Gender, family competence, and psychological symptoms
In this study (n =197) I examine the relationship between the presence of individual psychological symptoms as reported on the Symptom Checklist-90-R (SCL-90-R) and family competence as measured by the Beavers Interaction Scales, with particular emphasis on how these relationships vary according to gender. Despite the common belief that women need relationships more than men, the psychological health of men in this study was as strongly related to their perceptions of family quality as women’s, perhaps even more so. Male and female adolescents appeared to experience family interactions quite differently. Implications for research and practice are addressed.
Family systems theorists assume that individual health and well being are not separate from the quality of family relationships. The connection between family interaction processes and the onset and course of individual symptoms is an area of ongoing study within family therapy (e.g., Beavers & Hampson, 1990; Fisher & Ransom, 1995; Friedmann et al., 1997; Harvey & Bray, 1991; Olson, 1993; Pinsof & Wynne, 1995). Yet how these associations may vary from one family member to another is a relatively recent area of inquiry. Gender studies suggest they are not necessarily the same for women as for men (e.g., Hampson & Beavers, 1987; McGoldrick, Anderson, & Walsh, 1989; Mirkin, 1994). This study examines how the quality of family interaction processes relates to the presence of individual psychological symptoms and how these associations differ according to gender. Because of interest in basic family processes, the focus is on a broad range of families in both clinical and nonclinical populations.
Most models for describing and classifying variations in quality of family functioning focus on similar dimensions: joint problem solving, organization, and emotional climate (Group for the Advancement of Psychiatry, 1996). Whether measured by self-report or by an observer, they identify patterns of observable, interactive functioning and how well family members are able to respond to each other and adapt to change. The Beavers model of family interaction labels this functional ability to “perform the necessary tasks of organizing and managing itself’ (Beavers & Hampson, 1993, p. 74) “family competence.” This study uses the Beavers construct of family competence as an indicator of the quality of family interactions.
According to the Beavers classification system, approximately 75% of all families fall within the midto adequate range of functioning. Adequate family systems have clear boundaries and tend to be caring and effective parents. They are less skilled in negotiation and more focused on control than the less frequent (5%) optimal families. Midrange families have more boundary and control problems, and conflict is less likely to be resolved. They tend to exhibit stereotypic gender behavior. About 19% of American families, classified as borderline and severely dysfunctional, have much more difficulty coping and solving problems. Neither optimal nor severely dysfunctional families are included in this study.
Beavers and Hampson (1990) report fewer clinical problems as family competency increases. They found mild neurotic symptoms and behavior disorder among midrange families, with more serious disorders such as conduct disorders and sociopathology among borderline and severely dysfunctional families. This association between the psychological symptoms of family members and the quality of family functioning is supported by other studies (Fisher & Ransom, 1995; Olson, 1993).
Yet the relationship between family competence and individual health may not be the same for all family members or for all measures of personal well-being and psychological functioning, particularly among healthier families and persons with less severe psychological problems. Gontang and Erickson (1996) found that while some personality patterns on the Millon scale correlated with family competency as expected, the reverse was true for others. In the California Health Study, the relationships between symptoms and type of family functioning were different for wives than for husbands (Fisher & Ransom, 1995). Family factors were associated with differences in adolescent functioning, but in different ways than for adults (Ransom & Fisher, 1995).
Gender shapes how people experience themselves and others and influences their psychological health and well-being. However, the relationship between gender, family processes, and the presence of psychological symptoms is not clear. Men and women tend to approach relationships differently. Several studies of gender differences in scores on family measures have shown that women are more oriented to emotional expression and cohesion, while men appear more oriented toward independence, leadership, and organization (Hampson & Beavers, 1987; Kosek, 1998). Even what constitutes well-being and definitions of “appropriate” responses to distressing or problematic issues may be different for women than for men.
Women have been discouraged from showing power or expressing anger (Goodrich, 1991; Hare– Mustin, 1991; Miller, 1991). Female depression and anxiety have been related to internalized anger, inhibited assertiveness, and loss of emotional connection in key relationships as well as to feeling responsible for relationship failures and a sense of inauthenticity associated with doubting the appropriateness of one’s actions (Anderson & Holder, 1989; Kaplin, 1991). Other female symptoms such as addictions, somatic symptoms, and eating disorders have also been related to internal conflict regarding power and relationship issues (Bepko, 1989; Perlick & Silverstein, 1994).
In contrast, cultural directives for males to be assertive and “in control” encourage men to deny their dependency needs and can make anger the only acceptable male emotion (Meth, 1990; Miller, 1991; Pearson, 1993). According to Meth (1990), social expectations restrict the emotionality available to men and may lead to neglect or denial of internal needs and to behavior that is self-destructive and hurtful to others, despite men’s desire to feel relationally connected. These tendencies are associated with violence, addictive behavior, fear of intimacy, relationship problems, and unrecognized depression (Allen & Gordon, 1990).
For these reasons, it is sometimes assumed that men are less dependent on relationships or do not need others as much as women. It has also been hypothesized that relationship quality is particularly important to the well-being of women; that women need to experience mutual, two-way connections with others to facilitate their own development and will sometimes seek relationships at the expense of their own autonomy (e.g., Jordan, Kaplan, Miller, Stiver, & Surrey; 1991; Josselson, 1987). Yet because traditional gender stereotypes limit emotional functioning and relationship opportunities for both women and men, it is difficult to know how the correspondence between experiencing psychological symptoms and family competence may differ according to gender.
To determine how particular psychological symptoms are related to family competence, the scores of volunteer family members on the Symptom Checklist-90-R (SCL-90-R), a checklist of psychological symptoms, and their scores on the Beavers Self-Report Family Inventory and the observer-rated Beavers Family Interaction Scale (BFIS, competence dimension) were compared and differences between males and females identified.
Sample and Procedure
Eighty-two volunteer families solicited from Protestant and Catholic churches and 15 families who were currently in family therapy at a university training clinic participated in the study. Seventy-two percent of the white, working- and middle-class sample were first marriages, 20% were remarriages, and 7.7% were divorced families. Self-report data were collected from the adult family members and the oldest child living at home, age 12 or older. The total sample (n = 197) included 91 adult women, 74 adult men, 17 adolescent females, and 12 adolescent males. Young children were included in the study, but no self-report data were collected from them.
The interaction patterns within 22 of the families also were rated by observers. Following the protocol suggested by Beavers and Hampson (1993), these families were asked to discuss for 10 min what they would like to change about their families. These discussions, which occurred at the training clinic, were videotaped while the observers were out of the room and were rated by a trained pair of raters according to the BFIS. The observed families included eight “clinical” families who were currently participating in family therapy and 14 nonclinical families.
Family competence. Family competence was measured according to the BFIS. This scale, developed out of 25 years of research isolating the essential differences between dysfunctional, better-functioning, and healthy families and based on general systems theory, assesses family interaction according to two dimensions: competence and style (Beavers & Hampson, 1990). Only the competence dimension, designed to measure how well a family functions, was used for this study. Both an observational rating system, the Beavers Family Interactional Scales-Family Competence Dimension (FIS-C) and a Self-Report Inventory-Version II (SFI) were used.
The observer-rated FIS-C captures how well the family performs its necessary tasks. Families were rated from 1 to 5 on each of 12 items (overt power, parental coalitions, closeness, mythology, goal-directed negotiation, clarity of expression, responsibility for personal actions, permeability or responsiveness to others, range of feelings, mood and tone, unresolvable conflict, and empathy). They were also rated from 1 to 10 according to the observer’s assessment of the family’s overall, global competence. Low scores are indicators of better family functioning. Scores of global competence ranged from 1.8 to 7.7, with a mean of 4.0. Eighty-eight percent were scored in the midrange to adequate categories of the FIS-C. None rated optimal or severely dysfunctional.
Raters completed 12 hr of training using materials provided by the developers of the scale. Interrater reliability (i.e., correlations of raters scores) ranged from .71 to .79. Items with lower interrater reliability were not included in the study.
The SFI was developed by Beavers and his colleagues to measure the same theoretical constructs measured by the FIS. Family members assess their family on 36 items, rating each from 1 to 5. Low scores indicate higher family functioning (Beavers & Hampson, 1993). Scores in this sample ranged from 42 to 134, with a mean of 69.06. Thus, most of the sample reported “adequate” levels of functioning. Only 10% reported scores that suggested significant problems in functioning (i.e., a mean item response of 2.5 or total score > 90).
The SFI has five subscales:
1. Health (19 items) addresses happiness, optimism, problem-solving and negotiation skills, family love, strength of adult coalitions, autonomy/individuality, and minimal blaming/increased responsibility patterns.
2. Conflict (12 items) indicates low levels of unresolved conflict, fighting, blaming, and arguing, with higher levels of negotiation and acceptance of personal responsibility in solving conflict situations.
3. Cohesion (5 items) is a self-rating of family style. Low scores reflect satisfaction and happiness through togetherness and an emphasis on family closeness.
4. Leadership (3 items) involves strong and consistent patterns of adult leadership in the family.
5. Emotional expressiveness (5 items) involves perceptions of feelings of closeness, physical and verbal expressions of positive feelings, and the ease with which warmth and caring are expressed by family members.
Psychological symptoms. The presence of psychological symptoms was measured by the SCL-90-R symptom checklist. This 90-item self-report scale is one of the most popular means of operationally defining psychological health or distress (Derogatis, 1983). It yields an individual profile on each of the nine subscales listed below. Lower scores are indicators of fewer psychological symptoms.
1. Somatization (12 items) reflects distress arising from perceptions of bodily dysfunction, such as headaches, pain, and physical discomfort.
2. Obsessive-compulsive (10 items) focuses on thoughts, impulses, and actions that are experienced as unremitting, unwanted, and irresistible by the individual.
3. Interpersonal sensitivity (9 items) reflects feelings of personal inadequacy and discomfort and negative expectations concerning interactions with others.
4. Depression (13 items) reflects manifestations of clinical depression such as dysphoric mood and affect, withdrawal from life, lack of motivation and energy, hopelessness, and suicidal thoughts.
5. Anxiety (10 items) is composed of symptoms such as nervousness, tension, trembling, feelings of apprehension, and panic attacks.
6. Hostility (6 items) reflects feelings and thoughts associated with anger, rage, and resentment.
7. Phobic anxiety (7 items) reflects persistent fear in response to a specific person, place, object, or situation that is irrational or disproportionate to the stimulus.
8. Paranoid ideation (6 items) reflects thoughts such as projecting blame on others, hostility, suspiciousness, grandiosity, and fear of losing autonomy.
9. Psychoticism (10 items) represents a continuous dimension of interpersonal alienation from being withdrawn and isolated to symptoms such as hallucinations and thought broadcasting.
Several additional items are summed into the total score, including poor appetite, overeating, trouble sleeping, feelings of guilt, and thoughts of death.
Respondents rated themselves according to what extent each of the 90 problems had distressed or bothered them during the past 7 days. A score of 0 indicated “not at all.” A score of 4 indicated “extremely.” The total score (TSCL) is simply the sum of the responses and TSCL ranged from a minimum of 0 to a maximum of 211, with a mean of 37.1. A Global Severity Index (GSI), a Positive Symptom Index (PSI), and a Positive Symptom Distress Index (PSDI) were also calculated but are not reported here because they did not yield any different results than the TSCL.
The first level of analysis examined the correlations between the symptoms reported on the SCL-90 and each of the measures of family competence for the entire sample. Correlations were also determined for the clinical portion of the sample, for adolescents, and for respondents scoring above the mean on the SFI (i.e., reporting less family competence) and the SCL-90 (i.e., reporting more symptoms).
The second level of analysis examined the male and female samples independently for both the adults and the adolescents. Correlations between symptoms, SFT scores, and partners’ and parents’ scores were determined for each gender. Linear regression analyses to predict symptoms based on the SFI score and the partner’s SFI were also conducted for each gender.
Though our goal had been to link particular symptoms with specific, observed family interaction patterns, we found that the respondents’ overall perception of family health was more associated with psychological symptoms than any particular interaction patterns. The self-report of family competence was more significantly related to the number of psychological symptoms reported by respondents than was the observer-rated measure. This section, therefore, first presents a brief review of the comparisons between the self-report measure and the observer-rated measure and then focuses primarily on the SFI data and the differences found there between groups of respondents.
Self-Report versus Observer-Rating Scales
Correlation between family scales. Scores on the observer-rated FIS-C and the self-report SFI were .51 correlated. This relationship was significant at the .001 level and is similar to the relationship between scales reported by Hampson, Beavers, and Hulgus (1989) for a nonclinical population. Correlations with the clinical portion of the sample were slightly higher. Within the SFI, the health, conflict, and expression subscales were more highly correlated with the observer-rated scale than with the cohesion and leadership subscales.
Relationship to psychological symptoms. Total score on the SCL-90 correlated.46 with the SFI but only .19 with the observer-rated FIS-C. When the various dimensions of the FIS-C were examined in relation to particular psychological symptoms, only somatization and hostility were significantly related (.01 level) to specific family interaction patterns. Somatization correlated .37 with adult coalitions, .44 with tone, and .42 with global assessment of family health. In a stepwise multiple regression, 30% of the variance in somatization could be accounted for by the tone of the family interaction (P = .18), the degree to which family members take personal responsibility ((beta) = -.32), and the global assessment of family health .16). Hostility correlated .39 with the global assessment of family health and .44 with tone.
SFI and Psychological Symptoms
Scores on the SFI and the SCL-90 were significantly related at the .001 level for each of the symptom subscales except phobic anxiety. People reporting more positive family interactions reported fewer psychological symptoms. The relationship was .46 for the total sample, .54 for the clinical sample, and .64 for those persons who scored above the mean on the SFI (69.01). A summary of these correlations is shown in Table 1.
Mean scores on the SFT and the SCL-90 were similar for adult women and men, except that women reported more depression. The relationship between symptoms (TSCL) and perceptions of family interactions (SFI) was significantly correlated for both men and women. The magnitude of the correlation was greater for men on seven of the 10 scales and greater for women on two of the scales, as shown in Table 2.
In these separate analyses, interpersonal sensitivity, depression, anxiety, and hostility were significantly related to perceptions of family interactions for both genders, though the correlations were not as high for women. Phobic anxiety was not related to family interactions for either gender (except among the clinical population as shown above). Paranoid ideation and psychoticism were significantly related to family interactions at the .001 level among men, but not among women. Somatic and obsessive-compulsive symptoms were significantly related to family interactions for women at the .001 level but not for men. These gender differences in magnitude of correlations were significant (.OS level) for somatization, paranoid ideation, and psychoticism.
Gender differences according to subscales. Table 3, which shows the correlations between symptoms and perceptions of family interactions according to each subscale on the SFI, provides a more in-depth look at how these relationships compare for women and men. It is useful to note that although they were analyzed separately, these two samples are not independent of each other. The men and women in each sample are describing the same relationships. They report similar levels of symptoms and relationship quality. Yet men and women appear to be affected in some different ways.
Health. This subscale was the most consistently correlated to levels of psychological symptoms. The correlations with depression, anxiety, hostility, and interpersonal sensitivity were similar for women and men. Relationships between family health and somatization and obsessive-compulsive symptoms were significant for women but not for men. Magnitudes of correlations for paranoid ideation and psychoticism were significantly higher for men than women.
Conflict. The same trends in gender differences and similarities appeared.
Cohesion. Though Hampson and Beavers (1987) found that cohesion was particularly salient in determining young women’s perceptions of overall family competence, this subscale was not related to reports of psychological symptoms for adult women. However, correlations for men were significant (.01 level) for anxiety, paranoid ideation, and psychoticism.
Leadership. Strong and consistent patterns of adult leadership were not significantly related to any of the psychological symptoms for either men or women.
Emotional expressiveness. Correlations between psychological symptoms and emotional expressiveness for males were very similar to those described for the health subscale. Fewer of the symptoms were significantly related for females, and the magnitude of the relationship was smaller for women than for men on seven of the symptoms.
As scores on the SFI and the SCL-90 increased, so did partners’ scores. Adult partners’ scores on the SFI correlated .54; partners’ TSCL scores correlated .49. Knowledge of the partner’s score could predict 29% of the variance in SH and 24% of the variance in TSCL.
The relationship between partners’ scores on the SFI and the presence of symptoms was .28 (not significant) for women and .44 for men (significant at .001). Male symptoms were significantly related to how female partners viewed the quality of their relationships. Female symptoms were less related to how male partners viewed their relationships.
A different pattern emerged among couples with at least a 15-point discrepancy in the number of symptoms reported by each partner. When men were more symptomatic (n = 14), women’s symptoms correlated .60 with their partners’ view of the relationship. However, when women were more symptomatic (n = 15), men’s symptoms only correlated .07 with their female partner’s view of the relationship.
Table 4 shows the results of a regression analysis that used SFI scores and partner’s SFI scores (PSFI) to predict the presence of psychological symptoms among adults. The regression equation could predict 40% of the variance for men and 25% of the variance for women. Somatization and obsessive-compulsive were significantly predicted by the regression equation for women but not for men. Anxiety, paranoid ideation, and psychoticism were significantly predicted by the regression equation for men only. A significant interaction effect with PSFI is present for men but not for women. DSFI (the difference between partner’s scores) was excluded.
Adolescents. Adolescents reported more symptoms than adults across all of the subscales on the SCL90. The correlation between their symptoms and the symptoms of the adults in the household was relatively low (.27 for girls and .17 for boys). Similarly, boys’ perception of family quality scores (SFT) correlated only .13 with adults’ SFT (adult scores added together). Yet girls’ SFI correlated .43 with adults’.
The relationships between the adolescents’ psychological symptoms and perceptions of family competence varied considerably depending on the genders of the respondent and the adult family members. These relationships are shown in Table 5. Though these findings must be regarded with caution because the sample size for adolescents was small, they provide an interesting glimpse into the interior of the families in this study.
Girls’ symptoms were quite strongly related to how they perceived interactions within the family. Their TSCL scores correlated .56 (significant at .OS) with their SFI scores. In contrast, boys’ symptoms seemed to have little connection to how they described their families. Boys’ TSCL correlated only .20 (not significant) with their SFI scores.
Mothers’ SFI scores were not correlated with adolescent symptoms for either gender. However, fathers’ perceptions of family interactions correlated .48 with boys’ symptoms and .38 with girls’. Although not statistically significant in this small sample, the father’s experience or description of family interaction was considerably more connected to adolescent symptoms than the mother’s. The biggest father-mother difference was in relation to girls’ ratings of hostility. Girls’ hostility correlated .72 with father’s SFI (significant at .01), but not all with mother’s.
Summary of Findings
Overall, the presence of psychological symptoms was significantly related to perceptions of family competence, particularly among those reporting less family competence.
Except for somatic symptoms, observer scores were less connected to psychological symptoms than were self report scores.
Perceived family competence accounted for a greater amount of the variance in psychological symptoms for men than for women (40% compared to 25%). There was an interaction effect for partner’s perception of family competence for men but not for women.
Somatic and obsessive-compulsive symptoms were significantly related to family competence ratings for women but not for men. Paranoid and psychotic symptoms were significantly related to family competence scores for men, but not for women.
The psychological symptoms of adolescent girls were significantly correlated with their perceptions of family competence; boys’ symptoms were not.
The findings of this study are consistent in many ways with the results of other studies, but they also challenge some assumptions.
Link between Family Interactions and Psychological Symptoms
In this study the link between reported numbers of psychological symptoms and level of family competence is consistent with assumptions by family therapists over the years; that is, that individual psychological symptoms are associated with family functioning. Although this sample included primarily adults and adolescents not currently diagnosed with a mental disorder or participating in therapy, these findings, like those of Hampson et al. (1989), suggest that the relationships tend to be stronger among clinical populations and among families reporting less competence.
The Use of Observational versus Self-Report Data
Self-report and observational data are different subjective perspectives on the same phenomenon. When we designed the study we were concerned that studies about families often rely on only self-report data. We thought it was important to include observational data. Yet for this study the family members’ own perceptions were more useful. Our experiences also point to the difficulty in obtaining valid observational data for research. Training of the raters was time consuming, and accuracy of measurement depended on how the raters interpreted what they observed in a short, 10-min period. It is possible that with different training, different raters, a larger observed sample, or multiple observations the results would have been different. Furthermore, families were being observed in a situation contrived for research purposes. Research based on more in-depth observation may show quite different results. Observation in research is also different than observation in clinical practice. Clinicians have access to a much wider range of information on which to interpret their observations.
Men and Relationships
The most striking finding in this study is that male psychological symptoms were related to their perceptions of family competence as much as women’s. In fact, the magnitude of the correlation was higher for men on seven of the 10 subscales. This suggests that men are probably as dependent on relationships for their psychological well-being as women, possibly more so. The subscales describing levels of cohesion and emotional expressiveness, commonly believed to be female needs, were related to more psychological symptoms for men than women. At first glance this seems surprising. Much attention has been drawn to the importance of relationships to the psychological health and well-being of women (Jordan et. al., 1991). Yet, there is a wealth of research, beginning with Jessie Bernard’s (1972) “his and her marriage,” showing that couple relationships tend to provide a more protective function for men than women. Some scholars (e.g., Hare-Mustin 1991; Kaschak 1992) have argued that male autonomy is a myth that ignores female supports.
Paranoid ideation and psychoticism were significantly correlated with low family competence for men but not for women. Since this sample did not include persons with active paranoia or psychoticism, these symptoms can most usefully be thought of as indicators of how the self is perceived in relation to others. Paranoid ideation on the SCL-90 measures projection of blame on others, hostility, suspiciousness, grandiosity, and a fear of losing one’s autonomy. At lower levels of severity, psychoticism on the SCL-90 measures interpersonal alienation, withdrawal, and isolation. These symptoms reflect the experience of interpersonal distance as well as the behaviors or attitudes that might contribute to it. It is unclear why their correlation with less relationship quality would be greater for men than women. Perhaps these differences are the result of masculine gender orientations toward autonomy that keep them more separate and alienated within an intimate relationship. Whatever the case, it seems likely that gendered expectations regarding self and others contribute to these differences.
Women and Relationships
Distress arising from perceptions of bodily dysfunction, such as headaches, pain, and physical discomfort (somatization) was associated with poor relationship quality for the female sample but not the male. This is consistent with literature suggesting that women feel responsible for relationship quality but may also believe that they should not upset their partners or do not expect that their partners will respond to their needs. Internalizing relationship issues in ways that result in physical expression would seem to be a likely outcome.
The other symptom that related to low family competence for women, but not for men, was obsessivecompulsive. Unremitting, unwanted, and irresistible thoughts, impulses, and actions for women may also be related to taking less voice in a relationship. They could even be an indirect way of getting what one wants or expressing anger or dissatisfaction when not having access to more overt power. Whatever the reason for the difference, the results of this study suggest the need to consider context in order to understand the meaning or function of a symptom and to not assume they are the same for women as for men, even within the same relationships.
Adolescents in Families
These findings suggest different family experiences for adolescent boys and girls. The girls perceived family competence similarly to their adult family members, and their symptoms were related to those perceptions. This is consistent with the developmental argument that maintaining connection remains important to girls during adolescence (Apter, 1990; Brown & Gilligan, 1992). The boys’ reports, however, appeared less “in sync” with those of parents. Their perceptions of family competence did not correlate with those of their adult family members, and their symptoms were not statistically associated with their views of their family life. Perhaps the psychological symptoms of these boys were more strongly correlated with their experiences outside the family. Girls, socialized to be more relationally aware and tuned in to family, perhaps had more accurate perceptions of their families.
This study was based on a white, working- and middle-class sample in a small university town in the western United States. The adolescent portion of the sample was very small. Other studies are required to determine whether the trends found here extend beyond this sample and how they might vary among different populations. Despite these limitations, the findings suggest some useful clinical implications and directions for further research.
The findings remind us to examine the assumptions and conceptual frames that underlie our clinical work and to consider how they influence what we do. For example, for this particular sample, finding satisfaction and happiness through togetherness and closeness (cohesion), as measured by the SFI, were not related to the presence of psychological symptoms, especially for women. Yet many of our models of therapy emphasize the importance of connection for women. It seems important to ask whether, if these research participants were our clients, it would be appropriate to emphasize cohesion in our assessment and treatment of them, and how we would make this decision. Perhaps this model of cohesion means something different than the female need for connection often addressed in the literature. Perhaps the responsibilities for others that women associate with connection confound the possible association between cohesion and women’s psychological symptoms.
Finding that psychological symptoms were as related (or more) to perceptions of family quality for the men in the study than for women particularly gives pause. If clinicians make decisions based on a view that women need relationships more than men or that mothers influence children more than fathers, we may be perpetuating a myth. Similarly, the finding that female symptoms were less related to partner’s perceptions of the relationship than male symptoms is consistent with Lynch’s (1998) finding that women perceived more support from friends and children than they did from their spouses. Men tend to lack supportive friendships and are more reliant on wives. Future research should focus on the nature and implications of these differences.
The look into families afforded by this study suggests that it is important to not oversimplify the family role in teenaged symptoms and that the influential processes may work quite differently for boys than girls. While it was not possible to draw many statistical conclusions from the adolescent portion of the sample, the substantial differences in the responses of the teenaged boys and girls suggest a complex relationship between adolescent development, gender, family quality, and other factors. Future study should include the gender of both parents and teens.
This study shows the psychological symptoms of men, as well as women, statistically associated with perceptions of less family competence. It does not show direction of influence or qualitatively explain how gender, family competence, and psychological symptoms interact. Systemic conceptualizations suggest that the influences are likely to be interconnected and reciprocal. Future study will require that we draw on many different kinds of research findings, avoid simplistic gender and family stereotypes, and look more closely at the links between gender, age, mental health, and relationship patterns.
Allen, J., & Gordon, S. (1990). Creating a framework for change. In R. Meth & R. Pasick (Eds.), Men in therapy: The challenge of change (pp. 131-151 ). New York: Guilford.
Anderson, C., & Holder, D. (1989). Women and serious mental disorders. In M. McGoldrick, C. Anderson, & F. Walsh (Eds.), Women in families: A framework for family therapy (pp. 381-405). New York: Norton.
Apter, T. (1990). Altered loves: Mothers and daughters during adolescence. New York: St. Martin’s.
Beavers, W., & Hampson, R. ( 1990). Successful families: Assessment and intervention. New York: Norton.
Beavers, R., & Hampson, R. (1993). Measuring family competence: The Beavers systems model. In F Walsh (Ed.), Normal family processes (2nd ed., pp. 74-103). New York: Guilford.
Bepko, C. (1989). Disorders of power: Women and addiction in the family. M. McGoldrick, C. Anderson, & F. Walsh (Eds.), Women in families: A framework for family therapy (pp. 406-426). New York: Norton.
Bernard, J. (1972). The future of marriage. New York: Bantam.
Brown, L., & Gilligan, C. (1992). Meeting at the crossroads: Women’s psychology and girls’ development. Cambridge, MA: Harvard University Press.
Derogatis, L. (1983). SCL-90-R administration, scoring, and procedures manual-ll. Towson, MD: Clinical Psychometric Research.
Fisher, L., & Ransom, D. (1995). An empirically derived typology of families. I. Relationships with adult health. Family Process, 34, 161-182.
Freidman, M., McDermut, W, Solomon, D., Ryan, C., Keitner, G., & Miller, I. (1997). Family functioning and mental illness: A comparison of psychiatric and nonclinical famlies. Family Process, 36, 357-367.
Gontang, R., & Erickson, M. (1996). The relationship between Millon’s personality types and family system functioning. American Journal of Family Therapy 24, 215-226.
Goodrich, T. (1991). Women, power, and family therapy: What’s wrong with this picture? In T. Goodrich (Ed.), Women and power: Perspectives for family therapy (pp. 3-35). New York: Norton.
Group for the Advancement of Psychiatry (1996). Global assessment of relational functioning scale (GARF): Background and rationale. Family Process, 35, 155-172.
Hampson, R., & Beavers, R. (1987). Comparing males’ and females’ perspectives through self-report. Psychiatry, 60, 24-30.
Hampson, R., Beavers, R., & Hulgus, Y (1989). Insiders’ and outsiders’ views of family: The assessment of family competence and style. Journal of Family Psychology, 3, 118-136.
Hare-Mustin, R. (1991). Sex, lies, and headaches: The problem is power. In T. Goodrich (Ed.), Women and power: Perspectives for family therapy (pp. 63-85). New York: Norton
Harvey, D., & Bray, J. (1991). Evaluation of an intergenerational theory of personal development: Family process determinants of psychological and health distress. Journal of Family Psychology, 4, 298-325.
Jordan, J., Kaplan, A., Miller, J., Stiver, L, & Surrey, J. (1991). Women’s growth in connection. New York: Guilford.
Josselson, R. (1987). Finding herself.’ Pathways to identity development in women. San Francisco: Jossey-Bass.
Kaplin, A. (1991). The `self in-relation’: Implications for depression in women. In J. Jordan, A. Kaplan, J. Miller, I. Stiver & J. Surrey (Eds.), Women’s growth in connection (pp. 206-222). New York: Guilford.
Kaschak, E. (1992). Engendered lives: A new psychology of women’s experience. New York: Basic.
Kosek, R. ( 1998). Self differentiation within couples. Psychological Reports, 83, 275-279.
Lynch, S. (1998). Who supports whom? How age and gender affect the perceived quality of support from family and friends. Gerontologist, 38, 231-238.
McGoldrick, M., Anderson, C., & Walsh, F ( 1989). Women in families: A framework for family therapy New York: Norton.
Meth, R. (1990). The road to masculinity. In R. Meth & R. Pasick (Eds.), Men in therapy: The challenge of change (pp. 3-34). New York: Guilford.
Miller, J. (1991). The construction of anger in women and men. In J. Jordan, A. Kaplan, J. Miller, I. Stiver & 1. Surrey. (Eds.), Women’s growth in connection (pp. 181-198). New York: Guilford.
Mirkin, M. (1994). Women in context: Toward a feminist reconstruction of psychotherapy. New York: Guilford.
Olson, D. (1993). Circumplex model of marital and family systems: Assessing family functioning. In F Walsh (Ed.), Normal family processes (pp. 104-137). New York: Guilford.
Pearson, E. (1993). Male sexuality and power. In R. Glick & S. Roose (Eds.), Rage, power and aggression. New Haven, CT: Yale University Press.
Perlick, D., & Silverstein, B. (1994). Faces of female discontent: Depression, disordered eating, and changing gender roles. In P Fallon, M. Katzman, & S. Wooley (Eds.), Feminist perspectives on eating disorders (pp. 77-93). New York: Guilford.
Pinsof, W., & Wynne, L. (Eds.). ( 1995). The effectiveness of marital and family therapy [Special Issue]. Journal of Marital and Family Therapy 21 (4).
Ransom, D. & Fisher, L. (1995). An empirically derived typology of families: II. Relationships with adolescent health. Family Process, 34, 183-197.
Loma Linda University
Carmen Knudson-Martin, PhD, is Professor and Director of Marital and Family Therapy doctoral programs, Department of Counseling and Family Studies, Loma Linda University, Loma Linda, CA 92350.
This research was supported by a grant from the College of Education, Health, and Human Development, Montana State University. Special thanks to Stan Meloy, Barbara Johnson, Florence Guest, and June Odegard for their assistance conducting the study and to Gary Conti, Rick Johnson, and Willson Huang for their statistical support.
Copyright American Association for Marriage and Family Therapy Jul 2000
Provided by ProQuest Information and Learning Company. All rights Reserved