The Impact of Family of Origin Satisfaction

on Adult Attachment Formation

Elizabeth C. Hunter


The present study explored the relationship between adult attachment styles and level of satisfaction with family of origin functioning.  To test this relationship, one hundred McKendree College students from the Social Science Department completed a survey.  The survey consisted of two established psychological tests.  The first was the Family APGAR, a five-question scale that measures level of satisfaction with family of origin functioning (Smilkstein, 1992).  The second was the Self-Report Measure of Adult Attachment Style, which used scenario selection to test four styles of attachments:  secure, dismissing, preoccupied, and fearful (Bartholemew & Harowitz, 1991).  Findings indicated that people with low levels of satisfaction with family functioning were more likely to develop fearful attachment styles in adult relationships (r = -.264, p = .009).  In contrast, people with high levels of family of origin satisfaction were more likely to form secure attachments in adult relationships (r = .467, p = .001).  This study was unique in that it looked at the root of attachment on a subjective level.  Most studies of attachment suggest the importance of actual events in facilitating adult attachment style.  However, this study reveals that actual events within the family context do not have as significant an impact on attachment as one’s interpretation of those events.


      Various researchers and theorists suggest the importance of family in facilitating attachment patterns throughout the life course.  In fact, early theorists studying infant attachment focused almost exclusively on the role early childhood plays in the development of one’s attachment schema.  It has only been in the last couple of decades, that researchers have begun to speculate about the continuous nature of attachment formation.  More recent theorists implicate the importance of family in the development of attachment patterns, but they also suggest that other relationships can either reinforce or alter these existing patterns.  Despite this modification of attachment theory, most theorists still agree that it is difficult to reverse the patterns of attachment set forth in early parent-child relationships.   

     Hazen and Shaver (1994b) suggest the critical nature or the primary, or first, bond of life.  They contend that this bond will influence all subsequent bonds by setting up a self-confirming expectation about the availability of others.  In other words, individuals will begin to look for things in their present relationships that confirm either the acceptance or rejection received in their early interaction with parental figures.  In support of this assertion, one study found that adults with a secure attachment style described both their family of origin and their current family more positively than those with an insecure attachment style, in regards to family cohesion, warmth, and conflict resolution (Diehl, Elnick, Bourbeau, & Labouvie, 1998). 

     Given previous research, it is obvious that family does play a critical role in the development of attachment styles.  Since family is critical, the next question becomes, “Which aspects of family bonds are most critical in affecting attachment formation throughout the life course?”  The answer to this question has evolved over time.  Early theorists, particularly Bowlby, suggested that the actual events of infancy, such as the amount of time a child is held, are the most important determinants of later attachment patterns.  More recent theorists, however, have begun to argue that the individual’s self-perception of their attachment relationships in early years is more critical than the occurrence of actual childhood attachment related events. 

     As a result of this shift, measurements have shifted in recent years from more objective observational strategies, as those employed by Bowlby and Ainsworth, to more subjective measurement devices.  Self-report techniques have become increasingly popular for several reasons.  One reason may be the ease with which they allow adult attachment patterns to be assessed.  Longitudinal observational studies are extremely difficult to complete, and relying solely on such methods could inhibit us from understanding a variety of social problems that deserve immediate attention. 

     In addition to ease of assessment, self-report studies appear to have become more intrinsically valuable in recent years.  Attachment theorists are shifting away from exclusive focus on early childhood, as they become increasingly aware of the continuous nature of attachment formation over the life course.  Assessing attachment in infancy would obviously be better understood through observational studies of actual events, due to the underdeveloped nature of an infant’s interpretive cognitions.  However, as interpretive processes build and grow, is seems logical to conclude that the critical dimension of attachment becomes less about actual events and more about one’s interpretation of those events.  Whereas during infancy we respond to whether or not our needs are being met, as we grow we begin to respond to our interpretations of why those needs are or are not being met.

     Given all of this, it is obvious that more research needs to be directed towards understanding the interpretive processes of attachment formation, as opposed to the concrete evaluation of actual events.  This is especially true when looking at adult attachments.  The present study will seek to provide such an analysis, by utilizing the Self-Report Measure of Adult Attachment Style set forth by Bartholomew and Horowitz (1991).  In order to understand the importance of utilizing a self-report measure of adult attachment, it is important to first understand the evolution of attachment theory.

Review of the Literature

     Attachment theoretical models have consistently drawn upon the work of John Bowlby.  Bowlby (1977) defines attachment as “the propensity of human beings to make strong affectional bonds to particular others.”  According to Bowlby, these bonds develop as an evolutionary demand for infants to maintain close proximity to their caregivers during times of threat. 

     Anything that threatens to inhibit proximity evokes anxiety, which causes the infant to engage in whatever behavior is necessary to reestablish the desired closeness (Hazen & Shaver, 1994a).  These behaviors become the basis for future attachment patterns.  Thus, Bowlby not only views attachment as a behavioral system, but as the preeminent system in a set of behavioral patterns.  Similar to the hierarchy of human development proposed by Maslow, Bowlby believes that the activation of the attachment system precedes and either inhibits or enhances the development of other behavioral systems, primarily that of exploration.  Exploration is a necessary step towards feeling competent in mastering the social world, but in order for infants to proceed to this step, they must feel they have a “secure base” to retreat to in the event of danger.  The perception of whether or not this “secure base” is available can either inhibit or help the development of infants in their attachment systems, as well as in a broad range of other systems (Hazen & Shaver, 1990,1994a).

     After working extensively with institutionalized children, Bowlby concluded that the first three years of life are the most critical in the process of attachment formation (Hazen & Shaver, 1994a).  These years serve as the foundation for the development of one’s “internal working model of attachment” (Hazen & Shaver, 1994a).     Bowlby (1973) uses the term “internal working model” to refer to representations of self, others, and relationships, which develop over the course of infancy and provide the basis for continuity in attachment patterns over the life course.  Bowlby (1973) considers these internal working models to be fairly accurate depictions of actual childhood events.

     According to Bowlby, working models develop on the basis of two underlying dimensions (Bartholomew, 1997).  The first dimension refers to the child’s representation of others and assesses whether or not the primary attachment figure can be depended on for protection and support.  The second dimension refers to the child’s representation of self and determines whether or not the self is the type of person to whom anyone, particularly the attachment figure, is likely to react to in a positive manner.  Based on the messages received in these two systems, the infant begins to develop a particular attachment style. 

     This attachment style tends to become more stable as the individual grows, unless other disconfirming primary bonds are formed in childhood or adolescence (Kesner & McKenry, 1998).  As the individual’s internal working model becomes more stable, it begins to integrate into the person’s personality and determine the success or failure of his/her subsequent relationships (Kesner & McKenry, 1998).  Thus, the individual begins to seek out or try to reconstruct relationships that confirm their own “internal working model of attachment” (Hazen & Shaver, 1994a). 

     The next leading theorist in the development of childhood attachment theory was Mary Ainsworth.  Ainsworth built upon Bowlby’s models of self and other but developed a more specific interpretation of these models (Griffin & Bartholomew, 1994).  Whereas Bowlby conceptualized these models as being characterized by global expectations/evaluations, Ainsworth felt that attachment formation was characterized by more specific strategies of bonding, meaning that each pattern is associated with a specific set of interpersonal problems. 

     In addition, under Ainsworth attachment theory began to take a more subjective stance of human need fulfillment, suggesting that attachment systems evolved as a means of maintaining “felt security” as opposed to simply maintaining close physical proximity (Bartholomew & Horowitz, 1991). 

     Hazen and Shaver (1994a) posed the following questions in regards to Ainsworth’s notion of “felt security”:  “Can I count on my attachment figure to be available and responsive when needed?”  They go on to suggest that there are three possible answers to this question, which are yes, no, or maybe. 

     The answers listed above form the basis of Ainsworth’s three specific patterns of infant attachment formation, which includes secure, anxious resistant, and avoidant (Bartholomew & Horowitz, 1991).  Secure attachments are formed when a caregivers is consistently responsive to the child’s needs (Hazen & Shaver, 1994a).  Anxious-resistant patterns develop when a caregiver is sometimes unresponsive and at other times overly intrusive.  Avoidant attachments are the result of caregiving that is considerable unresponsive. 

    In order to test these attachment patterns and their link to infant exploration, Ainsworth developed a method of assessment called the Strange Situation (Hazen & Shaver, 1994a).  Under the Strange Situation, infants were separated form their caregivers and placed in an unfamiliar environment in order to activate the child’s attachment system.  The three critical components of this assessment include:  (a) proximity maintenance, which refers to the extent to which the infants sought proximity with their caregivers  (b) safe haven, which is the degree to which infants can be comforted by contact with the caregiver  (c) secure base behaviors, which involve exploratory behaviors that take place in the caregivers presence, such as playing with toys. 

            Findings revealed significant differences in infants based on attachment patterns in regards to these critical dimensions.  Ainsworth found that securely attached infants are distressed when their mother is absent, are easily comforted upon her return, and will engage in exploration as long as she is present (Hazen & Shaver, 1994a; Bartholomew & Horowitz, 1991).  Anxious-resistant children show ambivalent behavior in the presence of their mother (anxious, angry), are resistant to comfort upon her return, and are so preoccupied with their mother when present that exploration is inhibited (Hazen & Shaver, 1994a; Bartholomew & Horowitz, 1991).  Avoidant infants, on the other hand, show no signs of distress upon separation, avoided contact with their caregivers upon return, and direct most of their attention towards exploration (Hazen & Shaver, 1994a).  From these studies, it has been postulated that 60 percent of American samples have a secure attachment pattern, 15 percent have an anxious-resistant pattern, and 25 percent have an avoidant pattern of attachment (Hazen & Shaver, 1994a). 

            In recent years, attachment theorists have become increasingly aware of the continuous nature of attachment formation and have broadened their scope beyond the theories of childhood attachment postulated by Ainsworth and Bowlby.  In her work assessing child-rearing practices, Mary Main set forth three adult patterns of attachment, which she measured using her own assessment tool, the Adult Attachment Interview (Bartholomew, 1997).  A critical component of Main’s assessment is that it represented a shift away from the long-held notion that attachment patterns are rooted in actual events.  In her assessment, the content of descriptions of childhood were less important than the way in which individuals discussed their childhood experiences. 

     Around the same time period, Cindy Hazen and Phillip Shaver began constructing a theory of adult attachment, based on the three types set forth by Ainsworth.  They developed a simple self-report measure of attachment to assess adult love relationships (Volling, Notaro, Larsen, 1998).  According to Volling, Notaro, and Larson’s assessment of Hazen and Shaver, they believed that adult romantic love involved a bond similar to that which existed between infants and their caregivers.  They upheld that the three types of attachment set forth by Ainsworth, regarding infancy, were parallel to the patterns of attachment found in adulthood (Volling, Notaro, & Larsen, 1998).

         Under Hazen and Shaver, the secure adult was found to be more comfortable with intimacy and dependence on others, to have longer relationships, and to characterize the relationships as friendly, trusting, and happy (Brennan & Shaver; Volling, Notaro, & Larsen, 1998; Kesner & McKenry, 1998).  They are also more outgoing, report less loneliness, and have higher self-esteem than individuals with an anxious-ambivalent attachment patterns (Diehl, Elnick, Bourbeau, and Labouvie, 1998).  At work, they generally feel appreciated and do not worry about failure (Hazen & Shaver, 1990).  They maintain a healthy balance between work and friendships and do not have difficulty engaging in leisure activities or taking vacations.  Secure individuals also uphold the importance of love and relationships over work (Hazen & Shaver, 1990).  This is the most common pattern of attachment, comprising 55 percent of adults (Hazen & Shaver, 1994a). 

     The anxious-ambivalent adult tended to be characterized by a lack of confidence in the availability of others, causing them to be overly preoccupied with attachment issues (Brennan & Shaver 1998; Hazen & Shaver, 1994a).  They constantly seek to be closer than their partner will also and experience intense jealousy (Hazen & Shaver, 1994a).  This group tends to fall in love easily, have low self-esteem, and have difficulty maintaining a long-term relationship.  They also engage in inappropriate and excessive self-disclosure, in which they assert the importance of their own needs over their partner’s needs.  Their preoccupation with relationship issues causes their achievement potential to be lower than the other two groups.  At work, they are typically preoccupied, poorly organized, and undermotivated (Hazen & Shaver, 1994a).  They will often work to get praise but slack off once it is received (Hazen & Shaver, 1990).  This is the least common adult attachment style, comprising 20 percent of the adult population (Hazen & Shaver, 1994a).

     Avoidant adults also experience an array of interpersonal problems, but they are qualitatively different than those of the anxious-ambivalent.  While anxious-ambivalent people have an intense desire for closeness, avoidants have a fear of intimacy and closeness in relationships (Volling, Notaro, & Larsen, 1998; Kesner & McKenry, 1998; Hazen & Shaver, 1994a).  They also regard themselves as being excessively self-reliant (Brennan & Shaver, 1998, Hazen & Shaver, 1994a).  Others regard them as cold and hostile (Hazen & Shaver, 1994a).  In relationships, they have difficulty engaging in self-disclosure and maintaining long-term commitments (Hazen & Shaver, 1994a).  Kesner and McKenry (1998) also characterized them as being extremely jealous, although this is inconsistent with other research, which suggests anxious-ambivalents are the jealous type.  They are also more likely to use alcohol or illegal substances as a coping mechanism and to engage in uncommitted sexual activity (Hazen & Shaver, 1994a).  At work, they tend to be overly involved in order to avoid social interaction and they have trouble engaging in leisure activities (Hazen & Shaver, 1990, 1994a).  Avoidants will claim that work provides them with more personal fulfillment than love or relationships (Hazen & Shaver, 1990).  This is the second most common group of adults, comprising 25 percent of the adult population (Hazen & Shaver, 1994a).

     In addition to extending the three attachment types to explain adult relationships, Hazen and Shaver (1994a) reconceptualized attachment as a fluid process of relating to others that is affected and shaped by each new relationship.  This process is most pliable early in life and becomes more difficult to alter as time goes on.  Because of the fluidity of attachment, especially during the early years of formation, Hazen and Shaver propose that all it takes is one relationship experience in early childhood that disconfirms insecure attachment patterns to allow people to form secure attachments in adulthood.  Such disconfirming experiences often come from a teacher or another relative in childhood or from a romantic partner in adolescence/early adulthood.  Hazen and Shaver (1994a) also contend that the reverse pattern (secure-insecure) is less likely to occur, citing that numerous studies have shown that secure attachments are the most stable pattern. 

     However, Kim Bartholomew (1994) challenges this assertion.  She feels that the plethora of research, which upholds that change is more likely to occur in the direction of stability, may be overlooking the propensity of such a pattern to occur by chance alone.  Since more people have secure than insecure attachments, it is by chance more likely that more people will maintain secure attachments.  She feels that Hazen and Shaver may be acting prematurely by asserting that change occurs predominantly in only one direction (Bartholomew, 1994). 

     Despite this criticism, Hazen and Shaver have contributed much to our understanding of the changing nature of attachment over the life-course.  They have specified several fundamental differences between childhood and adult attachment bonds.  One difference specified is that childhood attachments are complimentary, whereas adult attachments are reciprocal.  Infants expect parents to provide care and parents understand that it is their role to provide, rather than seek care, from these infants.  Adult relationships, on the other hand, call for both partners to be providers and receivers of care.  In addition, attachment patterns shift over the life course from the childhood goal of maintaining physical proximity to the adult goal of maintaining emotional proximity in the form of “felt security.”  As this change occurs and peers begin to fulfill the needs once satisfied by parental figures, attachments are transferred one by one (Hazen & Shaver, 1994a).  Eventually, parental figures become secondary on one’s hierarchy of attachment figures, although their importance is never completely relinquished in this process.  This process, developed by Hazen and Shaver, explains why it is critical to assess perceptions/interpretations rather than actual events when assessing adult attachment.  Adults have already developed a schema of the world, which inhibits them from making completely objective evaluations about the occurrence of actual events.  They assess events based on messages received in the past and their individual interpretative processes.

     The present study will utilize a more recent self-assessment of adult attachment called the Self-Report Measure of Adult Attachment Style (Bartholomew & Horowitz, 1991).  This measurement device, developed by Kim Bartholomew and Leonard Horowitz (1991), is an extension of Hazen and Shaver’s earlier model of adult attachment.  Unlike Hazen and Shaver, Bartholomew and Horowitz (1991) conceptualize four possible patterns of attachment by exhausting Bowlby’s original dimensions of self and other.  Whereas Hazen and Shaver presented only the three types of attachment based on parenting styles, Bartholomew and Horowitz (1991) propose that there are four ideal types possible when one considers the importance of self-perception, as well as the importance of parenting styles.  Recall that the two critical dimensions of Bowlby’s theory are (a) the likelihood that the attachment figure will respond to calls for protection and support  (b) whether or not one view themselves as being the type of person to whom people are willing to respond in a helpful way (Bartholomew, 1997).         

     Given these two working models, four possible prototypes are likely to arise.  Two of these prototypes assess one’s perception of self, using polar extremes by looking at the extent to which the individual feels deserving of positive response from others (Griffin & Bartholomew, 1994).  The other two prototypes assess one’s perception of others’ availability by using bipolar ratings as well (Griffin & Bartholomew, 1994). 

    From this two-dimensional model, four possible combinations are derived.  The first prototype, known as secure, indicates a positive evaluation of self and others (Bartholomew & Horowitz, 1991).  The second, known as preoccupied, indicates a positive evaluation of others, but a negative evaluation of self (Bartholomew & Horowitz, 1991).  This type is consistent with Hazen and Shaver’s ambivalent type (Bartholomew & Horowitz, 1991).  The last two prototypes are more specific forms of the avoidant type laid forth by Hazen and Shaver (Brennan and Shaver, 1998).  The third category, the dismissive avoidant has a positive evaluation of self but a negative evaluation of others (Bartholomew & Horowitz, 1991).  Last, the fearful avoidant has a negative evaluation of both self and others (Bartholomew & Horowitz, 1991). 

     On a dimensional level, these are the basic descriptors of each type, however, on a deeper level each prototype is characterized by a unique set of motivations, thoughts, and behaviors.  The secure individual has a sense that other people are generally responsive and that they are generally deserving of love and attention.  According to Bartholomew and Horowitz (1991), they have the ability to maintain close relationships while maintaining a strong sense of self, to discuss relationships in a coherent manner, and to nurture intimate friendships. 

     In addition, another study using the Self-Report Measure of Adult Attachment Style found that secure individuals have a more positive evaluation of both their current family and family of origin, have a personality characteristic which indicated self-confidence and enhanced functioning in interpersonal relationships, and are less likely to resort to immature defense styles to resolve conflicts than people with insecure attachments (Diehl, Elnick, Bourbeau, & Labouvie, 1998).  The caregivers of adults with secure attachments are judged to be consistently responsive to the individual's needs (Bartholomew, 1997).

     The preoccupied individual tends to hold a negative perception of their own lovability, but still places faith in the availability of others as a source of love and support.  To disguise their own fragile self-concept, they continually strive for the acceptance and admiration of others.  They tend to idealize others and exaggerate emotionality when discussing relationships (Bartholomew & Horowitz, 1991).  They are overly consumed by their own attachment needs and are constantly seeking close relationships to fulfill these needs (Bartholomew, 1997).  As a result, they are more vulnerable to distress when their needs are not met than the other three groups (Griffin & Bartholomew, 1994).  They are also more likely to become victims of domestic violence.  Inconsistent parenting leads to this attachment style by making the individual feel they are to blame for those times that love is unavailable (Bartholomew, 1997). 

     The last two types share some similarities.  Both dismissive and fearful individuals avoid intimacy, especially when under stress.  Also, both attachment patterns typically develop in response to a history of unresponsive and rejecting attachment figures (Bartholomew, 1997).  However, there are some differences between the two.  Dismissing individuals appear to be assured of their own self-worth and lovability but, at the same time, deny the importance of close relationships (Lopez, Fuendeling, Thomas, and Sagula, 1997). 

     They come across as extremely independent and invulnerable, as a means of protecting themselves from disappointment.  They have restricted emotionality and lack clarity when discussing relationships (Bartholomew & Horowitz, 1991).  They score higher in self-esteem than fearful avoidants and are described by themselves and their peers as cold, “competitive, autocratic, and introverted” (Brennan & Shaver, 1998). 

     Fearful avoidants, on the other hand, are described as submissive, “subassertive, introverted, and exploitable” (Brennan & Shaver, 1998).  They do not believe in their own lovability and also do not feel that others view them as lovable (Lopez, Fuendeling, Thomas, and Sagula, 1997).  To avoid rejection, they avoid close contact and intimacy with others (Bartholomew & Horowitz, 1991).  Despite this shying away from intimacy, the fearful avoidant is highly dependent on others for their validation of self-worth (Griffin & Bartholomew, 1994).  This type has been labeled the angry attachment and has been associated with anger, jealousy, verbal abuse, and domestic violence (Kesner & McKenry, 1998).  Although both the fearful and dismissive have experienced rejection in childhood, the caregivers of the fearfully attached are thought to have been especially preoccupied and unavailable (Brennan & Shaver, 1998).  The parents of fearful avoidants often suffer from depression and/or alcoholism (Brennan & Shaver, 1998). 

     The most common of the four types are secure individuals, comprising 47 percent of the sample in one study.  In the same study, 18 percent were classified as dismissing, 14 percent as preoccupied, and 21 percent as fearful (Horowitz, Rosenberg, & Bartholomew, 1993).  Studies of the four types have found age and sex to have a significant impact on this distribution.  In Bartholomew and Horowitz’s (1991) original study, they found through the interview portion of their study that female subjects are more likely to form preoccupied attachments, while male subjects are more likely to form dismissive attachments.

     In a study conducted by Diehl and his colleagues (1998) significant differences on the basis of age were found.  Young and middle-aged adults were found to comprise a larger portion of the preoccupied and fearful categories.  They make the observation that both of these are considered “other oriented” styles of attachment and suggest that the overrepresentation of young people in these categories is due to the formative nature of young identities.  In other words, young people are influenced more greatly by the perception of others because they are still “trying on” their identities.  In contrast, a larger proportion of older adults describe their attachment style as being dismissive.  According to Diehl and his colleagues, this relationship is consistent with other findings that suggest that affect decreases with age, as an adaptation to the increase in age-related losses experiences by older adults.  Despite these differences, Diehl and his colleagues feel that it is too soon to simply accept these changes in attachment as being part of the developmental process.

            Two other variables that have been found to have an impact on attachment formation are parental death and divorce.  Research reveals that relationships with our primary attachment figure, typically our mother, have the greatest affect on the way in which our later attachments are constructed.  Several researchers suggest that separating a child from its mother significantly increases the likelihood of that child developing an avoidant attachment (Hazen & Shaver, 1994b).  It has also been found that children who are separated from their primary attachment figure, either through death or divorce, are more likely to suffer from depression, abuse alcohol, have a lower achievement level, and have worse overall health and sense of well-being (Hazen & Shaver, 1994a, 1994b). 

     One last important point to keep in mind is that all the types discussed are ideal types.  No one individual falls perfectly into any of these categories for this reason.  Bartholomew & Horowitz stress the importance of using continuous, rather than categorical ratings, when administering their assessment of adult attachment.  They feel that categorical ratings ignore the complexity and variability of attachment relationships (Bartholomew & Horowitz, 1991; Bartholomew, 1997).  Most often, continuous ratings have been obtained by using a five or seven point Likert scale (Bartholomew & Horowitz, 1991; Diehl, Elnick, Bourbeau, & Labouvie, 1998).

    The present study built upon the four types set forth by Bartholomew and Horowitz (1991) by using their Self-Report Measure of Adult Attachment Style scale.  These four types were used to test the hypothesis that people with a low level of satisfaction with family of origin functioning are more likely to form insecure attachments than people with a high level of satisfaction.  Conversely, people with a high level of satisfaction with family of origin functioning should be more likely to form secure attachments.  Family of origin functioning was assessed using the Family APGAR, which is a five-question scale that evaluates the individual’s overall subjective evaluation of family dynamics.   Both of the measurement devices used were carefully chosen for their ability to assess subjective interpretations of self and others, considering the more recent research that suggests the importance of subjective evaluations in the development of attachment patterns.



     The sample for this study consisted of 100 undergraduate students from McKendree College.  All participants were recruited from social science courses.  This population was chosen solely for convenience purposes.  Because this is a convenience sample, the results of this study cannot be generalized to the McKendree College student body. 

     Most of the participants in this study were Caucasian and under the age of twenty-five.  This study had a relatively equal number of males and females, with slightly fewer males.

Testing Materials

     Information for this study was obtained using a standardized survey.  The survey consisted of six parts.  The first part, questions 1 and 2, was used to determine each individual’s primary attachment figure.  Part 2, question 3 a-e, consisted of a measurement device known as the Family APGAR, which was used to assess the individual’s level of satisfaction with family of origin functioning (Smilkstein, 1992).  The independent variable was measured using this scale.  Part 3 was used to examine each participant’s interpretation of the care they received in early childhood (under age six).  Part 4, questions 6 and 7, assessed whether or not the individual had ever experienced parental separation due to the death or divorce or his/her caregiver(s).  Both death and divorce are critical variables to assess because previous research shows that both have a significant impact on attachment formation.  Part 5, questions 8-11, measured the dependent variable by determining each individual’s style of attachment.  Part 5 is a four-question measurement devise, known as the Self-Report Measure of Adult Attachment Style (Bartholomew & Horowitz, 1991).  The last part, questions 12-14, measured the control variable of race, sex, and age.


     This study evolved from the article “Adult Attachment Styles:  Their Relations to Family Context and Personality” published by Manfred Diehl and his colleagues (1998).   After formulating the hypotheses both measurement devices, the Family APGAR and the Self-Report Measure of Adult Attachment Style, were obtained.  Then, all other important variables were operationalized, including family type, family structure, and early childhood experience.  Before the surveys were distributed, they were pre-tested on a group of students in an Experimental Psychology class.  Finally, the surveys were administered in a classroom setting during regularly scheduled class times.  After administration, the completed surveys were entered into SPSS, a statistical program in which computer generated analyses can be conducted.  Results were analyzed using Levene’s Independent Sample T-Tests and Bivariate Correlations.


     Bivariate analyses were used to examine the relationship between the independent and dependent variables.  The findings were as follows.  Level of satisfaction with family of origin functioning, as measured by the Family APGAR, was significantly correlated to the secure attachment style at the .001 level.  Conversely, level of satisfaction with family of origin functioning was negatively correlated with the fearful attachment style at the .009 level.  In other words, the more satisfied a person is with his/her family of origin, the more likely he/she is to have a secure attachment style, and the more dissatisfied a person is, the more likely he/she is to be fearfully attached.  Mixed results were obtained for the two middle categories of attachment.  APGAR scores were not significantly related to the preoccupied attachment style, but the dismissive attachment style was found to be significant at .028. 

     A similar relationship was found between perceptions of parental availability during childhood, as measured by the Hunter Childhood Scale, and attachment style.  Feelings of parental availability were positively correlated to the secure attachment style at .001 and negatively correlated to the fearful attachment style at .026.  The preoccupied and dismissive attachment styles, on the other hand, were not significantly related to the Hunter Childhood Scale.

     Another interesting finding is that perception of parental availability during childhood is related at the .001 level to actual parental response during early childhood.  This finding settles the debate regarding the importance of actual versus perceived family events, by suggesting that the two are inextricably linked. 

     Research suggests that family type, sex, age, and the experience of parental death/divorce has a significant impact on attachment formation.  This study failed to support such findings.  No relationship was found between attachment style and sex or age.  The latter result was probably due to a lack of age variation within the sample.  The only relationship found for family type is that the traditional two-parent family is correlated to the secure attachment style at the .029 level.  Analyses failed to find significance for the death or divorce variables, with the exception of a significant negative .004 correlation between divorce and the secure attachment pattern.  The only relationship found between race and attachment is that African Americans are more likely than white people to form preoccupied attachments (.046*).  Additional bivariate analyses can be found in Table 1.



Table 1.  Zero-Order Correlations Among Study Variables


                                          1                 2                 3                 4                5                 6  



1.  APGAR                       

2.  Childhood                    .001*


3.  Memory                      .001*            .007*


4.  Secure                        .001*            .001*         .027*


5.  Dismissive                  -.028*           -.261          -.401            -.179     


6.  Preoccupied                .357             -.836          .625              -.273           -.273


7.  Fearful                       -.009*          -.026*        -.028*           -.001*            .203            .001*


*p < .05




     Table 2 provides the descriptive statistics for the study.  As can be seen in the table, participants were mostly young (under age 25) and mostly white.  The male-to-female ratio was almost equal, with slightly more females.  Most participants has never experienced the death of a caregiver and about 33 percent had experienced parental divorce. 

     Participants tended towards the high end on the family satisfaction scale, with an average score of 26 on a 5-35 point scale.  Participants scored even higher on feelings of parental availability during early childhood, with a mean of 29 on a 5-35 point scale.  The descriptives also reveal that, in most instances, one or both parents were available and supportive during an actual traumatic/upsetting childhood event.

     As for the four attachment styles, the most common was the secure attachment style, followed closely by the dismissive attachment.  Fearful attachments were the third most common, with preoccupied attachments being the least common.  These findings are fairly consistent with other research.  All of the descriptives can be found in the Table 2.


Table 2.  Descriptive Statistics for Study Variables


Variable                                Metric                                          Mean          S.D.       Range

APGAR                   5=low satisfaction; 35=high satisfaction              26.41          7.24          5-35

Childhood                5=low satisfaction; 35=high satisfaction               29.50          6.23         11-35

Memory                   1=both parents unsupportive,

                                 2=one parent supportive, 3= both supportive       2.54             .78           1-3

Secure                      1=very unlike me; 7=very like me                       4.71           1.67           1-7

Preoccupied             1=very unlike me; 7=very like me                        3.40           1.77           1-7

Dismissive               1=very unlike me; 7=very like me                         4.70           1.55           1-7

Fearful                     1=very unlike me; 7=very like me                        3.90           1.93           1-7

Divorce                    0=no, 1=yes                                                         .33             .47           0-1

Death                       0=no, 1=yes                                                         .22             .55           0-1

Sex                           1=male, 2=female                                              1.54             .50           1-2

Age                          1=below 25, 2= above 25                                    1.09              .29           1-2

Race                         1=white, 2=black, 3=other                                  1.09              .29           1-2






     The present study is important because it contributes to a relatively new and growing body of research examining the relationship between one’s self-perception of family functioning and attachment style.  Consistent with modern research, the findings of this study suggest that subjective evaluations of family relationships are critical determinants of one’s ability to form secure attachments to others.  Such research is necessary if psychologists are to ever understand the best way to approach the reversal of insecure attachment patterns.  As suggested earlier, insecure attachments contribute to a broad range of social problems.  The more we begin to understand the development and perpetuation of insecure attachments, the more we can do to alleviate such societal problems.

     It is important to note that this study cannot be generalized to the entire population.  It was drawn from a small, non-random sample, which lacked diversity in terms of age, race, and family type.  Future researchers should consider drawing from a larger and more representative sample.  Another suggestion for future research is that multiple measurement devices be used to assess respondents’ relationships with their family of origin.  In-depth interviews, used in conjunction with the Family APGAR, may provide a more accurate assessment of the relative importance of actual versus perceived events.  Also, future researchers may want to assess attachment patterns in terms of different relationships.  The Self-Report Measure of the Adult Attachment Style Scale could be used separately to study attachment patterns in relation to family, friends, acquaintances, and significant others.  Clumping this all together seems to ignore variation in the way individual’s bond in different contexts. Despite these limitations, this study provides a good baseline upon which to structure a more extensive analysis of adult attachment relationships.    




Bartholomew, K. & Horowitz, L. M.  (1991).  Attachment styles among young adults: a test of a

     four-category model.  Journal of Personality and Social Psychology, 61, 226-244.

Bartholomew, K.  (1994).  Assessment of individual differences in adult attachment. 

     Psychological Inquiry, 5, 23-28.

Bartholomew, K.  (1997).  Adult attachment processes: individual and couple perspectives. 

     British Journal of Medical Psychology, 70, 249-263.

Bowlby, J.  (1973).  Attachment and loss: Vol. 2. Separation.  New York:  Basic Books.

Bowlby, J.  (1977).  The making and breaking of affectional bonds.  British Journal of

     Psychiatry, 130, 201-210.

Brennan, K. A. & Shaver, P. R.  (1998).  Attachment styles and personality disorders:  their

     connection to each other and to parental death, and perceptions of parental caregiving. 

     Journal of Personality, 66, 835-879.

Diehl, M., Elnick, A. B., Bourbeau, L. S., & Labouvie, G.  (1998).  Adult attachment styles: 

     their relations to family context and personality.  Journal of Personality and Social

     Psychology, 74, 1656-1669.

Griffin, D. & Bartholomew, K.  (1994).  Models of the self and other:  fundamental dimensions         

     underlying measures of adult attachment.  Journal of Personality and Social Psychology

     67, 430-445.

Hazen, C. & Shaver, P. R.  (1990).  Love and work:  an attachment-theoretical perspective. 

     Journal of Personality and Social Psychology, 59, 270-280.

Hazen, C. & Shaver, P. R.  (1994).  Attachment as an organizational framework for research on

     close relationships.  Psychological Inquiry, 5, 1-22.

Hazen, C. & Shaver, P. R.  (1994).  Deeper into attachment theory.  Psychological Inquiry, 5,


Horowitz, L. M., Rosenberg, S. E., & Bartholomew K.  (1993).  Interpersonal problems,

     attachment styles, and outcome in brief dynamic psychotherapy.  Journal of Consulting &

     Clinical Psychology, 61, 549-560.

Kesner, J. E. & McKenry, P. C.  (1998).  The role of childhood attachment factors in predicting

     male violence toward female intimates.  Journal of Family Violence, 13, 417-433.

Lopez, F. G., Fuendeling, J., Thomas, K., & Sagula, D.  (1997).  An attachment-theoretical

     perspective on the use of splitting defenses.  Counseling Psychology Quarterly, 10, 461-473.

Smilkstein, G. (1992).  The APGAR questionnaire.  Screening for social support: family, friends,

     and work associates.  Unpublished Manuscript, University of California, Davis.

Volling, B. L., Notaro, P. C., Larsen, & J. J.  (1998).  Adult attachment styles:  relations with

     emotional well-being, marriage, and parenting.  Family Relations, 47, 355-368.






Appendix A.

Family Relations Survey


Thank you for participating in the following survey.  If at any time, you wish to withdraw your participation you are free to do so.  You are also free to leave any questions unanswered that make you uncomfortable.  The results of this survey are completely confidential.  Do NOT

write your name anywhere on this paper.  Thanks again.



1.  You were raised:


 ___  in a traditional two-parent family  (parents living in same home)


 ___  in a two-parent divorced/separated family (have a relationship with both parents,

          but they do not live in the same home)


      ___  in a single-parent family (relationship with only one parent)


      ___  in multiple families


      ___  Other




2.  Your primary caregiver(s) during childhood/adolescence included:

     (Check all that apply)

         *Primary Caregiver:  those who provided the most significant portion of your care.



a.       Biological Mother   ____                 g. Foster Mother            ____


b.      Biological Father    ____                  h.  Foster Father            ____


c.       Step Mother             ____                  i.  Grandmother              ____


d.      Step Father               ____                  j.  Grandfather               ____


e.       Adopted Mother     ____                   k.  Other Relative           ____


f.        Adopted Father      ____                    l.  Other Non-Relative   ____

Based on your relationship with the caregivers mentioned in question 2, please answer the following questions by circling the appropriate answer. 

* One = Never;   7 = Always


I feel that:


3.       I can turn to my family for help.                                          1     2     3     4     5     6     7


4.      My family talks over and shares their problems.                1     2     3     4     5     6     7


5.      My family supports my new activities.                                1     2     3     4     5     6     7


6.      My family is affectionate and responds to emotions.        1     2     3     4     5     6     7


7.      My family and I share time together.                                   1     2     3     4     5     6     7




If you can remember, please answer the following questions about your early childhood (under age 6) by circling the appropriate answer. 

* One = Never;   7 = Always


During my early childhood, my primary caregiver(s):


8.      Comforted me when I was physically hurt/sick.                 1     2     3     4     5     6     7


9.      Comforted me when I was sad/upset.                                   1     2     3     4     5     6     7


    10.  Comforted me when I had a bad dream.                               1     2     3     4     5     6     7


11.   Were affectionate with me in general.                                1     2     3     4     5     6     7


12.   Were available in times of need.                                         1     2     3     4     5     6     7




13.  If you can remember, please report your most traumatic/upsetting memory of early

     childhood (under age 6) and how your parental caregivers responded:






14.  Have you ever experienced parental separation/divorce?      Yes     or      No


            15.  If yes, at what age?   ______



16.  Have you ever experienced the loss of a primary caregiver,             Yes     or     No

       either through death/separation?   


            17.  If yes, at what age?  _____



Please respond to the following scenarios by circling the appropriate answer.



18. It is easy for me to become emotionally close to others.  I am comfortable depending   on others and having others depend on me.  I don’t worry about being alone or having others not accept me.


(Very Unlike Me)      1       2       3       4       5       6       7     (Very Like Me)



19. I am comfortable without close emotional relationships.  It is very important to me to feel independent and self-sufficient, and I prefer not to depend on others or have others depend on me.


(Very Unlike Me)      1       2       3       4       5       6       7     (Very Like Me)



20. I want to be completely emotionally intimate with others, but I often find that others are reluctant to get as close as I would like.  I am uncomfortable being without close relationships, but I sometimes worry that others don’t value me as much as I value them.


(Very Unlike Me)      1       2       3       4       5       6       7     (Very Like Me)



21. I am uncomfortable getting close to others.  I want emotionally close relationships, but I find it difficult to trust others completely, or to depend on them.  I worry that I will be hurt if I allow myself to become too close to others.


(Very Unlike Me)      1       2       3       4       5       6       7     (Very Like Me)




Please answer the following demographic questions.



22.  Sex:     Male   or   Female



23.  Age   ____         



24.  Race/Ethnicity  ___________