1731% of children with SM in
descriptive (Black & Uhde, 1995; Dummit et al.,
1997; Steinhausen & Juzi, 1996) and comparative
(Kristensen, 2000) studies. This discrepancy may
reflect different sampling strategies, our use of continuous versus categorical measurement, or the inability of the current sample to detect very small
differences between the SM and control groups.
Parents did not report differences in the ADHD,
ODD, or CD scores of the SM and control groups,
findings consistent with some descriptive (Black &
Uhde, 1995; Dummit et al., 1997) and comparative
studies (Bergman et al., 2002). Others, however,
have reported externalizing problems in children
with SM (Kristensen, 2001; Steinhausen & Juzi,
1996). Although parents of the SM group reported
difficulties with social cooperation, responsibility,
and control, these were apparently less severe than
the ADHD, ODD, and CD symptoms measured by
the OCHS-R scales.
At school, in contrast, teachers reported fewer
ODD (and ADHD) symptoms in the SM group.
Teachers also reported much lower ODD (and ADHD)
scores than parents of children with SM. This may
result from the inclusion of at least four speechrelated ODD (e.g., argues a lot with adults) and
three speech-related ADHD (e.g., talks excessively)
symptoms in the OCHS-R scales and DSM-IV criteria. Together, these results suggest that children

Selective mutism

with SM engage in impulsive or oppositional verbal

behavior at home but are inhibited at school.

Question 2: Is SM linked to parenting

or family dysfunction?
We found no differences in the marital status, economic resources, or support networks of families of
children with SM versus controls. This is consistent
with some previous large sample studies (Kristensen, 2001). Parents in our sample found children
with SM to be less cooperative and less compliant in
disciplinary settings, findings consistent with some
previous reports (Friedman & Karagan, 1973; Kristensen & Torgersen, 2001; Krohn et al., 1992). Although parents noted some disciplinary difficulties,
they did not report differences in the use of coercive
or permissive parenting strategies. Moreover, we
found no differences in family dysfunction or parental depression. This is consistent with Kristensens (2000) failure to find differences in psychiatric
disorders among the parents of 54 children with SM
and 108 controls. The absence of structural or
functional differences in families of children with SM
and controls questions the generality of case reports
linking SM to family dysfunction (Brown et al., 1975;
Goll, 1979; Pustrom & Speers, 1964; Rosenberg &
Linblad,1978; Wright, 1968).

Question 3: Will children with SM fail at school?

Contrary to predictions, the math and reading
scores of children with SM did not differ from those
of controls. Although teachers tended to rate
reading skills as lower in children with SM, the
absence of group differences on objective academic
tasks was supported by teacher ratings of mathematics and general classroom performance. These
results are consistent with those of Dummit et al.
(1997) who reported that only 11% of children with
SM had speech, language, or learning disabilities.
Interestingly, Kumpulainen et al. (1998) found that
children with SM whose academic performance was
at or above average were less likely to speak to
teachers.
This study suggests several mechanisms that may
protect children with SM from academic failures.
First, at school, the Social Cooperation scores of the
SM and control groups did not differ. The nonverbal
behaviors measured by the SSRS Social Cooperation
subscale (e.g., attending to instructions, following
directions, managing transitions, finishing assignments on time, producing correct work, waiting for
help, ignoring peer distractions, etc.) may allow
children with SM to succeed at classroom activities
without speaking. Second, children with SM evidenced fewer of the attentional or oppositional problems that might contribute to academic difficulties
(Hinshaw, 1992). Third, we observed no differences
in home activities (e.g., parentchild reading) that

1369

support academic development (Senechal & LeFevre,

2002). Fourth, a significant majority of the principal
caregivers of children with SM were employed, a
factor linked to early math and reading achievement
(Vandell & Ramanan, 1992). Finally, children with
SM were not differentially exposed to contextual
risks such as economic disadvantage, ineffective
parenting, or family dysfunction that might contribute to educational difficulties (Byrd & Weitzman,
1994).

Question 4: Will my child develop friendships

or be teased and bullied?
As predicted, parents and teachers reported deficits
in the verbally mediated social behavior of children
with SM. Children with SM scored significantly lower
on the SSRS Social Assertiveness scale which reflects the extent to which children joined groups,
introduced themselves, started conversations, or
invited friends to their house.
Children with SM were not victimized more than
controls. This seems consistent with Kumpulaninen
et al.s (1998) report that only 5% of children with
SM were victimized by peers. This study suggests a
combination of child characteristics and social
mechanisms that might protect children with SM
from victimization by peers. First, victimization by
peers is linked to submissive child behavior (Boivin,
Hymel, & Bukowski, 1995; Olweus, 1994; Schwartz,
Dodge, & Coie, 1993). Among children with SM, for
example, those rated more submissive by teachers
were more likely to be victimized at school,
r(48) .51, p < .001. In the present study, neither
parents nor teachers felt children with SM were more
submissive than controls. Similarly, children with
SM rated as more assertive by parents, r(48) ).31,
p < .001, and teachers, r(46) ).30, p .05, were
less likely to be victimized at school. Second, children with SM did not evidence the disruptive behavior associated with victimization by peers
(Schwartz, McFadyen-Ketchum, Dodge, Pettit, &
Bates, 1999). Children with SM with higher ODD
scores at school were more likely to be victimized by
peers, r(47) .32, p .028. Third, victimization by
peers is more closely related to depression than to
anxiety (Hawker & Boulton, 2000). While children
with SM were more anxious but not depressed, victimization at school increased as a function of parent-reported child anxiety scores, r(47) .31,
p .04, rather than depression.
There are also several family or social factors which
might prevent children with SM from being victimized
by peers. Children with SM, for example, were not
differentially exposed to the punitive discipline or
family dysfunction that has been linked to victimization by peers (Schwartz, Dodge, Pettit, & Bates,
2000). In the present study, however, coercive parenting was uncorrelated with victimization at school.
Finally, while children with SM were as likely as

1370

Charles E. Cunningham et al.

controls to be enrolled in the sports, recreational activities, and after-school playtimes with peers which
build the friendships that limit victimization by peers
(Hodges, Boivin, Vitaro, & Bukowski, 1999), these
factors were uncorrelated with victimization by peers.

Limitations of the study

To maximize our SM sample size, we employed a
case-finding approach to enrollment rather than a
prospectively screened community sample. It could
be argued that children receiving clinical service represent a more severely impaired subgroup of children with SM. Indeed, while we enrolled participants
who were several years younger than those in most
previous studies, the duration of mutism (6.9 years)
suggests our sample constituted a persistent subset
of the population.
Parents initiated contact with our service and
volunteered for this study. While we provide a publicly funded, universally available regional service,
our sample may include a disproportionate number
of motivated families. The composition (e.g., parents,
teachers, mental health professionals, etc.) of the
community workshops where we recruited participants does not allow us to determine the percentage of
families volunteering or the characteristics of those
who did not volunteer. While the similarity of our SM
and control groups on a wide range of demographic
measures suggests the SM group was representative
of the general population, the extent to which our
sample is representative of children with SM in
general is less certain.
Our approach to academic data collection also
deserves consideration. We assessed the arithmetic
and reading skills of both groups in the childs home.
For children with SM this is, presumably, a less
anxiety-provoking test situation than the classroom.
Moreover, when children with SM would not speak to
interviewers, parents administered reading vocabulary and verbal math questions. Although blind
coding of videotapes would have provided a more
reliable measure of these parent-administered tasks,
many children with SM refuse to speak when videotaped. While it could be argued that home assessments overestimated classroom performance, the
finding that teacher reports also failed to differentiate the academic skill of the two groups suggests this
setting may provide a better estimate of the actual
skills of children with SM.
While multiple informants (parents and teachers)
improve the measurement of victimization (Ladd &
Kochenderfer-Ladd, 2002), teachers do not detect
most bullying episodes (Olweus, 1994) and children
often fail to report bullying or harassment (Olweus,
1994). Observational studies need to determine
whether parents and teachers may underestimate
victimization in this sample.
Acknowledging these limitations, this study confirms the association between internalizing disorders

and SM. This study extends previous research by

exploring parenting and family functioning, measuring academic performance objectively, and examining social competence. Finally, this is the first
study of child characteristics and social factors that
may compensate for the failure to speak at school
and prevent children with SM from being victimized
by peers.

Author notes
Charles E. Cunningham, Department of Psychiatry
and Behavioural Neurosciences, McMaster University.

Acknowledgements
This research was supported by an Ontario Mental
Health Foundation Senior Research Fellowship to
Dr. Cunningham and the Jack Laidlaw Chair in Patient-Centred Health Care at McMaster University
Faculty of Health Sciences. We express our appreciation to the research teams who provided assistance with this project and to Donna Bohaychuk and
Laura Matsos for editorial assistance.

Correspondence to
Charles E. Cunningham, Chedoke Campus, Hamilton Health Sciences, Hamilton, Ontario, Canada,
L8N 3Z5; Email: cunningh@mcmaster.ca

References
American Guidance Service. Retrieved January 6,
2003 from http://www.agsnet.com./assessments/
technical/ssrs.asp.
American Psychiatric Association. (1994). Diagnostic
and statistical manual of mental disorders (4th edn).
Washington, DC: Author.
Anstendig, K.D. (1999). Is selective mutism an anxiety
disorder? Rethinking its DSM-IV classification. Journal of Anxiety Disorders, 13, 417434.
Bergman, R.L., Piacentini, J., & McCracken, J.T.
(2002). Prevalence and description of selective mutism in a school-based sample. Journal of the American Academy of Child and Adolescent Psychiatry, 41,
938946.
Black, B., & Uhde, T.W. (1995). Psychiatric characteristics of children with selective mutism: A pilot study.
Journal of the American Academy of Child and
Adolescent Psychiatry, 34, 847856.
Boivin, M., Hymel, S., & Bukowski, W.M. (1995). The
roles of social withdrawal, peer rejection, and victimization by peers in predicting loneliness and
depressed mood in children. Development and
Psychopathology, 7, 765786.
Boyle, M.H., Cunningham, C.E., Heale, J., Hundert, J.,
McDonald, J., Offord, D.R., & Racine, Y. (1999).
Helping children adjust A Tri-Ministry Study:

Selective mutism

I. Evaluation Methodology. Journal of Child Psychology and Psychiatry, 40, 10511060.

Boyle, M.H., Offord, D.R., Hofmann, H.G., Catlin, G.P.,
Byles, J.A., Cadman, D.T., Crawford, J.W., Links,
P.S., Rae-Grant, N.I., & Szatmari, P. (1987). Ontario
Child Health Study. I. Methodology. Archives of
General Psychiatry, 44, 826831.
Boyle, M.H., Offord, D.R., Racine, Y., Fleming, J.E.,
Szatmari, P., & Sanford, M. (1993). Evaluation of the
Revised Ontario Child Health Study Scales. Journal of
Child Psychology and Psychiatry, 34, 189213.
Bradley, S., & Sloman, L. (1975). Elective mutism in
immigrant families. Journal of the American Academy
of Child and Adolescent Psychiatry, 14, 510514.
Brown, B., Fuller, J., & Gericke, C. (1975). Elective
mutism: A review and a report of an unsuccessfully
treated case. Journal of the Association of Workers for
Maladjusted Children, 3, 4963.
Brown, G.J., & Lloyd, M.A. (1975). A controlled study of
children not speaking at school. Journal of the
Association of Workers for Maladjusted Children, 3,
4963.
Byles, J., Byrne, C., Boyle, M.H., & Offord, D.R. (1988).
Ontario Child Health Study: Reliability and validity of
the general functioning subscale of the McMaster
Family Assessment Device. Family Process, 27,
97104.
Byrd, R.S., & Weitzman, M.L. (1994). Predictors of early
grade retention among children in the United States.
Pediatrics, 93, 481487.
Cline, T., & Baldwin, S. (1994). Selective mutism in
children. San Diego: Singular Publishing Group, Inc.
Cunningham, C.E., Boyle, M., Offord, D., Racine, Y., &
Hundert, J. (1999). The emergence of victimization by
peers: Behavioral and social correlates. Presented at
the Society for Research in Child Development,
Albuquerque, New Mexico.
Cutrona, C.E., & Russell, D.W. (1987). The provision
of social relationships and adaptation to stress.
Advances in Personal Relationships, 1, 3767.
Dummit, E.S., Klein, R.G., Tancer, N.K., Asche, B.,
Martin, J., & Fairbanks, J.A. (1997). Systematic
assessment of 50 children with selective mutism.
Journal of the American Academy of Child and
Adolescent Psychiatry, 36, 653660.
Epstein, N.B., Baldwin, L.M., & Bishop, D.S. (1983).
The McMaster Family Assessment Device. Journal of
Marital and Family Therapy, 9, 171180.
Fast Track Project. (2002). Parenting Practice Inventory.
Retrieved December 2, 2002, from the Fast Track
Project Web site, http:/www.fasttrackproject.org/
techreport/p/ppi/parentingpractices.htm.
Ford, M.A., Sladeczek, I.E., Carlson, J., & Kratochwill,
T.R. (1998). Selective mutism: Phenomenological
characteristics. School Psychology Quarterly, 13,
192227.
Friedman, R., & Karagan, N. (1973). Characteristics
and management of elective mutism in children.
Psychology in the Schools, 10, 249252.
Fundudis, T., Kolvin, I., & Garside, R.F. (1979). Speech
retarded children: Their psychological development.
London: Academic Press.
Goll, K. (1979). Role structure and subculture in
families of elective mutism. Family Process, 18,
5568.

1371

Gresham, F.M., & Elliott, S.N. (1990). Social Skill Rating

System. Circle Pines, MN: American Guidance
Service.
Hawker, D.S., & Boulton, M.J. (2000). Twenty years
research on peer victimization and psychosocial maladjustment: A meta-analytic review of cross-sectional
studies. Journal of Child Psychology and Psychiatry,
41, 441455.
Hinshaw, S.P. (1992). Academic underachievement,
attention deficits, and aggression: Comorbidity and
implications for intervention. Journal of Consulting
and Clinical Psychology, 60, 893903.
Hodges, E.V., Boivin, M., Vitaro, F., & Bukowski, W.M.
(1999). The power of friendship: Protection against an
escalating cycle of peer victimization. Developmental
Psychology, 35, 94101.
Jastak, S., & Wilkinson, G.S. (1984). The Wide Range
Achievement TestRevised Administration Manual.
Wilmington: Jastak Associates, Inc.
Kolvin, I., & Fundudis, T. (1981). Elective mute
children: Psychological development and background
factors. Journal of Child Psychology and Psychiatry,
22, 219232.
Kopp, S., & Gillberg, C. (1997). Selective mutism:
A population based study: A research note. Journal
of Child Psychology and Psychiatry, 38, 257262.
Kristensen, H. (2000). Selective mutism and comorbidity with developmental disorder/delay, anxiety disorder, and elimination disorder. Journal of the American
Academy of Child and Adolescent Psychiatry, 39,
249256.
Kristensen, H. (2001). Multiple informants report of
emotional and behavioural problems in a nation-wide
sample of selective mute children and controls. European Journal of Child and Adolescent Psychiatry, 10,
135142.
Kristensen, H., & Torgersen, S. (2001). MCMI-II personality traits and symptom traits in parents of children
with selective mutism: A casecontrol study. Journal
of Abnormal Psychology, 110, 648652.
Krohn, D.D., Weckstein, S.M., & Wright, H.L. (1992). A
study of the effectiveness of a specific treatment for
selective mutism. Journal of the American Academy of
Child and Adolescent Psychiatry, 31, 711718.
Kumpulainen, K., Rasanen, E., Raaska, H., & Somppi,
V. (1998). Selective mutism among second-graders in
elementary school. European Journal of Child and
Adolescent Psychiatry, 7, 2429.
Ladd, G.W., & Kochenderfer-Ladd, B. (2002). Identifying victims of peer aggression from early to middle
childhood: Analysis of cross-informant data for concordance, estimation of relational adjustment, prevalence of victimization, and characteristics of
identified victims. Psychological Assessment, 14,
7496.
Olweus, D. (1994). Annotation: Bullying at school:
Basic facts and effects of a school based intervention
program. Journal of Child Psychiatry and Psychology,
35, 11711190.
Pustrom, E., & Speers, R.W. (1964). Elective mutism in
children. Journal of the American Academy of Child
Psychiatry, 3, 287297.
Radloff, L.S. (1977). The CES-D scale: A self-report
depression scale for research in the general population. Applied Psychological Measurement, 1, 385401.

1372

Charles E. Cunningham et al.

Rosenberg, J.B., & Linblad, M.B. (1978). Behavior

therapy in a family context: Treating elective mutism.
Family Process, 17, 7782.
Schwartz, D., Dodge, K.A., & Coie, J.D. (1993). The
emergence of chronic peer victimization in boys play
groups. Child Development, 64, 17551772.
Schwartz, D., Dodge, K.A., Pettit, G.S., & Bates, J.E.
(2000). Friendship as a moderating factor in the
pathway between early harsh home environment and
later victimization in the peer group. The Conduct
Problems Prevention Research Group. Developmental
Psychology, 36, 646662.
Senechal, M., & LeFevre, J.A. (2002). Parental involvement in the development of childrens reading skill:
A five-year longitudinal study. Child Development,
73, 445460.
Schwartz, D., McFadyen-Ketchum, S.A., Dodge, K.A.,
Pettit, G.P., & Bates, J.E. (1999). Early behavior
problems as a predictor of later peer group victimization: Moderators and mediators in the pathways of
social risk. Journal of Abnormal Child Psychology,
27, 191201.

Steinhausen, H., & Juzi, C. (1996). Elective mutism: An

analysis of 100 cases. Journal of the American
Academy of Child and Adolescent Psychiatry, 35,
606614.
Vandell, D.L., & Ramanan, J. (1992). Effects of early
and recent maternal employment on children from
low income families. Child Development, 63, 938949.
Wolke, D., Woods, S., Bloofield, L., & Karstadt, L.
(2000). The association between direct and relational
bullying and behaviour problems among primary
school children. Journal of Child Psychology and
Psychiatry, 41, 9891002.
World Health Organization. (1992). International statistical classification of diseases and related health
problems, 1989 revision. Geneva: World Health
Organization.
Wright, H.L. (1968). A clinical study of children who
refuse to talk in school. Journal of the American
Academy of Child Psychiatry, 7, 603617.
Manuscript accepted 1 October 2003