David H. Olson University of Minnesota

Laura Waldvogel and Molly Schlieff PREPARE/ENRICH

Circumplex Model of Marital and Family Systems:

An Update

This article provides an overview of the Cir-cumplex Model and how it was developed anddescribes the three major dimensions (cohesion,flexibility, and communication) and how theywere used to create the model. There are threemajor hypotheses and two assessments (FACESIV and the Clinical Rating Scale) used to test thehypotheses. The model is useful for describingcouple and family systems and for plotting howthey change over time. The model is also usedto describe other systems such as parenting andclassroom settings. Personal use of the model isdescribed and future directions and challengesfor the model presented.

Brief Overview of Circumplex Model

The Circumplex Model was initially designedto capture the curvilinear dimensions of cohe-sion and flexibility. These two dimensions werediscovered in the late 1970s by conceptual clus-tering more than 200 concepts in the marriageand family field (Olson, Sprenkle, & Russell,1979). Although most concepts in the fam-ily field are linear (meaning the higher thescore, the better), a basic discovery was thatcohesion and flexibility are curvilinear (veryhigh and very low are problematic). A third

Department of Family Social Science, University ofMinnesota, 977 Summit Avenue, St. Paul, MN 55105(dholson977@gmail.com).

Key Words: Circumplex Model, family systems theory, mar-riage.

dimension, communication, was also identifiedand is considered a linear and facilitating dimen-sion related to the model.

The Circumplex Model (Figure 1) has twodimensions of cohesion and flexibility, and eachdimension has five levels. The three central lev-els of cohesion and flexibility are referred to as“balanced” and the lowest and highest levels as“unbalanced.” Combining the two dimensionsorthogonal to each other resulted in nine bal-anced types (balanced on both dimensions), 12midrange types (balanced on one dimension andunbalanced on the other), and four unbalancedtypes (unbalanced on both dimensions).

There are three major hypotheses derivedfrom the Circumplex Model. First, balanced cou-ples and families tend to be more functional(happy and successful) than unbalanced sys-tems. Second, balanced couples and familieshave more positive communication than unbal-anced systems. Third, balanced couples and fam-ilies will more effectively modify their levels ofcohesion and flexibility to deal with stress anddevelopment change, as compared to unbalancedsystems.

The studies testing these hypotheses havemainly utilized the Family Adaptability & Cohe-sion Evaluation Scales (FACES) self-reportmeasure of cohesion and flexibility. A few stud-ies utilized the observational assessment calledthe Clinical Rating Scale (CRS). More than1,200 studies have been done on the CircumplexModel using the self-report family assessmentcalled FACES (Kouneski, 2000; Waldvogel

Journal of Family Theory & Review 11 (June 2019): 199–211 199DOI:10.1111/jftr.12331

200 Journal of Family Theory & Review

Figure 1. Circumplex Model of Marital and Family Systems.

& Schlieff, 2018), making it one of the mostresearched family models.

Historical Development of theCircumplex Model

Discovery of Three Dimensions

The initial discovery of the three dimensionsused with the Circumplex Model—cohesion,flexibility (initially called adaptability),and communication—was by the first author,

who was teaching a graduate seminar in maritaland family therapy. After listing more than200 concepts in the marital and family therapyfield, they clustered into the three dimensionsof cohesion, flexibility, and communication. Thethree dimensions were discovered by clusteringof concepts rather than empirical clustering.Some of the family professionals who createdthese terms are listed in Table 1.

While cohesion, flexibility, and communica-tion have been conceptually and operationally

Circumplex Model Update 201

Table 1. Theoretical models using cohesion, flexibility and communication

Cohesion Flexibility Communication

Beavers & Hampson (1990) Stylistic dimension Adaptability AffectBenjamin (1977) Affiliation InterdependenceEpstein et al. (1993) Affective involvement Behavior control

Problem solvingCommunication affective

responsivenessGottman (1994) Validation ContrastingKantor & Lehr (1975) Affect PowerLeary (1975) Affection hostility Dominance submissionLeff & Vaughn (1985) Distance Problem solvingParsons & Bales (1955) Expressive role Instrumental roleReiss (1981) Coordination ClosureWalsh & Olson (1989) Connectedness Flexibility Communication

defined in different ways (Doherty & Hovander,1990), one constant has been consensus on theimportance and value of the three dimensions.It is interesting that many family theorists haveindependently concluded that the dimensionswere critical for understanding and treating mar-ital and family systems.

Discovery of Curvilinear Dimensionsand Balance

A second discovery, that cohesion and flexibil-ity are curvilinear, occurred when the first authorlisted the concepts by various family profession-als within a given dimension. Family theoristshad different concepts for very low, moderate,and very high cohesion (Table 2).

A consistent and surprising finding by thesetheorists was that problem families tended to beeither very high or very low on cohesion. Hence,the discovery that too much or too little cohe-sion was problematic for families. Conversely,we hypothesized and found that most healthyfamilies tended to be in the middle range of cohe-sion. This led to creation of the concept of bal-ance, in that too much or too little cohesion andflexibility was problematic for couples and fam-ilies, and balancing between the extremes washealthier.

Creating Circumplex Model and ThreeDimensions

The Circumplex Model was created conceptu-ally rather than empirically. The two curvilin-ear circumplex dimensions of cohesion and flex-ibility (initially called adaptability) were puttogether in an orthogonal model. Five levels

of cohesion and flexibility were defined such thatthe very low and very high extremes were called“unbalanced,” and the three central cells, “bal-anced.” The 5 × 5 model resulted in 25 boxesin which the nine central cells were labeled “bal-anced,” 12 cells were balanced on one dimen-sion and unbalanced on the second dimension,and there were four unbalanced cells.

Although attempts were made conceptuallyto define communication in a curvilinear man-ner and to integrate communication as a thirdcircumplex dimension, it became too confusingconceptually and empirically. The decision wasmade to keep communication as a linear dimen-sion and consider it a facilitating dimension inmoving couples and families on the two otherdimensions.

Cohesion (Togetherness)

Cohesion is defined as the emotional bondingthat couple and family members have towardone another. Within the Circumplex Model,some of the specific concepts or variables thatcan be used to define and assess cohesionare emotional bonding, boundaries, coali-tions, time, space, friends, decision makingand interests, and recreation. Cohesion focuseson how systems balance separateness versustogetherness.

In the model’s balanced area, cohesive fami-lies are able to strike equilibrium between bothseparateness and togetherness. Individuals areable to be both independent from and con-nected to their families. Couples and familieswho present for therapy services often fall intoone of the extremes or unbalanced areas oftoo much separateness and/or togetherness.

202 Journal of Family Theory & Review

Table 2. Cohesion Dimension and Related Concepts

Author(s) Very Low Cohesion Balanced Cohesion Very High Cohesion

Bowen (1960) Emotional Divorce Differentiated Emotional FusionHess & Handel (1959) Separateness ConnectednessKantor & Lehr (1975) BondingReiss (1971) Distance Sensitive Environment Sensitive Consensus SensitiveRosenblatt & Titus (1976) Apartness TogethernessStierlin (1974) Centrifugal force Expelling Centripetal force BindingWynne (1958) Pseudo-hostility Mutuality Pseudo-mutuality

When cohesion levels are very high (enmeshedsystems), there is too much consensus or emo-tional closeness within the family and too littleindependence. At the other extreme (disengagedsystems), family members have a high levelof independence with limited attachment orcommitment to the family.

Extremely high levels of cohesion(enmeshed) and extremely low levels ofcohesion (disengaged) are hypothesized tobe problematic for individual and relation-ship development in the long run. In contrast,relationships with moderate scores are able tobalance being separate and together in a morefunctional way (i.e., better communication,more satisfied with the relationship).

Flexibility

Flexibility is the amount of change in its leader-ship, role relationships, and relationship rules.The specific concepts used to operationalizeflexibility include leadership (e.g., control,discipline), negotiation styles, role relation-ships, and relationship rules. Flexibility focuseson how systems balance stability with change.

Extremely high levels of flexibility (chaotic)and extremely low levels of flexibility (rigid) arehypothesized to be problematic for individualsand relationship development if they remain atthese levels for a lengthy duration. Relationshipsthat have moderate scores (structured and flexi-ble) are able to balance change and stability in amore functional way.

Communication

Communication is the third dimension in theCircumplex Model and is considered a facil-itating dimension, which means that goodcommunication helps couples and familiesalter their levels of cohesion and flexibility

to better deal with developmental or situationaldemands. Couple and family communica-tion is assessed by focusing on the groupwith regard to listening skills, speaking skills,self-disclosure, clarity, continuity tracking,and respect and regard. Listening skills includeempathy and active listening. Speaking skillsinclude speaking for oneself and not for others.Self-disclosure relates to sharing feelings aboutoneself and the relationship. Tracking refersto staying on topic, and respect and regardrefer to the affective aspects of communication.Several studies have investigated communi-cation and problem-solving skills in couplesand families and have found that systems bal-anced on cohesion and flexibility tend to havevery good communication, whereas systemsunbalanced on these dimensions tend to havepoor communication (Olson, 2000).

Hypotheses Derived From the CircumplexModel

One value of a theoretical model is thathypotheses can be deduced from the modeland tested in order to evaluate and furtherdevelop the model. These hypotheses weredescribed in the original paper on the Cir-cumplex Model (Olson, Sprenkle, & Russell,1979).

H1: Couples and families with balanced typeswill generally function more adequately acrossthe family life cycle than unbalanced types. Animportant aspect of the Circumplex Model isthe concept of balance. Individuals and fam-ily systems need to balance their separatenessversus togetherness on cohesion and their levelof stability versus change on flexibility. Eventhough a balanced family system is placed atthe four central levels of the model, these fami-lies do not always operate in a “moderate” man-ner. Being balanced means that a family system

Circumplex Model Update 203

can experience extremes on the dimension whenappropriate, as in times of trauma or stress, butthey do not typically function at these extremesfor long periods.

H1a: If a family’s expectations or culturalnorms support more extreme patterns, thefamily can function well if those unbalancedstyles fit within their cultural norms. Culturalnorms need to be seriously considered whenassessing family dynamics. What might appearto an outsider to be an “enmeshed” family,might be viewed by an insider as normativeand appropriate behavior for their family. Sounbalanced types of family systems are not nec-essarily dysfunctional, especially if the familynorms support these more extreme styles. Forexample, a specific ethnic group (e.g., Hmong,Hispanic) or religious group (e.g., Amish,Mormon) might have cultural norms that sup-port these more extreme behavior patterns oncohesion (overly connected) and flexibility(rigid).

H2: Positive communication skills will enablebalanced types of couples and families to changetheir levels of cohesion and flexibility. In gen-eral, positive communication skills are viewedas helping family systems facilitate and maintaina more balanced relationship on the two dimen-sions. Conversely, poor communication impedesmovement in unbalanced systems and increasesthe likelihood that these systems will remainextreme.

H3: Couples and families will modify theirlevels of cohesion and/or flexibility to effectivelydeal with situational stress and developmen-tal changes across the family life cycle. Thishypothesis deals with the capacity of the familysystem to change (second-order change) in orderto deal with stress or to accommodate changesin family members’ development and expecta-tions. The Circumplex Model is dynamic in thatit assumes that couples and families will changelevels of cohesion and flexibility, and thus familysystem type, and it is hypothesized that change isbeneficial to the maintenance and improvementof couple and family functioning.

When one family member’s needs or pref-erences change, the family system can eitherresist change or facilitate change. For example,if a stay-at-home dad decides to join the work-force again after his kids are in school, this maylead him to seek more independence and auton-omy from the family. The level of closenesshe had with his wife and children may shift.

The previous stay-at-home dad may prefer somechanges in other family members to deal with hisdecreased responsibility for parenting and homeduties. If his wife and children are unwilling tounderstand or assist with this desired change,the marriage and parent–child relationships maysuffer from stress and dissatisfaction.

Another common example of changingexpectations occurs when a child reachesadolescence and wants more freedom, indepen-dence, and power in the family system. Thesepressures to change the family system by onemember can facilitate change in the familydynamics or the family can resist any change,which can create more stress.

Assessment Tools: FACES and CRS

FACES is a self-report assessment for measur-ing the dimensions of cohesion and flexibility.Two major limitations of the first three versionsof FACES were that they did not tap the extremehigh or low levels of cohesion or flexibility,and they did not effectively assess curvilinear-ity. This led to the development of FACES IV,which is able to assess all aspects of the Circum-plex Model more effectively.

FACES IV

The FACES IV instrument (Olson, 2011) wasdeveloped to tap the full range of the cohesionand flexibility dimensions and is able to mea-sure the curvilinear dimensions of cohesion andflexibility. FACES IV has high levels of reli-ability and validity, and it is able to discrim-inate between healthy and unhealthy couplesand families (Olson, 2011). In creating FACESIV, work was done to develop items specifi-cally to tap the high and low extremes (unbal-anced) of the two dimensions (Tiesel, 1994).These items were then added to the moderatelyworded items of the previous versions of FACESin an attempt to develop scales that tappedthe full theoretical range of the dimensions(Gorall, 2002).

There are six scales in FACES IV with twobalanced scales (balanced cohesion and bal-anced flexibility) and four unbalanced scalesfor high and low cohesion (disengaged andenmeshed) and high and low flexibility (rigidand chaotic). These six scales provide a morecomprehensive picture of a couple and fam-ily system and provide a dimension score for

204 Journal of Family Theory & Review

plotting onto the Circumplex Model. Last, a ratioscore for cohesion and flexibility and a totalscore assesses the level of curvilinearity.

Family Communication Scale. Family commu-nication is the third dimension of the CircumplexModel. It is considered a facilitating dimensionbecause it helps couples and families changetheir level of cohesion and flexibility to dealwith ongoing issues.

Family Satisfaction Scale. Family satisfactionhas been added to the FACES IV scales as anadditional scale. This scale measures the sat-isfaction of all three dimensions of FACESIV (cohesion, flexibility, and communication)and serves as a very important outcome mea-sure. The indicators of satisfaction for cohesionwere emotional bonding, coalitions, time, space,friends, and interests. For flexibility, the indica-tors were leadership style, negotiation style, rolerelationship, and relationship rules. For commu-nication, the indicators were speaking and lis-tening skills, empathy, and negotiation skills.The final scale has been found in several studiesto have high reliability and validity.

Clinical Rating Scale

The CRS was initially developed in 1980to operationalize the dimensions of the threedimensions of the Circumplex Model. Thisobservational scale describes specific indica-tors for each level of the three dimensions.The current CRS was modified several times,and the latest version has been used in a varietyof studies (Olson, 2001). It is designed for useby therapists and researchers for rating coupleand family systems on the basis of clinicalinterviews or research observations of theirinteraction.

The CRS does tap the full continuum of thecohesion and flexibility dimensions and studiesusing the CRS have found a curvilinear rela-tionship with family functioning (Thomas &Olson, 1993, Thomas & Ozechowski, 2000).About 10 studies using the CRS have foundstrong support for the major hypothesis of theCircumplex Model, that balanced families func-tion more adequately than unbalanced families(Olson, 2011). Research has also shown thatthe scale produces the same factor structurewhen raters using the CRS scale are researchersor therapists (Lee, Jager, Whiting, & Kwantes,2000).

Validating Hypotheses Fromthe Circumplex Model

Since 1980, various researchers have validatedthe Circumplex Model on a wide varietyof topics. Most of these 525 studies have usedthe self-report scale called Family Adaptabilityand Cohesion Evaluation Scales (FACES I,II, and III), where higher scores on cohesionand flexibility on FACES represent balancedcouples and families. This means that there is alinear relationship between healthy functioningand scores on the earlier versions of FACES(Olson, 2000). Of the three hypotheses pro-posed in the original Circumplex Model, thefirst, regarding balanced and unbalanced cou-ples and families, is the most studied and mostoften supported.

In terms of topics investigated in the 525 stud-ies, the most frequent were family theory andassessment (22%), families with special prob-lems (16%), family dynamics (15%), marital andfamily therapy (12%), physical health (11%),development (10%), and sexuality (9%).

Hypothesis 1: Balanced Versus UnbalancedCouples and Families

The central hypothesis derived from the model isthat balanced couples and families function moreadequately than unbalanced couples and fami-lies. Most of the 525 studies (Kouneski, 2000;Olson, 2000) have supported this major hypothe-sis. These studies have generally compared cou-ples and families experiencing a variety of emo-tional problems and symptoms to nonclinicalfamilies.

Hypothesis 2: Balanced Couples or Familiesand Communication

Another hypothesis of the Circumplex Model isthat balanced couples and families have morepositive communication skills than unbalancedfamilies. Communication can be measuredat both the marital and the family level. In anational survey of 21,501 married couples whotook the ENRICH couple inventory, it wasfound that the happiest marriages were balancedon cohesion and flexibility and had very goodcommunication (Olson, Olson-Sigg, & Larson,2008). In a review of more than 20 studies ofcouples and families, Kouneski (2000) foundthat most of them provided strong supportfor the hypothesis that balanced couples and

Circumplex Model Update 205

families had more positive communication thanthose who were unbalanced.

Hypothesis 3: Changes in Levels of Cohesionand Flexibility to Deal With Stress

The Circumplex Model allows for the integra-tion of systems theory and family developmentaltheory. Most of the studies assessing change overtime have been done with couples and familiesin therapy were pre-test and post-test revealedthat couples changed to become more balancedat post-test (Kouneski, 2000; Olson, 2011). Inthe next section, we provide a case study of howthey changed their relationship over a few yearsto deal with developmental issues (Figure 2).

Changes in Couple and Family SystemsOver Time

The Circumplex Model is dynamic in that itassumes that changes can and do occur in thecouple and family types over time. Families canmove in any direction that the situation, stageof family life cycle, or socialization of fam-ily members may require. The model can beused to illustrate developmental change of acouple as they progress from dating to mar-riage; to pregnancy, childbirth, and child rear-ing; to raising and launching adolescents; and tomoving into life as a couple again. It can alsobe used to illustrate how a family moves throughthe model in times of high stress or managing atraumatic event.

Figure 2 illustrates the changes one youngcouple experienced in a period of 7–8 years fromdating to having their first child to when thechild was 4 years old. During their dating period(1), the couple had a very flexible and very con-nected relationship. They felt close (very con-nected) and had a very flexible style in termsof leadership and decision making. Since datingmoved them toward marriage, they have becomeincreasingly close and are trying out differentways of operating as a couple in term of flexi-bility.

During the first year of marriage (2), the new-lywed couple was best described as flexible oroverly connected. They were generally flexi-ble because they were still getting organized interms of their roles and leadership. Being in loveand enjoying spending maximum time together,they were still in the “honeymoon” phase andwere emotionally enmeshed.

By the end of their second year of marriage(3), the so-called honeymoon effect had wornoff, and the couple became somewhat flexibleor connected. Each person’s excitement with theother was not as great as it had been, and theirtogetherness became more balanced as eachgot more into his or her individual life. Theyalso developed more routines in their roles andlifestyle and became somewhat flexible.

During the third year of marriage, the couplehad a baby (4). The infant dramatically changedthe couple relationship as they became a veryflexible or somewhat connected family. Changewas high at the time, and the couple was forcedto adapt to the new challenges of parenting.Their life was in relative turmoil because theywere up each night to feed and attend to the baby.The infant’s unpredictable behavior often cre-ated chaos, and it was very difficult for the cou-ple to keep to a fixed routine; hence, they becamea very flexible family. The baby’s presence ini-tially increased the sense of bonding betweenhusband and wife, who felt united in their goalof rearing their child. As time went on, the infanttook a great deal of the mother’s time and energy,and the couple found it difficult to spend timeto stay connected as a couple. While the motherand infant were very close, the couple becamesomewhat connected.

By the time the child was 4 years old, life sta-bilized for the family (5). They are now function-ing as a flexible or connected family and experi-encing very few changes. Formerly a dual-careercouple, they shifted toward more traditional gen-der roles, with the mother staying at home, butshe returned to work part-time. While the hus-band spent little time with the infant, he has beenmore focused on his job and seeking a promo-tion. Both their closeness and flexibility havedropped a level and life has become more man-ageable for both of them.

In summary, this case study of developmen-tal changes in the family system over time illus-trates how the Circumplex Model can help mapchanges, thereby integrating system theory andfamily development theory. More specifically,this example illustrates how a couple’s rela-tionship can change from dating and across theearly stages of marriage. The changes can occurgradually over months or more rapidly afterthe birth of a child. These changes often occurwithout specific planning. However, couples cannegotiate the type of relationship they want andcan be more proactive in creating the type of

206 Journal of Family Theory & Review

Figure 2. Life-Cycle Changes.

relationship they both prefer. These changes ina couple or family system are a snapshot of thechanges that occur in couple or family levelsof cohesion and flexibility over the family lifecycle.

Five Parenting Styles Based on theCircumplex Model

One of the values of a theoretical model isto make conceptual comparisons to other related

models. Diana Baumrind (1995) identifiedfour styles of parenting: democratic (authori-tative), authoritarian, permissive, and rejecting(Figure 3). She also identified specific out-comes in children’s behavior for each of thesestyles.

Using the Circumplex Model, the compari-son revealed the same four styles as describedby Baumrind (1995), as well as a new fifth stylecalled “uninvolved,” not included in her fourstyles. The five styles as defined and used in the

Circumplex Model Update 207

Figure 3. Five Parenting Styles.

parenting version of the Circumplex Model: bal-anced, permissive, overbearing, strict, and unin-volved.

The balanced style of parenting, oftencalled “democratic,” is what Baumrind (1995)called the authoritative style. This style hasage-appropriate parenting and independenceis encouraged. Discipline tends to be consis-tent and fair. Parenting is warm and nurturingwithout being overindulgent. The permissive

style, the same name as used by Baumrind,allows children a wide range of freedom andchoice. Parents have a hard time saying no,creating boundaries, and enforcing rules. Thereis a high emotional connection to the child andresponsiveness to the child’s needs. This style isoften related to children becoming demandingand self-focused. The overbearing style, alsocalled authoritarian by Baumrind and others,is the opposite of permissive. There are high

208 Journal of Family Theory & Review

levels of parental control and high expectations.Children tend to have more anxiety, lowerself-esteem, and lower achievement as com-pared to children raised with other parentingstyles. The strict style, which Baumrind called“rejecting,” demands order but with little emo-tional connection. Children often feel uncaredfor and tend to have higher levels of rebellionand substance abuse. In the uninvolved style,which Baumrind did not include, children aregiven a great deal of freedom of choice withfew rules or boundaries. There is low emo-tional connection to and demands placed onthe child. Children experiencing this style oftenfeel isolated and uncared for by their parents.In summary, these five styles of parentingderived from the Circumplex Model reveala similarity to other parenting theorists likeBaumrind.

Other Applications of the CircumplexModel

Classroom Styles

One of the values of a theoretical model is thatit can potentially be applied in other settings.In addition to use of the Circumplex Modelto represent five parenting styles, it has alsobeen used to describe classroom settings. MarianFish and Elizabeth Dane (2000) used the Clin-ical Rating Scale from the Circumplex Modelto develop the Classroom Systems ObservationScale. Their assessment focused on classroomcohesion, flexibility, and communication. Theirstudies replicated the findings from couples andfamilies that balanced classrooms had more pos-itive outcomes for children.

Leadership Styles

The application of the Circumplex Modelto leadership styles of business teams was doneby a group headed by Robert Watson (2000),which developed leadership styles by revisingFACES items to apply to a work setting. Thefive leadership styles they found were balanced,permissive, micromanaging, controlling, anduninvolved. Their exploratory work found thatthe leadership styles they found were very simi-lar to what was discovered with the CircumplexModel for families. Also, balanced work groupswere found to be more productive and satisfiedwith work.

Recent Studies Using FACES IV

Kouneski’s (2000) review of research on the Cir-cumplex Model found that most studies usedthe self-report instruments FACES II and III.Both provided a linear measure of cohesion andflexibility so that a high score indicated themost balanced and a low score the least bal-anced. Since Kouneski’s (2000) article on theCircumplex Model using FACES IV, at least50 more studies and applications of the modelhave been published; these are summarized byLaura Waldvogel and Molly Schlieff (2018).Their review found that the studies were almostequally spread across the following categories:types of families, physical health, family coun-seling, stages of family life cycle, ethnicity, fam-ily theory, and families with special problems.

Several novel types of families have beenstudied with the use of FACES IV, and oneof these studies captured the era of today:millennial families. Rebecca Ristow (2015)studied millennial families and their percep-tions on cohesion, flexibility, satisfaction, andcommunication. The families described them-selves as having balanced levels of cohesionand flexibility (Ristow, 2015). Families alsodescribed healthy boundaries in their familialunits. Because of their balanced flexibility, thefamilies showed signs of egalitarian leadershipand a democratic approach to decision making.Roles were shared, and there was fluid changewhen necessary in their family.

The Circumplex Model has also been uti-lized to assess families dealing with physicalhealth difficulties. Pereira and Teixeira (2013)found FACES IV to be valid when assessingPortuguese families with parents having cancer.Their study found that in a parental cancersituation, adult children caregivers with lowereducational levels seemed to have more prob-lems dealing with unbalanced systems (i.e.,enmeshed and chaotic). However, adult childrencaregivers and their parents with higher educa-tion levels were shown to be more balanced thanthose with lower education.

The Circumplex Model can be used assessdifferent family types related to individualdevelopment. A study by Everri, Mancini,and Fruggeri (2016) found six family typesdescribed from the viewpoint of adolescents:rigidly balanced, flexibly oscillating, flexiblychaotic, cohesively disorganized, rigidly dis-engaged, and chaotically disengaged. Theyfound that the families did not differ in terms of

Circumplex Model Update 209

adolescent age, gender, and family structure, butthey did differ in the extent to which the variousstyles had different behavioral outcomes.

There are several international studies thathave studied the family in different countriesof varying ethnicity, education, and religion.Three international studies validated FACES IVwith specific ethnic groups, including Hungar-ian, Romanian, and Greek families. The studyby Mirnics, Vargha, Toth, and Bagdy (2010)included 249 Hungarian couples and found sim-ilar family types as found in earlier CircumplexModel studies. Studying 1,359 young Roma-nian family members, Cornelia Rada (2017)found similar family profiles to those in theresearch in the United States. The Greek versionof FACES IV was found to be valid and reliableand useful in understanding underlying familydynamics in Greece (Koutra, Triliva, Roumelio-taki, Lionis, & Vgontzas, 2012). These studiesand many more are described on the FACESIV website (www.facesiv.com) under “Researchstudies.”

For international studies using FACES IV,it is highly recommended that they obtain alarge-enough database (several hundred respon-dents) so that they can renorm cohesion and flex-ibility. This renorming will change the centroidof the Circumplex Model, thereby making thefindings more accurate for that culture. Moredetails on this process is described in the FACESIV Manual, available on the FACES IV website.

Immigrant families have also been studiedusing FACES IV to assess level of cohesionin relationship to trauma. The study by Singh,Lundy, Vidal de Haymes, and Caridad (2011)looked at 122 Mexican immigrant men andwomen living in a Mexican community in theMidwestern United States. This study identi-fied the positive role the family can play intrauma prevention for immigrant families (Singhet al., 2011). Immigration created movement andadjustment within the family system, and it alsoappeared to create greater cohesion in the fami-lies.

Religion was also used as a specific societalcontext for understanding families and theirdynamics. Michael Messina (2008) researchedthe relationship among nine religion variablesand family variables assessed by FACES IV.Among 152 Christian participants, Messina(2008) found that seven of the religiosity scaleswere positively related to family cohesion,five of those seven were positively related to

communication, and none was related to familyflexibility.

Several studies were done with families expe-riencing a wide assortment of challenges, fromalcoholism to eating disorders. One study byTafa et al. (2016) measured family functioningin families with female adolescents who haveeating disorders. Their study found that adoles-cents and their parents differ in their perceptionof family functioning. More specifically, ado-lescents with anorexia perceived their family ashighly disengaged, rigid, and with poor commu-nication, whereas parents tended to describe thefamily as more balanced on cohesion and flexi-bility with good communication.

Personal Use of Circumplex Model

There are a variety of ways that the CircumplexModel can help a person or couple better under-stand and improve their relationship. One optionis to purchase FACES IV and take the assess-ment, score it, and plot the results (www.facesiv.com). The simplified version of the Circum-plex Model is called the Couple and FamilyMap, and it is built into the PREPARE/ENRICHcouple assessment and program (www.prepare-enrich.com) or the Couple Checkup book andprogram (www.couplecheckup.com). Once theassessment is complete, results can reveal howeach person perceives the relationship.

The model can also be used to demonstratehow a relationship has changed over time. Anassumption of the Circumplex Model is that rela-tionships change, and plotting key events in therelationship (e.g., birth of a child, car accident,illness, loss of job) often reveals elasticity orresiliency in the relationship. The CircumplexModel can also reveal where a couple or familyis in terms of cohesion and flexibility and helpthem think about where they would like to be inthe future (i.e., setting goals).

Theoretical Contributions of theCircumplex Model

The Circumplex Model provides a varietyof benefits to family professionals (Olsonet al., 2008). First, the model identifies threesignificant dimensions that conceptually sum-marize many family concepts. Second, thetwo dimensions of cohesion and flexibilitywere hypothesized to be curvilinear, not linear.The curvilinear discovery led to the important

210 Journal of Family Theory & Review

concept of balance. Putting the five levelsof cohesion and flexibility together created adescriptive model identifying 25 types of coupleand family systems. It is also a dynamic model,able to illustrate change in systems over time.

The Circumplex Model also helped to createtestable hypotheses. To test them, assess-ment scales were developed on the basis ofthe conceptual and operational definitions ofthe concepts. This resulted in an improvedself-report assessment, FACES IV, and an obser-vational assessment, the Clinical Rating Scale(CRS).

The Circumplex Model, its historical roots,basic concepts, and dimensions are grounded insystems theory. The Circumplex Model’s dimen-sions are systemic and thus have been appliedto understanding couple and family systems,as well as parenting styles, classrooms, workgroups, and other ongoing systems.

The model has also been used with diversecouple and family systems in terms of ethnic-ity, race, marital status (married), living arrange-ment (cohabitating), family structure (single par-ent, stepfamilies), sexual orientation (gay andlesbian couples), stage of family life cycle (par-enting to empty nest), and social class and educa-tional levels. Changes that couples and familiesexperience developmentally and in reaction tostressors can also be illustrated using the model.

Future Directions

Many of the studies from the years 2000–2018have been with families from other coun-tries. Many of these international studies havefocused on translating and validating FACESIV and determining the value of the assessmentfor their specific setting. These internationalstudies have not focused as much on testingthe three major hypotheses as on assess-ing the reliability and validity of FACES IVin an international context. Hopefully, futureinternational studies can focus more on test-ing the hypotheses with a variety of familyproblems.

A limitation of many of the studies in theUnited States is that they are conducted mainlywith Caucasian, Christian, middle-class fami-lies. Future studies should focus on familiesfrom a wider variety of ethnic groups, religiousorientations, and social classes. Future studiesneed to also include more family members ratherthan just studying one person in a relationship,

which will reveal the complexity of couple andfamily systems.

References

Baumrind, D. (1995). Child maltreatment and opti-mal caregiving in social context. New York, NY:Garland.

Beavers, W. B., & Hampson, R. B. (1990). Successfulfamilies: Assessment and intervention, New York,NY: Norton.

Benjamin, L. S. (1977). Structural analysis of a fam-ily in therapy. Journal of Counseling Clinical Psy-chology, 45, 391–406.

Bowen, M. (1960). The family as the unit of study andtreatment. American Journal of Orthopsychiatry,31, 40–60.

Doherty, W., & Hovander, D. (1990). Why don’tmeasures of cohesion and control behave the waythey’re supposed to? American Journal of FamilyTherapy, 18, 5–18.

Epstein, N. B., & Bishop, D. S. (1993). The McMas-ter Assessment Device (FAD). In F. Walsh (Ed.),Normal family processes. New York, NY: GuilfordPress.

Everri, M., Mancini, T., & Fruggeri, L. (2016). Therole of rigidity in adaptive and maladaptive fami-lies assessed by FACES IV: The points of view ofadolescents. Journal of Child and Family Studies,25, 2987–2997.

Fish, M. C., & Dane, E. (2000). The classroom sys-tems observation scale. Learning EnvironmentsResearch, 3, 67–92.

Gorall, D. (2002). FACES IV and Circumplex Model.Unpublished doctoral dissertation. St. Paul, MN:Family Social Science, University of Minnesota.

Gottman, J. M. (1994). Why marriages succeed or fail.New York, NY: Simon & Schuster.

Hess, R., & Handel. G. (1959). Family worlds: Apsychological approach to family life. Chicago, IL:University of Chicago Press.

Kantor, D., & Lehr, W. (1975). Inside the family. SanFrancisco, CA: Jossey-Bass.

Kouneski, E. (2000). Circumplex model andFACES: Review of literature. Unpublishedmanuscript. Roseville, MN: PREPARE/ENRICH.Retrieved from http://www.lifeinnovations.com/familyinventoriesdatabase.html

Koutra, K., Triliva, S., Roumeliotaki, T., Lionis, C.,& Vgontzas, A. (2012). Cross-cultural adaptionand validation of the Greek version of the fam-ily adaptability and cohesion evaluation scales IV(FACES IV Package). Journal of Family Issues,34(12), 1647–1672.

Leary, T. (1975). Interpersonal diagnosis of person-ality. New York, NY: Ronald Press.

Lee, R. E., Jager, K. B., Whiting, J. B., & Kwantes,C. T. (2000). Clinical assessment using the Clinical

Circumplex Model Update 211

Rating Scale: Thomas and Olson revisited. Journalof Marital & Family Therapy, 26, 535–537.

Leff, J., & Vaughn, C. (1985). Expressed emotion infamilies. New York, NY: Guilford Press.

Messina, M. (2008). The relationship between reli-giosity and the family dynamics of the CircumplexModel of marital and family systems. Unpublisheddoctoral dissertation. Los Angeles, CA: AlliantInternational University.

Mirnics, Z., Vargha, A., Toth, M., & Bagdy, E. (2010).Cross-cultural applicability of FACES IV. Journalof Family Psychotherapy, 21, 17–33.

Olson, D. H. (2000). Circumplex model of marital andfamily systems. Journal of Family Therapy, 22(2),144–167.

Olson, D. H. (2001). Clinical Rating Scale for Cir-cumplex Model. St. Paul, MN: Family Social Sci-ence, University of Minnesota.

Olson, D. H. (2011). FACES IV and the CircumplexModel: Validity study. Journal of Marital & Fam-ily Therapy, 3(1), 64–80.

Olson, D. H., & Olson, A. K. (1999). PRE-PARE/ENRICH Program: Version 2000. InR. Berger & M. T. Hannah (Eds.), Preventiveapproaches in couple therapy (pp. 196–216).Philadelphia, PA: Brunner/Mazel.

Olson, D. H., & Olson, A. K. (2000). Empoweringcouples: Building on your strengths. Roseville,MN: PREPARE/ENRICH.

Olson, D. H., Olson-Sigg, A., & Larson, P. J. (2008).The couple checkup. Nashville, TN: Thomas Nel-son.

Olson, D. H., Russell, C. S., & Sprenkle, D. H. (1989).Circumplex model: Systemic assessment and treat-ment of families. New York, NY: Haworth Press.

Olson, D. H., Sprenkle, D. H., & Russell, C. S. (1979).Circumplex model of marital & family systemsI: Cohesion & adaptability dimensions, familytypes, & clinical applications. Family Process, 18,3–28.

Parson, T., & Bales, R. F. (1955). Family socializationand interaction process. Glencoe, IL: Free Press.

Pereira, M. G., & Teixeira, R. (2013). Portuguesevalidation of FACES IV in adult children care-givers facing parental cancer. Contemporary Fam-ily Therapy, 35, 478–490.

Rada, C. (2017). Latent class analysis approach for thefamily adaptability and cohesion evaluation scale

IV among young people from Romania: The firststep for validation. Journal of Family Issues, 39(6),1598–1615.

Reiss, D. (1981). The family’s construction of reality.Cambridge, MA: Harvard University Press.

Ristow, R. (2015). The Circumplex Model and the mil-lennial family system. Unpublished master’s thesis.Omaha, NE: University of Nebraska.

Rosenblatt, P., & Titus, S. (1976). Together and apartin family. Humanitas, 12, 367–379.

Singh, S., Lundy, M., Vidal de Haymes, M., &Caridad, A. (2011). Mexican immigrant families:Relating trauma and family cohesion. Journal ofPoverty, 15, 427–443.

Stierlin, H. (1974). Separating parents and adoles-cents. New York, NY: Quadrangle.

Tafa, M., Cimino, S., Ballarotto, G., Bracaglia, F.,Bottone, C., & Cerniglia, L. (2017). Femaleadolescents with eating disorders, parental psy-chopathological risk and family functioning.Journal of Child & Family Studies, 26, 28–39.

Thomas, V., & Olson, D. H. (1993). Problem familiesand the Circumplex Model: Observational assess-ment using the Clinical Rating Scale. Journal ofMarital & Family Therapy, 19, 159–175.

Thomas, V., & Ozechowski, T. J. (2000). A test of theCircumplex Model of marital and family systemsusing the Clinical Rating Scale. Journal of Marital& Family Therapy, 26, 523–534.

Tiesel, J. (1994). FACES IV: Reliability and Validity.Unpublished doctoral dissertation. St. Paul, MN:Family Social Science, University of Minnesota.

Waldvogel, L., & Schlieff, M. (2018). Updated Cir-cumplex Model studies, 2000–2018. Unpublishedmanuscript. Roseville, MN: PREPARE/ENRICH.

Walsh, F., & Olson, D. H. (1989). Utility of the Cir-cumplex Model with severely dysfunctional fam-ily systems. In D. H. Olson, C. S. Russell, &D. H. Sprenkle (Eds.), Circumplex model: Systemicassessment and treatment of families (2nd ed., pp.51–78). New York, NY: Haworth.

Watson, R. (2000). Team organizational profile (TOP)leadership styles. Unpublished manuscript. Den-ver, CO: United Launch Alliance.

Wynne, L. (1958). Pseudo-mutuality in the fam-ily relations of schizophrenia. Psychiatry, 21,205–222.