JOURNAL OF APPLIED BEHAVIOR ANALYSIS 1977, 101 183-195 NUMBER 2 (SUMMER) 1977

SOCIAL-SKILLS TRAINING FOR UNASSERTIVE CHILDREN:

A MULTIPLE-BASELINE ANALYSIS1
MITCHELL R. BORNSTEIN, ALAN S. BELLACK, AND MICHEL HERSEN
UNIVERSITY OF PITTSBURGH

The effects of social-skills training consisting of instructions, feedback, behavior re-

hearsal, and modelling were examined in a multiple-baseline analysis in four unassertive
children. The treatment was effective in that the behaviors selected for modification
changed markedly. The effects of treatment generalized from trained to untrained items
(interpersonal situations requiring assertive responding) and gains were maintained at
the two- and four-week followup probe sessions. In addition, overall assertiveness in all
subjects increased from baseline assessment to the conclusion of treatment and into
followup.
DESCRIPTORS: verbal skills, social skills, training, multiple baseline, unassertive
children

In recent years, increased attention has been and his colleagues relating psychiatric disorder
accorded to the measurement and modification to social competence (Levine and Zigler, 1973;
of social-skill deficits in a variety of popula- Phillips and Zigler, 1961, 1964; Zigler and
tions (see Hersen and Bellack, 1976; Hersen Levine, 1973; Zigler and Phillips, 1960, 1961,
and Bellack, (in press); Hersen and Eisler, 1962). Among the important findings resulting
1976; Hersen, Eisler, and Miller, 1973, for from this interrelated series of studies is the sug-
reviews). Although much of the present ex- gestion that a psychiatric patient's posthospital
perimental work in social-skills training owes success is directly related to his premorbid level
its historical roots to the pioneering clinical of social competence or skill. Argyle and Ken-
efforts of Salter (1949), Wolpe (1958, 1969), don (1967) also presented data relating psy-
and Wolpe and Lazarus (1966) on assertion chiatric disorder to level of social skill. At the
training, a further source of inspiration stems clinical level, Gladwin (1967) underscored the
from the extensive studies conducted by Zigler importance of teaching social skills to patients
in order to improve their resistance to environ-
'This article is based on M.R.B.'s master's thesis, mental stresses.
conducted under the direction of the second and third Therapies such as assertive training that are
authors, and submitted to the Graduate Faculty of
Arts and Sciences, University of Pittsburgh, in partial aimed at ameliorating clients' and patients'
fulfillment of the requirements for the degree of social-skill deficits have been carefully evaluated
Master of Science. Reprints may be obtained from both in analogue studies (e.g., Hersen, Eisler,
either Alan S. Bellack, Clinical Psychology Center,
University of Pittsburgh, Pittsburgh, Pennsylvania Miller, Johnson, and Pinkston, 1973; Kazdin,
15260, or Michel Hersen, Department of Psychiatry, 1974; McFall and Marston, 1970) and in clini-
Western Psychiatric Institute and Clinic,University of cally oriented investigations using group com-
Pittsburgh School of Medicine, Pittsburgh, Pennsyl-
vania 15261. The authors thank Bill MacDonald of parison designs (e.g., Percell, Berwick, and
Falk Elementary School for his extensive cooperation, Beigel, 1974) and single case experimental
and Adam Adams, Sheila Barber, Rocco Cugine, strategies (e.g., Hersen, Turner, Edelstein, and
Margie Leiberton, Paul Pentalas, Trina Plakosh, Kay- Pinkston, 1975). The component techniques
lee Shank, Joe Thomas, John Watts, and Martin Wil-
liams for their varied contributions throughout this comprising assertive training have been exam-
project. ined by McFall and his colleagues with unasser-
183
184 MITCHELL R. BORNSTEIN, ALAN S. BELLACK, and MICHEL HERSEN
tive college students (McFall and Lillesand, who were shy, experienced difficulty communi-
1971; McFall and Marston, 1970; McFall and cating their anger, did not stand up for their
Twentyman, 1973) and by Hersen and his col- rights, conformed to the wishes of others, and
leagues with unassertive psychiatric inpatients experienced difficulty speaking clearly when
(Eisler, Hersen, and Miller, 1973; Hersen et al., frustrated.
1973). In addition, specific strategies for remedi- Lack of social skill appears to generate social
ating dating-skill deficiencies in college students failure. Poor competency as a child may set the
have been evaluated (e.g., Christensen and Arko- stage for inappropriate interpersonal functioning
witz, 1974; Curran, 1975; Curran and Gilbert, as an adult, accentuating the potential for psy-
1975; Twentyman and McFall, 1975). In gen- chiatric disorder. Along these lines, O'Connor
eral, these studies indicate that the most ac- (1969) pointed out that "a child who is grossly
tive ingredients in social-skills training leading deficient in social skills will be seriously handi-
to behavioral change include instructions, coach- capped in acquiring many of the complex behav-
ing, feedback, behavior rehearsal, and modelling. ioral repertoires necessary for effective social
In addition to the recent interest in the evalu- functioning.... Such negative experiences would
ation and training of interpersonal social skills be expected to reinforce interpersonal avoidance
in college students and psychiatric patients, the responses which, in turn, further impede the
need for research investigating the development development of competencies that are socially
of social skills in children has frequently been mediated" (p. 15). These findings strongly sug-
articulated. Zigler and Phillips (1961, 1962) gest the importance of effective treatment pro-
presented cogent arguments for conceptualizing grams for children deficient in social skills (cf.
psychopathology in a developmental context. Cowen, Gardner, and Zax, 1967).
In a longitudinal study in which the stability of The term "social skills" has generally been
childhood characteristics was examined, Kagan used inexplicitly in reference to the compound
and Moss (1962) found that "passive withdrawal abilities necessary for effective interpersonal
from stressful situations, dependency on the functioning. As described above, social-skill defi-
family, ease of anger arousal, involvement in cits of children are primarily described in a man-
intellectual mastery, social interaction anxiety, ner similar to assertiveness deficits. The term
sex-role identification, and pattern of sexual be- "assertiveness" refers to a subcategory of social
havior in adulthood were each related to reason- skills in which the emphasis is on the ". . . ability
ably analogous behavior dispositions during the to express both positive and negative feelings in
early school years" (p. 266). the interpersonal context without suffering con-
Several other investigators have described sequent loss of social reinforcement" (Hersen
populations of children who lack appropriate and Bellack, (in press) b). Children deficient in
social skills (e.g., Patterson, 1964; Robinson, this regard would require training to increase
Vitale, and Nitsche, 1962; Ross, Lacey, and their ability to stand up for their rights and ex-
Parton, 1965). Patterson (1964), using factor press both anger and positive feelings, such as
analytic techniques, identified withdrawn chil- appreciation.
dren who characteristically were cooperative but The traditional therapeutic approach with un-
passive and shy, looked off into space, and spoke assertive children has attempted to overcome
and reacted slowly. Robinson et al. (1962) iden- their inhibitions by providing a supportive atmo-
tified passive children who generally were reluc- sphere while waiting for the desired social behav-
tant to take initiative in addition to being appre- ior to emerge. Bandura and Walters (1963)
hensive about expressing antagonistic feelings. criticized this approach as being minimally ef-
Ross et al. (1965), also using factor analysis, fective and time consuming, and proposed alter-
described a population of withdrawn children natives to this traditional treatment, arguing that
 SOCIAL-SKILLS TRAINING 185

social imitation or modelling can be most effec- referred by their teachers. Of these 12, four met
tive in the modification of inhibited behavior in the criteria of at least three deficient verbal and
children. Analogue studies confirm that model- nonverbal target behaviors (e.g., poor eye con-
ling techniques can be used effectively to help tact, short speech duration, inaudible responses,
such children acquire relevant social behaviors and inability to make requests) during baseline
(e.g., Bandura, Grusec, and Menlove, 1967; assessment (see Eisler, Miller, and Hersen, 1973).
Bandura and Huston, 1961; Bandura, Ross, and Jane was an 8-yr-old female third grader re-
Ross, 1963; O'Connor, 1969, 1972). In addi- ferred because of her difficulty relating to peers.
tion to short-term analogue studies, there are a She was described as passive and she experienced
few clinical reports in which skill approaches difficulty expressing anger when appropriate,
to treatment have been applied successfully to was unable to refuse unreasonable requests, was
shy and withdrawn children (e.g., Patterson, oversensitive to criticism, and rarely volunteered
1972). in class.
However, the efficacy of the entire social-skill Tom was an 8-yr-old male third grader re-
treatment package (instructions, feedback, be- ferred because of difficulty in interpersonal situa-
havior rehearsal, and modelling), such as that tions. When in conflict with a peer, he would
applied to psychiatric patients by Hersen and his cry and report the incident to his teacher rather
colleagues (e.g., Hersen et al., 1975; Hersen, than directly confront his antagonist. His passiv-
Eisler, and Miller, 1974), has not been experi- ity led to derision by peers and inhibited the
mentally evaluated with unassertive children. development of close interpersonal relationships.
Moreover, in previous studies with children, Mary was a 10-yr-old female fifth grader re-
the sequential effects of treatment on the com- ferred because of her inability to stand up for
ponent behaviors comprising social skill have her rights. When in conflict with a peer, she
not been examined. In the present study, the assumed the passive role and reported the inci-
effects of social-skills training (specifically asser- dent to her teacher. She failed to express anger
tion training) on verbal and nonverbal compo- when appropriate.
nent measures were assessed in a multiple-base- Alice was an 11-yr-old sixth grader referred
line design across behaviors in four withdrawn because of hei inability to refuse unreasonable
children. Standard training situations involving requests, to express hostility when warranted,
interpersonal encounters, as in the Behavioral and to volunteer in class whether in small or
Assertiveness Test (Eisler, Miller, and Hersen, large groups. Typically, when in a conflict situa-
1973) and the Behavioral Assertiveness Test- tion, she sought the teacher as arbitrator.
Revised (Eisler, Hersen, Miller, and Blanchard,
1975), were adapted for use with children. In Experimental Setting
addition to evaluating generalization, a brief The behavioral assessments and training ses-
followup was conducted to assess durability of sions were conducted in a videotape studio that
results. included three chairs arranged in a triangle to
accommodate the subject and two role models.
METHOD Male and female role models (undergraduate
reseach assistants) sat next to the subject and
Subjects provided predetermined prompts to facilitate re-
Unassertive children ranging in age from 8 to sponses. An adjoining room separated by a one-
11 yr were recruited from the Falk Elementary way mirror contained videotape recording equip-
School of the University of Pittsburgh. Twelve ment. Role-played scenes were narrated by the
children described as excessively cooperative, senior author over the intercom from the control
passive, shy, unassertive, and conforming were room. Similarly, the senior author provided di-
186 MITCHELL R. BORNSTEIN, ALAN S. BELLACK, and MICHEL HERSEN
rection, instructions, and feedback from the con- back to you and says that she broke the
trol room during training sessions. Point.
Prompt: "I broke the point."
Behavioral Assessment 4. Narrator: "Imagine you're about to go to
Because previous studies with adult psychi- Art Class when Cindy asks you if she can
use your desk while you're gone. You
atric patients demonstrated that overall level of agree to let her use it, but tell her that
assertiveness and component behaviors could be you'll need it when you get back. When
reliably identified on the basis of role-played you come back from Art, Cindy says she
responses to a series of standard interpersonal still needs to use your desk.
situations (Eisler et al., 1975; Eisler et al., 1973), Prompt: "I still need to use your desk."
the Behavioral Assertiveness Test used in these Male or Female Model
prior investigations was modified and adapted 5. Narrator: Your class is going to put on a
for use with children. The resulting Behavioral play. Your teacher lists the parts, asking
Assertiveness Test for Children served as the for volunteers. She reads a part you like
dependent measure for the present study. It con- and you raise your hand. But (Steve/
sists of nine scenes, five involving a same-sex Sue) raises (his/her) hand after you and
role model and four involving an opposite-sex says that (he/she) would like to get the
role model. An attempt was made to include part.
situations that the subjects were likely to en- Prompt: "I want to play this part."
gage in daily with other children. Similarity Male Model
between the scenes and typical daily encounters 6. Narrator: You're playing a game of kick-
was expected to facilitate the subjects' ability ball in school and it's your turn to get
to respond as they might in the natural environ- up. But Bobbie decides he wants to get
ment. Six of the scenes (1, 2, 3, 6, 7, 8), ran- up first.
domly selected from the pool of nine, were used Prompt: "I want to get up."
as training items for all subjects. The remaining
7. Narrator: Imagine you're playing a game of
four squares in gym. You make a good
three scenes (4, 5, 9) comprised the gener- serve into Barry's square. But he says
alization series. Listed below are the nine that it was out and keeps the ball to
scenes of the Behavioral Assertiveness Test for serve.
Children. Prompt: "It's my turn to serve."
8. Narrator: You're in school and you brought
Female Model your chair to another classroom to watch
1. Narrator: You're part of a small group in a movie. You go out to get a drink of
science class. Your group is trying to water. When you come back Mike is sit-
come up with an idea for a project to ting in your seat.
present to the class. You start to give Prompt: "I'm sitting here."
your idea when Amy begins to tell hers 9. Narrator: Imagine you're standing in line
also. for lunch. Jon comes over and cuts in
Prompt: "Hey, listen to my idea." front of you.
2. Narrator: Imagine you need to use a pair of Prompt: "Let me cut in front of you."
scissors for a science project. Betty is During the initial assessment period, all sub-
using them, but promises to let you have jects were told about the research nature of the
them next. But when Betty is done she
gives them to Ellen. program as well as their volunteer status. It was
Prompt: "Here's the scissors, Ellen." explained that the program was expected to in-
3. Narrator: Pretend you loaned your pencil crease the interpersonal skills of those children
to Joannie. She comes over to give it who took part. If the subjects and their parents
 SOCIAL-SKILLS TRAINING 187

gave consent for participation, then assessment the different situations. When he/she is done
began. talking, you say to him/her what you would
The initial assessment consisted of three ad- say if you were really doing something with
ministrations of the entire Behavioral Assertive- that person."
ness Test for Children. The procedure described
for the initial assessment phase was the same as Following the practice scene, probe sessions
the procedure for all other probe administra- then proceeded as follows: (a) the narrator pre-
tions. The subject, along with a male and female sented a scene, (b) the role model delivered a
role model, sat in the videotape studio, while prompt (standard lead-in), (c) the subject then
the therapist provided instructions from the con- responded to the role model. The subjects' re-
trol room. In general, the subject was instructed sponses to the nine scenes were videotaped on
to respond as realistically as possible to the situa- three separate occasions per week (for four
tions from the Behavioral Assertiveness Test for weeks) and retrospectively rated on three verbal
Children. The following instructions were used and nonverbal components of assertive behavior
with each subject: and for overall assertiveness.
"Hi. I'm next door. Can you hear me? See Selection and Scoring of Target Behaviors
that box on the floor? Well, my voice comes
through there. You must imagine things some- A behavioral analysis of the subjects' re-
times. You probably pretend a lot of things. sponses was conducted to establish which com-
For instance, maybe you pretend that you're ponent behaviors appeared at a disproportion-
doing something with someone else. Well, ately low rate. To be considered deficient, a
we're going to imagine a lot of different behavior had to be consistently rated as low
things with you. At times we will pretend that across all three probe administrations of the
you are doing something in your classroom, in Behavioral Assertiveness Test for Children. As
gym, or in the lunchroom. Each imaginary sit- there are no objective criteria on which to iden-
uation will be about you doing something with tify low assertiveness, it was anticipated that
someone from school. When I describe each the selection would be based on subjective clini-
situation I want you to really try to pretend
that you are part of the situation. cal criteria. In fact, all four subjects gave mini-
"To make it even more real, John and Susan mal responses to all scenes, simplifying the
(the role models) who are in the room with selection process. By the conclusion of the assess-
you will pretend that they are the other peo- ment phase, three deficient target behaviors had
ple in the situations. They will be someone been identified for each of the four subjects.
from school in the different situations. They Jane, Alice, and Mary were rated deficient on
will say something to you and you try really eye contact, loudness of speech, and requests for
hard to imagine that they are those people. new behavior. Tom was rated deficient on eye
When they are finished talking, you say what contact, duration of speech, and requests for
you would say if you were really doing some- new behavior.
thing with that person. Do you know what I Listed below are the operational definitions
mean?
"O.K., why don't we try a situation. Let's of the target behaviors selected for modification
imagine you're with your friend in an ice and rated retrospectively from videotapes by one
cream store. He/she wants to know what kind pair of judges.
of ice cream you want." Role Model Prompt: Ratio of eye contact to speech duration. The
"What kind of ice cream do you want?" Sub- total length of time in seconds that the subject
ject responds. Narrator: "Can you taste it? Is it looked at the interpersonal partner while he/
good? . . . O.K., now remember you pretend she (the subject) was speaking was measured
that John or Susan is someone from school in for each scene. The ratio was computed by di-
188 MITCHELL R. BORNSTEIN, ALAN S. BELLACK, and MICHEL HERSEN

viding total duration of eye contact while speak- havior. As suggested by Hersen and Barlow
ing by the duration of speech. (1976, Chapter 4), two types of percentage
Loudness of speech. Loudness of speech for agreements were calculated for the dichotomous
each scene was rated on a five-point scale from occurrence-nonoccurrence judgments of requests
1 (very low) to 5 (appropriately loud). Raters for new behavior. First, in cases where both oc-
were trained so that a response judged to have currence and nonoccurrence of the behavior were
appropriately loud volume was rated 5. scored, percentage agreement was calculated by
Requests for new behavior. Verbal content dividing number of agreements by total number
requesting new behavior from the interpersonal of agreements plus total number of disagree-
partner was scored as occurring or not occurring ments multiplied by 100. Second, to obviate the
in each scene. Responses scored in this category possibility of an inflated estimate of agreement
required more than mere noncompliance. The due to agreements on the nonoccurrence of the
subject had to show evidence that he/she wanted behavior, percentage agreement was also calcu-
his/her interpersonal partner to change his/her lated for instances where only occurrence of the
behavior (e.g., he had to ask the boy who cut behavior was recorded. Percentage agreement
in front of him at the movie to step to the end was again obtained by dividing number of
of the line; or he had to request that his sister agreements by total number of agreements plus
wait her turn to sit next to their mother at total number of disagreements multiplied by
dinner). 100.
Overall Assertiveness Procedure
After all the above behaviors were rated, two Following baseline assessment, all subjects
additional judges not familiar with the purposes received three weeks of social-skills training con-
of the study rated the subjects' responses to sisting of three 15- to 30-min sessions per week.
scenes for overall assertiveness using a five-point Consistent with multiple-baseline strategies,
scale with 1 indicating "very unassertive" and training was applied sequentially and cumula-
5 indicating "very assertive". These global judg- tively to the three target behaviors over the three-
ments were based on Wolpe's (1969) definition week period. Specifically, during the first week,
of "hostile" and "commendatory" assertiveness, all subjects received training geared toward
which emphasizes the socially appropriate ex- increasing eye contact. In the second week, at-
pression of feelings. tention was directed to loudness of speech for
Jane, Mary, and Alice, and to speech duration
Reliability of Behavioral Measures (number of words spoken) for Tom. In the final
Interrater agreement was evaluated during all week, training for all subjects focused on increas-
experimental phases for half of the videotaped ing the number of requests, with attention still
probe sessions. The probe sessions used to calcu- directed to maintaining the second target behav-
late reliability were selected randomly. To pre- ior for Jane, Mary, and Alice.
vent observer drift and bias (Kent, O'Leary, Dia- Throughout training, six of the scenes on the
ment, and Dietz, 1974), both the primary and Behavioral Assertiveness Test for Children
secondary raters were uninformed as to which served as training scenes and the remaining
probe sessions were used. Pearson product mo- three scenes were used to assess generalization
ment correlations were calculated for ratings of from trained to untrained scenes (for all sub-
speech duration, eye contact, loudness of speech, jects). Generalization scenes were selected to
and overall assertiveness. parallel training scenes in content (e.g., sex of re-
Percentage of agreement between raters was spondent, setting of interaction). Order of scene
established for ratings of requests for new be- presentation was randomized throughout all as-
 SOCIAL-SKILLS TRAINING 189

sessment and training phases of the program. TRAYN SCENES

At the end of each treatment session, the entire
Behavioral Assertiveness Test for Children was BsIn. Social Skills Training Follow-up
administered as a probe. 1.00 I I
Social-skills training specifically involved the
following: (a) The therapist presented one of raS .40 .60
the scenes from the Behavioral Assertiveness 1i
; .20
Test for Children, the model delivered a prompt,
and the subject responded. (b) The therapist
provided the subject with feedback on his/her
performance, with reference to the specific target
behavior. (c) The therapist then discussed feed-
I
Go~s
1~~~~~~~~~
back with the subject to ensure that he/she
understood. (d) The role models then modelled
responses, with specific attention to the target
behavior. (e) Specific instructions were then
m
C

i
=

6_
A

given by the therapist concerning the target be- . ~ ~ . -

havior, followed by the subject responding a 5 ....,A

second time. (f) Rehearsal continued for a scene
until the therapist felt that the criterion for that
target behavior had been reached. (g) Training
then advanced to a new interpersonal situation,
ra

Iy ~'

proceeding in a similar fashion through all train-

ing scenes.
Once training was completed for all three tar- 1 3 5 7 9 11 2-4-
get behaviors, the program shifted to followup Probe Sessions Weeks
assessment. Followups consisted of probe assess- Fig. 1. Probe sessions during baseline, social-skills
ments at two- and four-week intervals after treat- Atreatment, and followup for training scenes for Jane.
multiple-baseline analysis of: ratio of eye contact
ment. while speaking to speech duration, loudness of speech,
number of requests, and overall assertiveness.
RESULTS ness of speech, and requests for new behavior
The results were similar for all four children. resulted in considerable increases over baseline
While individual results are presented graphi- levels. These changes were obtained without cor-
cally below, only the results for Jane, which responding changes in untreated baselines: ratio
were representative, are discussed. of eye contact to speech duration increased from
Interrater reliabilities were uniformly high 0 to 1.00, loudness of speech increased from
for all subjects. Pearson correlations were for the ratings between 1 and 2 to ratings of about 4,
most part in the mid 0.90s; percentage agree- and number of requests increased from zero to
ment ranged from 85 % to 100% for all ratings six. In addition, overall assertiveness gradually
calculated. increased during training, with a sizeable in-
The results of the multiple-baseline analysis crease when treatment for requests was intro-
for training scenes with Jane are presented in duced at about probe 10 (this pattern was evi-
Figure 1. Data points represent averages across dent with the other subjects as well). Ratings of
all scenes for each probe session. The sequential overall assertiveness increased from about 1 to
introduction of treatment for eye contact, loud- almost 5. Enhanced levels of performance were
190 MITCHELL R. BORNSTEIN, ALAN S. BELLACK, and MICHEL HERSEN

GEERALIZATION SCENES ment effectively enhanced performance in the

verbal and nonverbal components of assertive-
BsIn. Social Skills Training Follow-up ness that had been identified as deficient (i.e.,
aI

1.00 eye contact, loudness of speech, speech duration,

so..M .80 and requests for new behavior). Multiple-base-
,9I3 .60 line analyses indicated that targeted behaviors
-2
Z.40 were predominantly independent and that
.20
0 changes from baseline levels generally occurred
'3
only when training was directed to a specific be-
a havior. Thus, the sequential and cumulative ef-
fects of treatment were documented. In all cases,
'~i....
3 .........
gains obtained on treated scenes generalized to
untreated scenes. Finally, changes in both treated
3
I
-3
.. ..
.' and untreated scenes persisted over a one-month
posttreatment period.

I
._

I 2.......
. l1.00
BsIn.
TRAINING SCENES
Social Skills Training Follow-up

O-% = .806
*

* 40

C=
w.201
.j

1 3 5 7 9 112-4- V)
I-,
Weeks
Probe Sessions
@3
Fig. 2. Probe sessions during baseline, social-skills
treatment, and followup for generalization scenes for E

Jane. A multiple-baseline analysis of: ratio of eye

contact while speaking to speech duration, loudness k-
of speech, number of requests, and overall assertive- E
4-,

ness.
=n

e
maintained at two- and four-week evaluations E

a
after treatment. The results of the multiple-base-
(-3@3
line analysis for generalization scenes are pre- q

sented in Figure 2. These data parallel those

=2

@30
CA
obtained for the training scenes.
(A

DISCUSSION 1 3 5 7 9 11 2-4-
The results of this multiple-baseline analysis Probe Sessions Weeks
indicate that social-skills training generated con- Fig. 3. Probe sessions during baseline, social-skills
siderable improvement in both component be- treatment, and followup for training scenes for Tom.
A multiple-baseline analysis of: ratio of eye contact
haviors and overall assertiveness for all four while speaking to speech duration, number of words,
subjects in this direct replication series. Treat- number of requests, and overall assertiveness.
 ~!
SOCIAL-SKILLS TRAINING 191

wE mK rS Despite our reservation about external validity,

the screening battery used here seems to repre-
BsIn. Social Skills Training Follow-up sent a systematic method for identifying inade-
DOO . __ I quate social functioning in children. This would
present a substantial improvement over tradi-
tional classification of such behavior through
i
overall clinical impressions and/or teacher ob-
01 servations.
.4
b-
Second, the extent of modification and stabil-
8 ity of change in the targeted behavior provides
.1 123064 '. . . ........ . .. . . . i . strong support for the effectiveness of the social-
skills training package used (i.e., instructions,
2
2
feedback, behavior rehearsal, and modelling).
E
2 Gains in performance over baseline levels indi-
cate each child's improved facility to express
AI;
i_
3 ',-' TRAINING SCENES
. .
BsIn. Social Skills Training Follow-up
Ia _
1.0110 I. I

0 1:
D .= .610 I.
Ii .410
.210
1 . ... _ _ _ _ _ _
0
1 3
5 7 9 11 2-4- 5
Probe Sessions Weeks C-
Fig. 4. Probe sessions during baseline, social-skills CO)
treatment, and followup for generalization scenes for .4
44
Tom. A multiple-baseline analysis of: ratio of eye
contact while speaking to speech duration, number of
words, number of requests, and overall assertiveness. Is . .i
co i
w .......... .,.

Two major implications may be drawn from 4

___
these results. First, the data suggest that it is 2
possible to identify specific deficiencies in verbal
and nonverbal components of assertiveness in 0
children globally described as low in interper- 5
sonal social skill. These behaviors (eye contact,
loudness of speech, speech duration, and requests 3
for new behavior) were identified through the
use of a screening battery composed of 10 of 1
the 14 behavioral components of assertiveness 1 3 5 7 9 11 2- 4-
isolated in adult psychiatric patients by Eisler Probe Sessions Weeks
et al. (1973). However, there is a need to in- Fig. 5. Probe sessions during baseline, social-skills
vestigate the relationship between deficits in these treatment, and followup for training scenes for Mary.
components and children's level of interper- A multiple-baseline analysis of: ratio of eye contact
while speaking to speech duration, loudness of speech,
sonal functioning in their natural environment. number of requests, and overall assertiveness.
192 MITCHELL R. BORNSTEIN, ALAN S. BELLACK, and MICHEL HERSEN

Several questions have been raised by the

present findings. One of the foremost concerns
Bsln. Social Skills Training Follow-up the degree of generalization of the behavior
KI v IKI change obtained. While a measure of generaliza-
tion was included in the present study, it in-
volved role play with adults in situations similar
to those used in training. The degree of generali-
zation to interactions with other children in the
natural environment is, therefore, uncertain.
There are several ways in which such generali-
zation might be better assessed. Economic and
time limitations precluded the use of naturalistic
observations of subjects in this study. Such ob-
servations should be included in any replications,

TRIING SCENES
Bsin. Social Skills Training

1 3 5 7 9 11 2-4-
Probe Sessions Weeks
Fig. 6. Probe sessions during baseline, social-skills
treatment, and followup for generalization scenes for
Mary. A multiple-baseline analysis of: ratio of eye
contact while speaking to speech duration, loudness
of speech, number of requests, and overall assertive-
ness.

negative assertion to male and female role

models enacting hypothetical peers. However,
given the rapidity of behavioral change, along
with absence of severe disturbance in this subject
population, it seems unlikely that all of the
component techniques included in the treatment
package were necessary. Such a conclusion seems
consistent with previous data collected in ana-
logue studies with psychiatric patients (e.g., 1 3 5 7 9 11 2-4-
Hersen et al., 1973). Therefore, future investi- Probe Sessions Weeks
gations should focus on establishing the differen- Fig. 7. Probe sessions during baseline, social-skills
treatment, and followup for training scenes for Alice.
tial effectiveness of the component techniques A multiple-baseline analysis of: ratio of eye contact
when applied to a subject population of with- while speaking to speech duration, loudness of speech,
drawn children. number of requests, and overall assertiveness.
 SOCIAL-SKILLS TRAINING 193

inznmSCEcS nation of which deficient behaviors merit exten-

sive treatment is currently based on subjective
Bsin. Social Skills Taining Follow-up evaluation. Empirical criteria should be devel-
1.00 oped. For example, the present data suggest that
there is a high correlation between overall as-
sertiveness and requests for new behaviors. This
.6.0
. . . . . . could be a function of the definition of overall
assertiveness employed by the raters, or it could
be that verbal content makes a greater contri-
bution to assertiveness in children than does non-
X *5 I'- !
z~~~~* verbal behavior. Normative data should be col-
lected and evaluated so as to establish criteria
for assertiveness in normal children and to de-
termine the relative contributions of component
behaviors to overall assertiveness or effectiveness
A_'~~~~
' ' '~~~ ' ' ' ''' ' * 'S
of responses.
In addition to normative data on performance,
a pool of relevant scenes needs to be established
for the Behavioral Assertiveness Test for Chil-
dren, so that a standardized screening device can
be constructed for differentiating high and low
assertive children (cf. Eisler et al., 1975; Eisler
et al., 1973). Finally, it would be valuable to
develop more elaborate training procedures that
1 3 5 7 9 11 2-4- would allow for sequential responding such as
Probe Sessions Weeks rebuttal of the child's responses by the role
Fig. 8. Probe sessions during baseline, social-skills models. This type of interaction would simulate
treatment, and followup for generalization scenes for
Alice. A multiple-baseline analysis of: ratio of eye possible stress situations encountered by the
contact while speaking to speech duration, loudness of child in his natural environment and might lead
speech, number of requests, and overall assertiveness. to responding more representative of typical
functioning.
if at all possible. The use of other children as
role models might be more realistic than adult
models; however, it is unlikely that children REFERENCES
could be recruited who were both consistent Argyle, M. and Kendon, A. The experimental analy-
across trials and available for such a large num- sis of social performance. In L. Berkowitz (Ed),
ber of sessions. Finally, the presentation of en- Advances in experimental social psychology:
tirely novel situations after training could also Volume III. New York: Academic Press, 1967.
Bandura, A., Grusec, J. E., and Menlove, F. L. Vi-
indicate the generalizability of the new skills. carious extinction of avoidance behavior. Journal
Bellack, Hersen, and Turner (1976) employed of Personality and Social Psychology, 1967, 5,
this latter procedure with adult psychiatric pa- 16-23.
tients and found that the degree of generaliza- Bandura, A. and Huston, A. C. Identification as a
process of incidental learning. Journal of Ab-
tion varied across target behaviors. normal and Social Psychology, 1961, 63, 311-
Another major question pertains to the selec- 318.
tion of target behaviors for treatment. Both the Bandura, A., Ross, D., and Ross, S. A. Imitation of
film-mediated aggressive models. Journal of Ab-
identification of deficient behaviors and determi- normal and Social Psychology, 1963, 66, 3-11.
194 MITCHELL R. BORNSTEIN, ALAN S. BELLACK, and MICHEL HERSEN
Bandura, A. and Walters, R. H. Social learning and M. B., and Pinkston, S. G. Effects of practice,
personality development. New York: Holt, Rine- instructions, and modeling on components of as-
hart & Winston, 1963. sertive behavior. Behaviour Research and Therapy,
Bellack, A. S., Hersen, M., and Turner, S. M. Gener- 1973, 11, 443-451.
alization effects of social skills training in chronic Hersen, M., Turner, S. M., Edelstein, B. A., and
schizophrenics: an experimental analysis. Behav- Pinkston, S. G. Effects of phenothiazines and
iour Research and Therapy, 1976, 14, 391-398. social skills training in a withdrawn schizophrenic.
Christensen, A. and Arkowitz, H. Preliminary re- journal of Clinical Psychology, 1975, 31, 588-
port on practice dating and feedback as treatment 594.
for college dating problems. Journal of Counseling Kagan, J. and Moss, H. A. Birth to maturity: a
Psychology, 1974, 21, 92-95. study in psychological development. New York:
Cowen, E., Gardner, E. A., and Zax, M. (Eds) Emer- Wiley, 1962.
gent approaches to mental health problems. New Kazdin, A. E. Effects of covert modeling and model
York: Appleton-Century-Crofts, 1967. reinforcement on assertive behavior. Journal of
Curran, J. P. Social skills training and systematic Abnormal Psychology, 1974, 83, 240-252.
desensitization in reducing dating anxiety. Be- Kent, R. N., O'Leary, K. D., Diament, C., and Dietz,
haviour Research and Therapy, 1975, 13, 65-68. A. Expectation biases in observational evalua-
Curran, J. P. and Gilbert, F. S. A test of the relative tion of therapeutic change. Journal of Consult-
effectiveness of a systematic desensitization pro- ing and Clinical Psychology, 1974, 42, 774-780.
gram and an interpersonal skills training program Levine, J. and Zigler, E. The essential-reactive dis-
with date anxious subjects. Behavior Therapy, tinction in alcoholism: a developmental approach.
1975, 6, 510-521. Journal of Abnormal Psychology, 1973, 81, 242-
Eisler, R. M., Hersen, M., and Miller, P. M. Effects 249.
of modeling on components of assertive behavior. McFall, R. M. and Lillesand, D. B. Behavior re-
Journal of Behavior Therapy and Experimental hearsal with modeling and coaching in assertion
Psychiatry, 1973, 4, 1-6. training. Journal of Abnormal Psychology, 1971,
Eisler, R. M., Hersen, M., Miller, P. M., and Blanchard, 77, 313-323.
E. B. Situational determinants of assertive be- McFall, R. M. and Marston, A. R. An experimental
haviors. Journal of Consulting and Clinical Psy- investigation of behavior rehearsal in assertive
chology, 1975, 43, 330-340. training. Journal of Abnormal Psychology, 1970,
Eisler, R. M., Miller, P. M., and Hersen, M. Com- 76, 295-303.
ponents of assertive behavior. Journal of Clinical McFall, R. M. and Twentyman, C. T. Four experi-
Psychology, 1973, 29, 295-299. ments on the relative contributions of rehearsal,
Gladwin, T. Social competence and clinical practice. modeling, and coaching to assertion training.
Psychiatry, 1967, 30, 30-43. Journal of Abnormal Psychology, 1973, 81, 199-
Hersen, M. and Barlow, D. H. Single case experi- 218.
mental designs: strategies for studying behavior O'Connor, R. D. Modification of social withdrawal
change. New York: Pergamon Press, 1976. through symbolic modeling. Journal of Applied
Hersen, M. and Bellack, A. S. Social skills training Behavior Analysis, 1969, 2, 15-22.
for chronic psychiatric patients: Rationale, re- O'Connor, R. D. Relative effects of modeling, shap-
search findings, and future directions. Compre- ing, and the combined procedures for modification
hensive Psychiatry, 1976, 17, 559-580. of social withdrawal. Journal of Abnormal Psy-
Hersen, M. and Bellack, A. S. Assessment of social chology, 1972, 79, 327-334.
skills. In A. R. Ciminero, K. S. Calhourn, and Patterson, G. R. An empirical approach to the classi-
H. E. Adams (Eds), Handbook for behavioral fication of disturbed children. Journal of Clinical
assessment. New York: Wiley, (in press). Psychology, 1964, 20, 326-337.
Hersen, M. and Eisler, R. M. Social skills training. Patterson, R. L. Time-out and assertive training for
In W. E. Craighead, A. Kazdin, and M. J. Ma- a dependent child. Behavior Therapy, 1972, 3,
honey (Eds), Behavior modification: principles, 466-468.
issues, and applications. Boston: Houghton Mifflin, Percell, L. P., Berwick, P. T., and Beigel, A. The
1976. Pp. 361-375. effects of assertive training on self-concept and
Hersen, M., Eisler, R. M., and Miller, P. M. Develop- anxiety. Archives of General Psychiatry, 1974, 31,
ment of assertive responses: Clinical, measure- 502-504.
ment, and research findings. Behaviour Research Phillips, L. and Zigler, E. Social competence: The
and Therapy, 1973, 11, 505-521. action-thought parameter and vicariousness in
Hersen, M., Eisler, R. M., and Miller, P. M. An normal and pathological behaviors. Journal of
experimental analysis of generalization in asser- Abnormal and Social Psychology, 1961, 63, 137-
tive training. Behaviour Research and Therapy, 146.
1974, 12, 295-310. Phillips, L. and Zigler, E. Role orientation, the
Hersen, M., Eisler, R. M., Miller, P. M., Johnson, action-thought dimension and outcome in psychi-
 SOCIAL-SKILLS TRAINING 195
atric disorder. Journal of Abnormal and Social Wolpe, J. and Lazarus, A. A. Behavior therapy
Psychology, 1964, 68, 381-389. techniques. New York: Pergamon Press, 1966.
Robinson, J. F., Vitale, L. J., and Nitsche, C. J. Be- Zigler, E. and Levine, J. Premorbid adjustment and
havioral categories of childhood. American Jour- paranoid-nonparanoid status in schizophrenia.
nal of Psychiatry, 1962, 117, 806-810. Journal of Abnormal Psychology, 1973, 82, 189-
Ross, A. O., Lacey, H. M., and Parton, D. A. The 199.
development of a behavior checklist for boys. Zigler, E. and Phillips, L. Social effectiveness and
Child Development, 1965, 36, 1013-1027. symptomatic behaviors. Journal of Abnormal and
Salter, A. Conditioned reflex therapy. New York: Social Psychology, 1960, 61, 231-238.
Capricorn, 1949. Zigler, E. and Phillips, L. Social competence and
Twentyman, C. T. and McFall, R. M. Behavioral outcome in psychiatric disorder. Journal of Ab-
training of social skills in shy males. Journal of normal and Social Psychology, 1961, 63, 264-271.
Consulting and Clinical Psychology, 1975, 43, Zigler, E. and Phillips, L. Social competence and the
384-395. process-reactive distinction in psychopathology.
Wolpe, J. Psychotherapy by reciprocal inhibition. Journal of Abnormal and Social Psychology, 1962,
Stanford: Stanford University Press, 1958. 65, 215-222.
Wolpe, J. The practice of behavior therapy. New
York: Pergamon Press, 1969. Received 3 February 1976.
(Final acceptance 2 June 1976.)