JOFT 22.

2 - Main Text 14/4/0 10:59 am Page 128

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 The Association for Family Therapy 2000. Published by Blackwell Publishers, 108 Cowley
Road, Oxford, OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.
Journal of Family Therapy (2000) 22: 128–143
0163–4445

The Beavers Systems Model of Family Functioning

Robert Beaversa and Robert B. Hampsonb

Family competence and family style are the two main dimensions of the
Beavers Systems Model of Family Functioning. The competence dimen-
sion ranges from optimal through adequate, midrange and borderline to
severely dysfunctional. The style dimension ranges from centripetal to
centrifugal. When the two dimensions are combined, they diagramatically
define nine distinct family groupings, three of which are relatively func-
tional and six of which are thought to be sufficiently problematic to
require clinical intervention. A family’s status on the competence and style
dimensions may be established with the Beavers interactional scales. The
self-report family inventory may be used to evaluate family members’
perceptions of their status on the competence dimension. The reliability
and validity of the self-report instrument and observational rating scales
have been documented in over thirty papers and books published by the
Beavers research team since 1970. The model has proved useful in train-
ing, research and clinical work.

Overview
The Beavers Systems Model offers a cross-sectional perspective on
family functioning. Family competence is conceptualized as falling
along one dimension and family style is viewed as falling along a
second orthogonal dimension. Figure 1 is a diagram of this model.
The horizontal axis – family competence – relates to the struc-
ture, available information and adaptive flexibility of the system. In
systems terms, this may be called a negentropic continuum, since
the more negentropic (flexible and adaptive) a family, the more the
family can negotiate, function and deal effectively with stressful situ-
ations. High competence requires both structure and the ability to
change structures. There is a complex interaction of morphogenic

a Executive Director, Robert Beavers Family Studies Center, Southern

Methodist University, Dedman College, PO Box 750442, Dallas TX 75275-0442,
USA.
b Associate Professor of Psychology, Southern Methodist University, Dallas,
USA.

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 129

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 129

Figure 1. The Beavers Systems Model of Family Functioning

and morphostatic features. Capable families intuitively have a

systems approach to relationships, with an appreciation of the inter-
changeability of causes and effects and the circularity of systems
phenomena. When a family is not bound to rigid behaviour
patterns and responses, it has more freedom to evolve and differ-
entiate. Bertalanffy (1968) wisely said, ‘system sickness is system
rigidity’.

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 130

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
130 Robert Beavers and Robert Hampson
The vertical axis – family style – relates to a stylistic quality of family
interaction. With respect to healthy functioning, it is not a unidirec-
tional continuum. Rather, it bears a curvilinear relationship to healthy
functioning and ranges from centripetal to centrifugal. Centripetal
families view most relationship satisfactions as coming from within the
family rather than from the outside world. Conversely, centrifugal
families see the outside world as holding the most promise of satisfac-
tion and the family as holding the least. The arrow shape of the
diagram is designed to illustrate that extremes of style – either
profoundly centrifugal or centripetal – are associated with poor family
functioning. As a family becomes more competent, excessive
centripetal or centrifugal styles diminish. Competent families change
and adapt in various ways in order to meet individual members’ needs.
For example, a family with small children is appropriately more
centripetal. As the family matures and children reach late adoles-
cence, a more centrifugal pattern is expected to be optimally adaptive.
Nine family groupings may be defined on the basis of families’
positionings along the dimensions of competence and style. The
following description of the characteristics of the nine different
family groupings specified in our model is based on both clinical
observation and empirical research (Beavers, 1977, 1981a, 1981b,
1982, 1985, 1989; Beavers and Hampson, 1990, 1993; Hampson and
Beavers, 1996a, 1996b; Lewis et al., 1977).

Group 1. Optimal families

Optimal families serve as our model for effective functioning. The
family members have what can be described as a systems orienta-
tion. They realize that many causes interact to produce a given
result, and that causes and effects are interchangeable (e.g. harsh
discipline leads to aggressive behaviour and aggressive behaviour
invites harsh discipline). Intimacy is sought and generally found. It
is a function of frequent, equal-powered transactions along with
mutual respect for differing family members’ viewpoints. Individual
choice and perceptions are respected, allowing for capable negoti-
ation and excellent group problem-solving. Individuation of each
person is highly evolved and boundaries are clear. There is conflict,
but it is usually resolved quickly.

Group 2. Adequate families

Adequate families are contrasted with optimal families in that the

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 131

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 131
former are more control oriented and often attempt to resolve
conflict by intimidation and direct force. Therefore, greater overt
power is sought by family members and the parental coalition is less
emotionally rewarding, though usually effective. While still
adequate, there is less intimacy and trust in these families, less joy
and less spontaneity. Role stereotyping, particularly sex-role stereo-
typing, with conventional, powerful, unemotive males counter-
pointed by relatively less powerful, emotive and frequently
depressed women, is usual.

Groups 3, 4 and 5. Mid-range families

The first three groups of dysfunctional families are termed ‘mid-
range’. These families usually contain functional but vulnerable
children, and both parents and children are susceptible to psycho-
logical problems. Mid-range families are concerned with control
and overt power differences. Power struggles and discipline without
negotiation are usual. Members of mid-range families assume that
people are basically antisocial and therefore their control efforts
are believed to be essential. Family members do not have boundary
problems. Although there are frequent projections, the family roles
allow for rebuttal, and invasion of one member’s inner space by
another is resisted. Further, one often sees favourite children in
mid-range families. These favourites may be different for each
parent – mother selects a son, father a daughter – or they may team
up and select an agreed-upon favourite and possibly a scapegoat.
Ambivalence is frequently handled by denying one half of a pair
of strong feelings and using repression or projection for the other.
For example, ‘I like to go out and you like to stay at home’ or ‘You
are too strict with the kids’ versus ‘No, you are too lenient’. There
is a pervasive belief that people ‘really’ have one feeling: ‘He really
loves me, though he is contemptuous’ or ‘She really hates me
though she tries to be nice’. Three types or styles of mid-range fami-
lies – centripetal, centrifugal and mixed – will be discussed below.
Members of a mid-range centripetal family expect overt, author-
itarian control to be successful. Parental manipulation or indirect
control is minimal. The expression of hostility is not approved and
is therefore covert. Expressions of caring are approved. There is
only modest spontaneity and great concern for rules and authority.
Sex stereotyping is at a maximum in this group. Childlike women
and strong, silent males abound.

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 132

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
132 Robert Beavers and Robert Hampson
Mid-range centrifugal families also attempt to use control by
intimidation but do not expect their efforts to be successful. Open
hostility, blame and attack are frequent. Expressions of warmth and
caring are anxiety-provoking. Easily unleashed negative feelings
provide the energy for change in centrifugal families. Parents spend
little time in the home, and children move out into the neighbour-
hoods and streets much earlier than the norm. The parental coali-
tion is tenuous, with unresolved power issues openly displayed.
Mid-range mixed families have enough alternating and compet-
ing centripetal and centrifugal behaviour to disqualify them from
an extreme stylistic position and thus are in a mixed position within
the mid-range.

Groups 6 and 7. Borderline families

Borderline families present with chaotic overt power struggles alter-
nating with ineffective but persistent efforts to establish domi-
nance/submission patterns. Individual family members have little
skill in meeting emotional needs, either their own or those of
others. The families are neither as dysfunctional as the severely
disturbed group nor as effective in establishing control-oriented
stability as the mid-range families.
In borderline centripetal families the chaos is more verbal than
behavioural, and control battles are intense but usually covert.
Open rebellion or covertly expressed rage is not expected, that is,
not within the family rules. Severely obsessional and anorectic
patients may sometimes be found in these families.
Borderline centrifugal families are much more open in the
expression of anger. The parental coalition is notably poor, and
stormy battles occur regularly. Children learn to manipulate the
unstable but oscillating parental subsystem and sometimes receive a
label of borderline personality disorder.

Groups 8 and 9. Severely dysfunctional families

The severely dysfunctional family’s greatest deficit is in the domain
of communication and its greatest need is for communicational
coherence. Consequently, this group is most limited in negotiating
and adaptive capacity. Family members have little ability to resolve
ambivalence and to choose and pursue goals. There is a lack of a
shared focus of attention in discussion and an emotional distancing

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 133

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 133
that precludes satisfying encounters. Overt power is not clearly held
by anyone in the family. Family functioning appears chaotic, since
control is carried on by a variety of covert and indirect means.
Severely disturbed centripetal families have a tough, nearly
impermeable outer boundary, and the family may be seen by neigh-
bours as unusual. Children may be delayed in their progression
through normal sequences of emotional development. In these
families there is a powerful conflict between the developmental
need for separation/individuation and the family’s insistence on
togetherness and extreme family loyalty.
Severely disturbed centrifugal families have a tenuous boundary
between the family and the community, with frequent member
leave-taking, much open hostility, and great contempt for depen-
dency, vulnerability, human tenderness and warmth. This contrasts
with the severely disturbed centripetal family’s characteristics, but
the confused, incomplete transactions and severely disturbed level
of adaptability are quite similar. Children from severely disturbed
centrifugal families may be as limited in social-emotional develop-
ment as those from severely disturbed centripetal families.

Description of the self-report instruments and clinical

rating scales
A family may be classified into one of the nine categories or family
groupings described in the previous section on the basis of their
scores on the Beavers interactional scales (Beavers and Hampson,
1990). A family’s position along the competence dimension may be
established with the self-report family inventory. A description of
these two instruments will be given below.

Beavers interactional scales

There are two Beavers interactional scales: The Beavers
Interactional Competence Scale and the Beavers Interactional
Style Scale. Both are designed for use by trained raters who have
observed an episode of family interaction in which family
members discuss the following question for ten minutes: What
would you like to see changed in your family? Each interactional scale
is made up of a number of five- or ten-point subscales, with ver y
concrete descriptions of what type of families should be given
particular ratings.

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 134

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
134 Robert Beavers and Robert Hampson
Beavers Interactional Competence Scale
The Beavers Interactional Competence Scale is designed to assess a
family’s overall level of health and competence (Beavers and
Hampson, 1990). The family competence scale is composed of the
following thirteen subscales.
1. Structure of the family:
• Overt power (from chaotic to egalitarian)
• Parental coalitions (from parent–child coalition to strong
parental coalition)
• Closeness (from indistinct boundaries to distinct boundaries).
2. Mythology (from congruent to incongruent).
3. Goal-directed negotiation (from extremely efficient to extremely
inefficient).
4. Autonomy:
• Clarity of expression (from very clear to unclear)
• Responsibility (from regular to rare acceptance of responsibil-
ity for actions)
• Permeability (from very open to unreceptive).
5. Family affect:
• Range of feelings (from direct expression of a wide range to
little expression)
• Mood and tone (from warm and optimistic to cynical and
pessimistic)
• Unresolvable conflict (from severe unresolved conflict to
none)
• Empathy (from consistent empathy to none).
6. Global health pathology (from pathological to healthy).
In one of our investigations (Beavers and Hampson, 1990) there
were three pairs of raters who reached at least 90% overall reliabil-
ity in training and maintained a minimum of 85% reliability
throughout the study. Interrater reliabilities expressed as Kappa
coefficients ranged from .76 (closeness scale) to .88 (range of feel-
ings), and a Kappa coefficient of .86 was obtained for the global
competence rating. The scale also shows a high degree of internal
consistency across the thirteen subscales, with a Chronbach’s alpha
of .94.
The validity of the competence scale has been demonstrated in a
number of investigations. The original Timberlawn study (Lewis et al.,
1976) found that the competence scale successfully discriminated

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 135

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 135
families with hospitalized adolescents from non-clinical families.
The competence scale has also shown a high degree of construct
validity and correlated with the Self-Report Family Inventory
(r = .62), and with the general functioning subscale of the McMaster
Family Assessment Device (r=.68: Epstein et al., 1993).

Beavers Interactional Style Scale

The Beavers Interactional Style Scale is designed to assess a family’s
style which may range from centrifugal to centripetal, with
members of centrifugal families looking outside the family for their
needs to be met and members of centripetal families looking exclu-
sively within the family for need fulfilment (Beavers and Hampson,
1990). The family Style Scale is composed of the following eight
subscales.
1. Meeting dependency needs (from needs ignored to met alertly).
2. Managing conflict (from open to covert).
3. Use of space (from much space between members to very close).
4. Appearance to outsiders (from try to make a good impression to
unconcerned).
5. Professed closeness (emphasize closeness to deny closeness).
6. Managing assertion (discourage to encourage assertion).
7. Expression of positive and negative feelings (mainly positive to
mainly negative).
8. Global style (from centripetal to centrifugal).
For the eight style subscales, interrater reliabilities expressed as
Kappa coefficients ranged from .76 (Adult Conflict Scale) to .88
(Positive versus Negative Feelings), with a coefficient of .81 for the
overall Style Scale (Beavers and Hampson, 1990). The Style Scale
has good internal consistency reliability across the eight subscales
with a Chronbach’s alpha of .88. For research purposes, a style
factor consisting of two of the subscales (Social Presentation and
Balance of Positive/Negative Feelings) may be used. This factor
shows a higher predictive and clinical validity than does the overall
Style Scale (Daniels, 1995).
Validation research on the Style Scale is still in progress. Pilot
data from families assessed in a psychiatric emergency room indi-
cate that style was a significant predictor of internalizing versus
externalizing diagnoses of patients. In a diagnostically heterogen-
eous sample of patients all six unipolar depression cases were clas-

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 136

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
136 Robert Beavers and Robert Hampson
sified as centripetal and all seven substance abuse and all four
borderline personality disorder cases were classified as centrifugal
(Beavers and Hampson, 1990) .

Self-report Family Inventory

The Self-report Family Inventory (Beavers and Hampson, 1990) is a
thirty-six-item, self-report instrument which may be completed by
family members of 11 years of age and older. The Self-report Family
Inventory measures five family domains: health/competence,
conflict, cohesion, leadership, and emotional expressiveness. The
health/competence subscale includes nineteen content items
involving family affect, parental coalitions, problem-solving abilities,
autonomy and individuality, optimistic versus pessimistic views, and
acceptance of family members. The conflict subscale includes
twelve content items involving overt versus covert conflict, includ-
ing arguing, blaming, fighting openly, acceptance of personal
responsibility, unresolved conflict, and negative feeling tone. The
cohesion subscale includes five content items dealing with family
togetherness, satisfaction received from inside the family versus
outside, and spending time together. The leadership subscale
includes three content items involving parental leadership, direc-
tiveness, and degree of rigidity of control. Finally, the emotional
expressiveness subscale includes six content items dealing with
verbal and nonverbal expression of warmth, caring and closeness
(Hampson and Beavers, 1988). Respondents answer all Self-report
Family Inventory items except the last two on a Likert-type scale,
with 1 being ‘Yes: Fits our family well’; 3 being ‘Some: Fits our family
some’; and 5 being ‘No: Does not fit our family’.
The Self-report Family Inventory has high internal consistency
reliability with Cronbach alphas between .84 and .93 and test–retest
reliabilities of .85 or better. The Self-report Family Inventory also
has good validity with canonical correlations of .62 or better
between the Self-report Family Inventory Competence scores and
the observer-rated Beavers Interactional Competence Scale
(Hampson et al., 1989).
The clinical validity of the Self-report Family Inventory has
been shown by its capacity to discriminate groups of psychiatric
patients with differing diagnoses (Hampson and Beavers, 1990:
61). For example, in this study of forty-six diagnostically hetero-
geneous cases, all ten cases with schizophrenia were classified as

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 137

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 137
being within the severely dysfunctional centripetal family group-
ing and all four cases of borderline personality disorder fell into
the borderline centrifugal family grouping as predicted by the
Beavers Model.
The Self-report Family Inventory also corresponds well with
other self-report family scales measuring conceptually similar
domains. For instance, the Self-report Family Inventory
health/competence subscale correlates with the general function-
ing subscale of the Family Assessment Device (r=.77: Miller et al.,
1985) and Self-report Family Inventory. The cohesion subscale of
the Self-report Family Inventory correlates with the cohesion scale
from FACES III, a self-report scale for the Circumplex Model of
Marital and Family Functioning (r= -.67: Beavers and Hampson,
1990; Olson, 1986).

Summary of research
Our research programme has shed light on the distribution of fami-
lies in terms of the main dimensions of the Beavers Model; the rela-
tionship between the Self-report Family Inventory, the Beavers
Interactional Scales and instruments derived from other models of
family functioning; and the relationship between the dimensions of
the model and treatment process and outcome. A summary of the
results of this research follows.

Normative data
Since the 1970s we have accumulated Beavers Interactional Scales
data on over 1,800 families from both clinical and non-clinical
populations. From these data we have found that 5% of the families
studied fell into the optimal range; 38% fell into the adequate
range; 38% fell into the mid-range; 16% fell into the borderline
range; and 3% fell into the severely dysfunctional range on the
Beavers Interactional Competence Scale (Beavers and Hampson,
1993). We accept that our data are not drawn from a normative
stratified random sample, but believe that they provide an approxi-
mate indication of the distribution of families along the central
dimension of our model. These results have confirmed our expec-
tation that adequate and mid-range levels of competence are rela-
tively common, while optimal and dysfunctional levels of
functioning are relatively rare.

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 138

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
138 Robert Beavers and Robert Hampson

Cross-model, cross-method study

A cross-model, cross-method study (Hampson et al., 1995) was
conducted to compare the validity of the Beavers assessment instru-
ments with that of assessment instruments for the McMaster
(Epstein et al., 1993) and Circumplex models (Olson, 1993). In this
study parents and children over 11 years old in a group of forty-five
clinical and forty-five non-clinical families completed self-report
instruments from all three models (Beavers and Hampson, 1990;
Miller et al., 1985; Olson, 1986). Family interaction was rated by
trained raters, who achieved acceptable levels of interrater reliabil-
ity using observational ratings scales for all three models (Beavers
and Hampson, 1990; Miller et al., 1994; Olson and Killorin, 1985).
In addition, parents in all families completed the Dyadic
Adjustment Scale, a self-report measure of marital satisfaction and
adjustment (Spanier, 1976). All families comprised at least three
members, with one child being 11 years or older, and 97% of parti-
cipants were two-parent families. Clinical families were recruited
through the Southwest Family Institute, Dallas and non-clinical
families were recruited through churches and schools in the same
district.
There were four main findings in this study which support the
validity of the Beavers Systems Model’s observational and self-report
assessment instruments. First, for the competence dimension of the
Beavers Systems Model, the self-report and observational rating
scales correlated highly (r=.71). Thus, there was considerable cor-
respondence between the way in which family members and inde-
pendent raters described family competence. Second, self-reported
family competence as assessed by the Beavers Self-report Family
Inventory competence subscale correlated highly with other self-
report measures of global marital and family adjustment including
the Dyadic Adjustment Scale (r=-.44: Spanier, 1976), the general
functioning scale of the McMaster Family Assessment Device (r=.87:
Miller et al., 1985), and the cohesion scale of the Family Adaptability
and Cohesion Scales (r=-.82: Olson, 1986). This indicates that the
Beavers Self-Report Family Inventory competence subscale
measures marital and family strengths assessed by other self-report
instruments derived from other models of marital and family func-
tioning. Third, family competence as assessed by the Beavers
Interactional Competence Scale correlated highly with self-report
measures of global marital and family adjustment derived from

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 139

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 139
other models, including the Dyadic Adjustment Scale (r=-..44:
Spanier, 1976) and the general functioning scale of the McMaster
Family Assessment Device (r=.71: Miller et al., 1985). This indicates
that the Beavers Interactional Competence Scale measures marital
and family strengths assessed by self-report instruments derived
from other models of marital and family functioning. Fourth, in a
discriminant function analysis, a subset of the Beavers Interactional
Scales and Self-report Family Inventory Scales correctly classified
91% of clinical and non-clinical families. Scales that made signifi-
cant contributions to this discriminant function were observer-rated
competence, self-reported competence, self-reported cohesion and
self-reported emotional expressiveness. These results show that
scores on this list of observational and self-report scales may be
usefully employed in screening families with clinically significant
difficulties, without too many false positives or false negatives.
Taken together, the four key results from this study provide strong
support for the validity of the Beavers Family Systems Model.

Clinic family therapy studies

We examined factors associated with positive therapeutic outcome
in family therapy in two studies (Hampson and Beavers, 1996a,
1996b). The first of these two studies involved a cohort of 434 fami-
lies who sought therapy at a sliding-fee clinic in Dallas, Texas, over
an eight-year period (Hampson and Beavers, 1996b). The therapists
were interns from various universities and disciplines (psychology,
marital and family therapy, social work and psychiatry) who were all
trained in the Beavers Systems Model.
With respect to the overall outcome of treatment, 75.8% of these
families improved to some extent, with at least a few goals met,
based on ratings made by therapists in final therapy sessions. When
families who attended only one session were omitted from our
analysis, the overall improvement rate was 86.6%. The improve-
ment rate for families who attended six or more sessions was 93.8%.
The most powerful predictors of goal attainment identified in a
multiple regression analysis were: number of sessions, researcher
ratings of family competence on the Beavers Interactional Scales,
researcher ratings of family style on the Beavers Interactional
Scales, self-reported competence on the Self-report Family
Inventory, and therapists’ ratings of the degree of partnership
shared with the family in treatment. The families which achieved

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 140

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
140 Robert Beavers and Robert Hampson
the highest level of goal attainment in family therapy had the follow-
ing profile. They were more competent, with a more centripetal
style, attended at least six sessions, and formed a good therapeutic
partnership with the therapist. It was noteworthy that the following
demographic variables were unrelated to therapy outcome: family
income, family size, family structure (single versus two-parent
households), family race and therapist gender.
In the second of our two studies on therapy outcome, we exam-
ined the degree to which the outcome of therapy was related to the
match between family type (as defined by the Beavers Interactional
Scales) and therapist style in a cohort of 175 families which
attended at least three therapy sessions (Hampson and Beavers,
1996a). These families were a subset of those who participated in
the first study for whom complete datasets were available. In each
case after the third session, therapists rated their therapeutic style
in working with the family on three dimensions. These dimensions
were openness in disclosing the therapeutic strategy to the family
(from very open to guarded); power differential in the relationship
with clients (from egalitarian to maximally hierarchical); and part-
nership in the therapeutic alliance (from close and co-operative to
distant and directive).
Families rated as more competent and families which were char-
acterized by a centripetal style fared best when their therapists
were more open about their therapeutic strategy, more egalitarian
in the power differential they established with their clients, and
more joined in partnership with families within the therapeutic
alliance. Families rated as more dysfunctional and more centrifu-
gal in their style made greater therapeutic progress when their
therapists were less open about their therapeutic strategy, and
established a more hierarchical therapeutic relationship charac-
terized by interpersonal distance and directiveness. These results
confirm that different therapeutic styles are appropriate for differ-
ing types of families as defined by the Beavers Model of Family
Functioning.

Clinical implications
Family assessment and goal specification lays a solid foundation for
effective family therapy. If there are not goals which require assess-
ment, preferably goals determined by negotiation with family
members, therapy can add to, rather than subtract from family

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 141

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 141
confusion and dysfunction. The Beavers Family Systems Model was
developed to help novices in family therapy learn systems thinking
and to provide a structure to guide assessment and therapy.
The Beavers Interaction Scales and Self-report Inventory can
serve to identify high-risk families as well as evaluate the results of
clinical intervention. Our accumulated data show that most families
fall within the adequate and mid-range categories. Borderline and
severely dysfunctional families requiring clinical attention have
been found in our work to account for about 20% of all 1,800 fami-
lies we studied.
We can have a fair degree of confidence that families screened as
requiring clinical input using the Beavers Interaction Scales and
Self-report Inventory have difficulties in terms of overall function-
ality or style, because our instruments correlate well with similar
dimensions from instruments based on other models such as the
McMaster Model of Family Functioning (Epstein et al., 1993).
The results of our first family therapy outcome study show that we
can be confident that families who score higher on competence and
families who have a centripetal style will usually benefit most from
therapy. These factors and not demographic characteristics are
important determinants of a family’s capacity to benefit from family
therapy. The results of this study also highlight the importance of
taking steps to develop good partnerships with clients and keeping
them engaged in therapy, since families who form a good alliance
and stay in therapy for at least six sessions tend to make the most
therapeutic gains.
The results of our second family therapy outcome study support
the view that partnership, openness and low power differential are
the hallmarks of an effective therapeutic alliance with adequate,
mid-range and borderline families. It is only with severely dysfunc-
tional families that a clinician maximizes therapeutic effectiveness
by maintaining an overt power differential and by not disclosing
strategy. Further, centrifugal families and severely dysfunctional
families do not tend to make as much therapeutic progress as other
family groups. Setting concrete goals and holding moderate expec-
tations of therapeutic success with these families may reduce the
incidence of therapist burn-out.
Common sense would suggest that not all families require the
same interventions any more than all needful individuals respond
well to the same therapy. Our results suggest that a therapist will do
well to assess the family and be prepared to vary power differential,

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 142

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
142 Robert Beavers and Robert Hampson
disclosure of strategy and partnership with family members,
depending upon the family’s characteristics.
Clinical intervention styles developed by pioneers in family ther-
apy fall into distinct categories as determined by these three vari-
ables. Boszormenyi-Nagy’s contextual therapy (Boszormenyi-Nagy et
al., 1991), Bowen’s (1978) family-of-origin approach and social
constructionist approaches are low on power differential, high on
disclosure of strategy, and high on inviting partnership. Our results
suggest that this approach should work well with adequate and mid-
range families. Carl Whitaker, with his stories, tangential comments,
and relative warmth and openness, provides a model for working
with borderline centripetal families (Neill and Kniskern, 1982;
Roberto, 1991). Avoiding power struggles is most important in
treating these families. Jay Haley (1976, 1980, 1984) recommends a
therapeutic style that has a high power differential, secrecy about
therapeutic strategies, and a modest effort at developing a partner-
ship. Our results suggest that such an approach is well suited to
severely dysfunctional families.

Note
Copies of the Beavers Interactional Scales and the Self-report Family
Inventory are contained in W.R. Beavers and R.B. Hampson (1990)
Successful Families: Assessment and Intervention. New York: W.W. Norton.

References
Beavers, W.R. (1977) Psychotherapy and Growth: A Family Systems Perspective. New
York: Brunner/Mazel.
Beavers, W.R. (1981a) A systems model of family for family therapists. Journal of
Marital and Family Therapy, 7: 299-307.
Beavers, W.R. (1981b) Un modello sistemico di famiglia per terapisti familiari.
Terapia Familiare, 9: 42-57.
Beavers, W.R. (1982) Healthy, midrange and severely dysfunctional families. In F.
Walsh (ed.) Normal Family Processes. New York: Guilford Press.
Beavers, W.R. (1985) Successful Marriage: A Family Systems Approach to Marital
Therapy. New York: W.W. Norton.
Beavers, W.R. (1989) The Beavers Systems Model of Family Assessment. In C.N.
Ramsey (ed.) Family Systems in Medicine. New York: Guilford Press.
Beavers, W.R. and Hampson, R.B. (1990) Successful Families: Assessment and
Intervention. New York: W.W. Norton.
Beavers, W.R. and Hampson, R.B. (1993) Measuring family competence: the
Beavers Systems Model. In F. Walsh (ed.) Normal Family Processes (2nd edn). New
York: Guilford Press.

 2000 The Association for Family Therapy and Systemic Practice

JOFT 22.2 - Main Text 14/4/0 10:59 am Page 143

14676427, 2000, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1467-6427.00143 by Instituto Politecnico De Leiria, Wiley Online Library on [11/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Beavers Systems Model 143
Bertalanffy, L. von (1968) General Systems Theory. New York: Braziller.
Boszormenyi-Nagy, I., Grunebum, J. and Ulrish, D. (1991) Contextual therapy. In
A. Gurman and D. Kniskern (eds) Handbook of Family Therapy. Volume 11. New
York: Brunner/Mazel.
Bowen, M. (1978) Family Therapy in Clinical Practice. Northvale, NJ: Jason Aronson.
Daniels, L. (1995) Honing the concept of family style. Unpublished Masters Thesis.
Southern Methodist University, Dallas, TX.
Epstein, N.B., Bishop, D.S., Ryan, C., Miller, I. and Keitner, G. (1993) The
McMaster Model: view of healthy family functioning. In F. Walsh (ed.) Normal
Family Processes (2nd edn). New York: Guilford Press.
Haley, J. (1976) Problem Solving Therapy. San Francisco, CA: Jossey Bass.
Haley, J. (1980) Leaving Home. The Therapy of Disturbed Young People. New York:
McGraw Hill.
Haley, J. (1984)Ordeal Therapy. San Francisco, CA: Jossey Bass.
Hampson, R.B. and Beavers, W.R. (1988) Comparing males’ and females’ perspec-
tives through family self report. Psychiatry, 50: 24–30.
Hampson, R.B. and Beavers, W.R. (1996a) Family therapy and outcome: relation-
ships between therapist and family styles. Contemporary Family Therapy, 18:
345–370.
Hampson, R.B. and Beavers, W.R. (1996b) Measuring family therapy outcome in a
clinical setting. Family Process, 35: 347–360.
Hampson, R.B., Beavers, W.R. and Hanks, C. (1995) Family assessment: cross-
model, cross-method comparisons. Texas Psychological Association Meeting,
November 1995, Dallas, Texas.
Hampson, R.B., Beavers, W.R. and Hulgus, Y.F. (1989) Insiders’ and outsiders’
views of family: the assessment of family competence and style. Journal of Family
Psychology, 3: 118–136.
Lewis, J.M., Beavers, W.R., Gossett, J.T. and Phillips, V.A. (1976) No Single Thread:
Psychological Health in Family Systems. New York: Brunner/Mazel.
Miller, I.W., Epstein, N.B., Bishop, D.S. and Keitner, G.I. (1985) The McMaster
Family Assessment Device: reliability and validity. Journal of Marital and Family
Therapy, 11: 345–356.
Miller, I.W., Kabacoff, R.I., Epstein, N.B., Bishop, D.S., Keitner, G.I., Baldwin, L.M.
and van der Spuy, H.I.J. (1994) The development of a clinical rating scale for
the McMaster Model of Family Functioning. Family Process, 33: 53–69.
Neill, J. and Kniskern, D. (1982) From Psyche to System: The Evolving Therapy of Carl
Whitaker. New York: Guilford Press.
Olson, D. (1986) Circumplex model VII: validation studies and FACES III. Family
Process, 25: 337–351.
Olson, D.H. (1993) Circumplex Model of Marital and Family Systems: assessing
family functioning. In F. Walsh (ed.) Normal Family Processes (2nd edn). New
York: Guilford Press.
Olson, D.H. and Killorin, E. (1985) Clinical Rating Scale for the Circumplex Model of
Marital and Family Systems. St. Paul, MN: University of Minnesota, Department of
Family Social Science.
Roberto, L. (1991) Symbolic-experiential family therapy. In A. Gurman and D.
Kniskern (eds) Handbook of Family Therapy. Volume 11. New York: Brunner Mazel.
Spanier, G. (1976) Measuring dyadic adjustment: new scales for assessing the qual-
ity of marriage and similar dyads. Journal of Marriage and the Family, 38: 15–28.

 2000 The Association for Family Therapy and Systemic Practice