Introduction

Serious emotional disturbance (SED) is an umbrella term that captures several severe and persistent mental disorders experienced by children and adolescents, including anxiety, mood, disruptive, and substance-use disorders. This population is grossly underserved due to financial limitations, poor assessment procedures, and a shortage of well-trained professionals (Marsh and Fristad 2002). Those who do enter treatment exhibit a range of mental health problems. Among the most common diagnoses are Oppositional Defiant Disorder (ODD), Pervasive Developmental Disorders (PDD), Mood and Anxiety Disorders, Thought Disorders, and Substance Abuse. The prevalence rate of children and adolescents who struggle with emotional disturbance in the United States is estimated to be between 9% and 19%. The children on the severe end of this spectrum have considerable difficulty in regular classrooms and are often referred to treatment centers for psychologic intervention and academic assistance (Anderson and Mohr 2003).

Children and adolescents with SEDs often face multiple risk factors. According to recent research, these children and youth continue to have one of the poorest long-term outcomes of any disability group, and their difficulties extend into numerous areas of their lives: school performance, behavioral functioning, and social relationships (Satcher 2000). They are continually cited as a high-risk population with increased risk for school drop-out and incarceration. Furthermore, given their diverse needs and the various levels of disturbance severity they experience, they are a difficult population for agencies to serve (Anderson and Mohr 2003).

Although much is known about children with SEDs and their complex needs, little is known about their families, especially the strengths they have and the risk factors they face (Beresford 1994; Green et al. 2001). Historically, researchers have tended to design child-focused studies that do not account for important family variables (Green et al. 2001). Given that families are integral to the treatment of children with SEDs and have strengths in addition to weaknesses, many theorists have highlighted the importance of identifying both resiliency and risk factors. Resiliency factors such as effective problem-solving strategies and adaptive responses to negative life events can lead to positive outcomes for families and their children with SEDs, while risk factors that increase family stress, such as psychologic distress and lack of social support, can lead to negative outcomes (McCubbin et al. 1998; McLoyd 1990; Murry et al. 2001; Walsh 1998, 2003). A model of risk and resiliency will be primarily used in this study to demonstrate the importance of shifting the paradigm of “families as damaged” to “families as challenged” and, thus, address the ways in which families can use their strengths to overcome adversity even in the midst of elevated levels of stress (Walsh 1996; Waters and Lawrence 1993).

In terms of the known risk factors for families of children with SEDs, some studies have shown that the impact of having a child with an SED can exacerbate stressors on the family, including strained parent–child or sibling relationships and increased parenting stress (Smith et al. 2001), all of which have been shown to adversely affect the quality of care giving, child behavior, and school adjustment (Crawford and Manassis 2001). Stress levels can also be exacerbated by financial strain and social isolation (Satcher 2000). Although resiliency factors have not been well studied for families of children with SEDs, the best empirically supported protective factors for at-risk families in general are parenting skills (e.g., discipline, negotiation, and supervision), positive mother-child relationships, and social support (Beresford 1994; Murry et al. 2001; Walsh 1996).

Given the dearth of research on these families’ normative strengths, clinicians may inadvertently base their assessment of the family solely on the risk factors that are immediately apparent during the initial family meeting (Walsh 1996). A practitioner’s focus on families’ or parents’ weaknesses, without acknowledging both their normative stress levels and strengths, may subtly communicate that they are incompetent caregivers. A primary focus on families’ difficulties with their child may also lead parents to assume a more passive role in the assessment and treatment process as they relinquish the role of expert on their children’s and families’ needs to service providers (Lawrence 2006; Walsh 1996). Yet research on day and residential treatment for children with SEDs suggests that family involvement in treatment is significantly associated with desirable child outcomes (Anderson and Mohr 2003; Brannan et al. 2003; Brinkmeyer et al. 2004; Green et al. 2001; Grizenko 1997; Pfeiffer and Strzelecki 1990) and lack of involvement is associated with poorer treatment outcomes (Osher et al. 2002). Thus, it is imperative that clinicians find ways to help these families feel integral to the assessment and treatment process by seeking to understand how they perceive both their resources and stressors.

In fact, many have suggested that interventions for children with SEDs and their families are most compelling and effective when treatment providers assess the parents’ perceptions of their child’s and family’s most pressing needs in order to develop treatment goals that the family finds meaningful (Brannan et al. 2003; Brinkmeyer et al. 2004; Hansen et al. 2002; Kernan et al. 2003; Osher et al. 2002; Stroul and Friedman 1986). For example, Anderson and Mohr (2003) argue that a comprehensive assessment must recognize the expertise that parents perceive themselves already having in caring for their child with an SED. Professionals who champion the families’ perception of their greatest assets and most pressing needs not only encourage their collaboration early in the treatment process, but also strengthen the families’ sense of competency (McCubbin et al. 1998; McLoyd 1990; Murry et al. 2001; Walsh 2003; Waters and Lawrence 1993).

Current Study

The current study will begin to address the dearth of research on parent perceptions of family resiliency and risk factors. Although there is significant heterogeneity among families who care for children with SEDs (Anderson and Mohr 2003), it is useful to determine whether there are patterns of resources and stressors that these families report at the outset of treatment. Thus, this study has two aims: 1) to delineate a possible pattern of resources and stressors that parents caring for children with SEDs experience and 2) to determine whether the protective factors that have been shown to be important for at-risk families, namely parenting skills and social support, also serve as important protective factors for families of children with SEDs.

Method

Participants

The sample consisted of 128 parents or caretakers from New York City—mainly the Bronx and its surrounding areas—80 who had a child with an SED and were considered the clinical group and 48 who did not have a child with an SED and were considered the comparison group. The participants in the clinical group all had a child who had been diagnosed with an SED and was receiving residential or day treatment from a multi-site facility in New York. This agency adopts a family-oriented philosophy of care and provides long-term service to children and adolescents struggling with SEDs.

The 80 children in the clinical sample were predominantly male (88%), ranged in age from 4 to 15 (mean age = 10), and were racially diverse (Table 1). Treatment was almost equally divided between residential care (51%) and day-treatment (49%), and there were no significant demographic differences between these two treatment conditions. As a group, the children had diagnoses that ranged from ODD and Conduct Disorder to Bipolar Disorder and psychosis. The adult respondents identified themselves as parents, grandparents, or legal guardians of the children receiving treatment and were predominantly female (84%), unmarried (66%), and middle-age (mean age = 42). The majority of the adult respondents were employed (60%) and, of those for whom financial data was available (N = 55), the mean household income was between $20,000 and $25,000.

Table 1 Sample characteristics separate by group and respondent
Full size table

The comparison group consisted of 48 parents or caretakers who worked either at the treatment agency or another human service agency in the Bronx and had a child living in their household. This sample was chosen because of the demographic similarities (e.g., location, income, and racial identity) between agency employees and the families in the clinical group, and inclusion was limited to families whose children had not been diagnosed with an SED and were not receiving psychologic and/or special educational services. In addition, none of the comparison participants had ever administered the Family Inventory of Resources and Stressors (FIRST) as part of their job responsibilities. Given that children from this sample had a larger age range than those in the clinical group, one child from each comparison family was selected for analysis on the child-centered scales on the FIRST based on the age range of the identified children in the clinical group. Given the disproportionate number of males who receive services for SEDs, boys from the comparison group who fell within the appropriate age range were selected for inclusion over girls whenever possible. These selection criteria were established to minimize the possibility that comparisons between the two samples on this scale were just capturing developmental differences between age groups or differences between families with male versus female children.

The comparison group participants and their children had sample characteristics similar to those in the clinical group with some exceptions (Table 1). Even though boys from the comparison group who fell within the appropriate age range were selected for inclusion over girls whenever possible, there were more male children in the clinical group. In addition, adults in the clinical group more frequently reported being grandparents of the identified child and, therefore, had a higher mean age. Finally, given that the comparison group was recruited from agencies where they were employed, these respondents had higher employment rates, more education, and slightly higher incomes.

Procedure

For the clinical families, mental health professionals (e.g., social workers, psychologists, psychiatrists, intake coordinators, child-care and health-care workers) administered the FIRST to parents and guardians as a structured interview during the intake phase of their children’s treatment at the agency. This agency used the FIRST as standard intake protocol for all families who were enrolling a child in an agency program. The agency elected to use the interview version of the FIRST rather than the self-report version to establish a collaborative working alliance with families during the early stages of treatment. They asked the family to select a spokesperson and conducted the intake interview with this person. For the study sample, this respondent was a parent, grandparent or legal guardian and was designated as Adult 1. Since it is often difficult for agencies to allocate intake time to gather each family member’s perceptions of the family, the questions on the FIRST were designed to capture the spokesperson’s reflections on the entire family’s resources and stressors (Lawrence 2006). Although individual family member’s perceptions may vary across questions on the FIRST, the assumption is that, given limited agency resources, it is most important to initially capture the perceptions of the family leader who typically assumes the bulk of the responsibility for developing and following through on treatment plans.

Before administering the structured interview, the mental health professionals obtained parents’ verbal consent to complete the FIRST. The professionals then read aloud each question on the FIRST and recorded the parent’s response. After reviewing the adult’s responses, the professionals met with the family to share the results and develop a treatment plan based on the FIRST responses. Once the FIRSTs were collected, the agency deleted all identifying information and released the anonymous inventories to the research staff.

Participants in the comparison group were invited via an e-mail advertisement to complete the FIRST. The e-mail was distributed to nearly 300 employees and recruitment resulted in a 22% return rate. While all employees received the e-mail, only those who had a child living in their household and were not receiving mental health services were eligible. The e-mail provided participants with a link to a website containing the self-report version of the FIRST. Once they anonymously completed the inventory online, their data was electronically sent to the researchers. Each employee who completed the survey was mailed a $10 gift certificate to a local toy store. The study was approved by the human subjects Internal Review Board at the university where the authors conducted the research.

Measures

Family Inventory of Resources and Stressors (FIRST)

The FIRST is a 104-item broad-based assessment measure of family functioning intended to help clinicians and family members evaluate a family’s resources and stressors for the purpose of goal-oriented intervention (Lawrence 2006). The FIRST was based on a competence model which proposes that people seeking help are often more competent than they realize and their problems are often the result of “adaptive attempts gone awry” (Waters and Lawrence 1993, p. 9). This approach suggests that intervention works best when clinicians understand and help families build on their competence and the healthy intentions that underlay unhealthy behavior as they address problematic behavior. It is similar to the risk and resilience model of intervention, which promotes the importance of increasing protective factors while simultaneously reducing risk factors (McCubbin et al. 1998; Walsh 1996, 2003).

The FIRST was specifically designed to assist low-income, multi-problem families in recognizing their areas of resiliency while simultaneously identifying their needs. It enables clinicians to develop family-driven treatment plans that capitalize on the families’ perceived strengths in order to effectively address their problems. It is a family-centered rather than child-centered assessment tool as items ask about family functioning rather than just individual functioning with two exceptions. There is a child-specific scale that measures each child’s academic needs and performance and an adult-specific scale that measures each adult’s employment needs and performance.

The FIRST was selected for use with the families in this study because of its unique capability to assess low-income, multi-problem family strengths and needs. The Resource factor on the FIRST is comprised of 70 items divided into seven scales: basic life, personal, parent–child, social support, partner relationship, child education, and adult employment. However, the Partner Relationship scale was not used in analyses because it was developed mid-way through the study and was not available for all subjects. The Stressor factor on the FIRST is comprised of 34 items divided into two scales: Emotional and Family/Community. Item responses on the FIRST are based on a four-point Likert scale, with higher scores indicating higher levels of resources on the Resource scales and higher levels of stress on the Stressor scales. Studies of the FIRST’s psychometric properties have found it to have adequate internal consistency with eight of the nine original scales exceeding Cronbach alpha’s of .70 (Lawrence 2006; Vanderwielen 2003). In addition, confirmatory factor analysis supported the proposed factor structure of the FIRST on the domain level (Vanderwielen) and preliminary discriminant factor analysis revealed that the FISRT adequately predicted membership in non-clinical, at-risk, and clinical groups (Heller 2005).

Resources

The six resource scales of the FIRST used in this study assess areas of family life that are thought to be important predictors of positive functioning. Items for each scale are further divided into subscales to ease translation into treatment planning. Basic Life Resources consists of three subscales that assess the adequacy and stability of the family’s housing, its ability to provide for financial needs, and the quality of its health care. Personal Resources evaluates how well the family functions on an emotional level (e.g., hopefulness, assertiveness, and relatedness), as well as its ability to generate and execute solutions to daily problems. Parent–Child Resources assesses parents’ ability to provide structure, enforce rules, and use skills to get their children to behave. In addition, it evaluates the level of cooperation between parents and children and the quality of their time together as a family. Social Support Resources examines the quantity and quality of support adults and children feel from others outside the family (e.g., relatives, friends, community organizations, and babysitters). Child Education Resources evaluates the degree to which each child’s needs are being met in school, his or her level of academic performance, and the family factors that contribute to academic well-being (e.g., parental cooperation with the school and good attendance records). Adult Employment Resources assesses each adult’s employment status, job satisfaction, benefits, and perception of job skills.

Stressors

The two stressor scales of the FIRST examine areas of family life determined to be risk factors that impede functioning. As occurred with the resource scales, the stressor items are further divided into subscales. The Emotional Stressors scale identifies the presence and severity of emotional and psychologic distress exhibited by any member of the family in the 6 months preceding the assessment. Its subscales are depression, anxiety, substance abuse, physical safety, and sexual safety. The Family/Community Stressors scale is further divided into two subscales (Family and Community) that identify risks within the family system (e.g., family conflict and problems with work, money, and childcare) and risks within the family’s community (e.g., the level of neighborhood safety, difficulties related to racism or cultural barriers, and problems with the law).

While the items in the Family/Community Scale have a lower internal consistency than is typically recommended (Cronbach 1951), the scale continues to be included because the degree of internal consistency may not be the best measure of the scales’ value. The 8 items on this subscale were designed to assess the presence and severity of a broad range of different types of community stress rather than equal measures of the same construct. The presence of any risks as well as the additive effect of more risks is a useful indicator of a family’s community stress.

Results

Internal consistencies of the factors, scales, and, in some cases, the subscales were calculated for this sample using Cronbach’s alpha (Table 2). Both factors and all scales and subscales had acceptable inter-item consistency (α > .70) with the exception of four Stressor subscale scores (i.e., Depression, Physical Safety, Family Stressors, and Community Stressors). Researchers proceeded with analyses using these subscales given that the items in these subscales were designed to assess the presence and severity of particular stressors and were not meant to be equal measures of a particular construct. Composite scores for the factors and scales were tabulated by calculating means and standard deviations (Table 2) as recommended in the FIRST manual (Lawrence 2005). For the purpose of this study, respondents’ highest level of education was captured in one of six categories ranging from some/all grade school to post-college, and marital status was categorized as not married or married/partnered. While most FIRST responses pertained to the family as a whole, child race was determined by the identified child’s racial identity, categorized as African-American, Caucasian, Latino/Hispanic, Asian/Pacific Islander, or Biracial. Household income was classified by 18 non-equal categories based on US poverty guidelines and ranged from $0-4,999 to $100,000 or above.

Table 2 Results of one-way ANCOVAa analyses for clinical and comparison groups
Full size table

Analysis 1: Differences Between Families With and Without a Child with SED on Perceived Resources and Stressors

It was hypothesized that families of children with SEDs would report fewer overall resources and more overall stressors than those of children without SEDs. An exploratory analysis of the differences between the two groups was conducted on the eight scales of the FIRST in order to better understand the specific strengths and needs of families of children with SEDs. In order to ensure that these results were not skewed as a result of demographic differences between the two groups, univariate analyses were conducted that controlled for sampling variables. Of the four variables considered (i.e., respondent’s education level and marital status, the identified child’s race, and household income), only income and education level differed significantly across the two groups and were entered as covariates. The participants who did not report their education level or household income were excluded from these analyses. Group differences were assessed through a series of one-way ANCOVAs (Table 2). Levene’s test of homogeneity of variance indicated that the error variance of each dependent variable was equal across groups (all p’s > .05). Effect sizes using Cohen’s d were evaluated by the standard criteria (0.20 = small, 0.50 = medium, 0.80 = large).

On overall resource scores, no significant differences were found between the two groups of families F (3, 97) = 0.00, p > .05. Of the six resource scales, Child Education Resources was the only scale on which families of children with SEDs reported significantly lower resource scores than their counterparts F (3,82) = 11.98, p < .01, and Cohen’s d revealed a large effect size (0.91). Surprisingly, the families of children with SEDs reported significantly higher resources than their counterparts on the Adult Employment scale F (3,84) = 3.90, p < .05, and Cohen’s d revealed a small effect size (0.46).

As predicted, the families of children with SEDs had higher overall stressor scores than those without children with SEDs, F (3, 97) = 9.73, p < .01. Cohen’s d revealed a medium to large effect size (0.78). In order to test whether this pattern also held for the scales and subscales, further analyses were conducted on these scores. The families of children with SEDs had higher scores on the Emotional Stressors scale F (3, 97) = 8.86, p < .01, d = 0.73 and on the majority of its subscales, including Depression F (3, 95) = 7.54, p < .01, d = 0.71, Anxiety F (3, 96) = 7.94, p < .01, d = 0.67, and Physical Safety F (3, 97) = 5.45, p < .05, d = 0.58. They also reported more stressors than the comparison group on the Family/Community Stressors scale F (3, 97) = 5.81, p < .05, d = 0.61. Of this scale’s two subscales, Family Stressors and Community Stressors, they only reported more stressors on the Family Stressor subscale F (3, 97) = 9.50, p < .01, d = 0.74.

Given the assumption that children in residential care exhibit more severe symptoms than those in day-treatment programs, it was hypothesized that families of children receiving residential treatment would report higher levels of stress and lower levels of resources than those receiving day treatment. However, when the clinical group was divided into day and residential clinical groups, no differences were found between groups on reports of overall stressors F (1, 76) = 2.63, p > .05 or overall resources F (1, 76) = 0.42, p > .05.

Analysis 2: Social Support and Parent–Child Relationships as Promotive Factors

To test the hypothesis that families of children with SEDs who reported greater social support and parent–child resources would also report fewer stressors, a correlation matrix was obtained (Table 3). As predicted, it was found that these families’ Social Support and Parent–Child Resource scores were significantly negatively correlated with their overall Stressor and Stressor scale scores (i.e., Emotional and Family/Community) with one exception: the correlation between Emotional Stressors and Social Support Resources was negative but not significant.

Table 3 Results of one-way ANCOVAa analyses for day and residential treatment groups
Full size table

Discussion

The results from this study suggest that families of children with SEDs perceive themselves as having more emotional and family stressors or risk factors than those in the comparison group but similar levels of overall resiliency. While it is not surprising that these families report more overall stressors, it is important to note the pattern of stressors they experience in an initial attempt to identify what might be normative risk factors for families of children with SEDs. They reported greater risk related to their own and other family members’ emotional and physical well-being. They were more likely to perceive a family member to be depressed and anxious than the families of children without SEDs and more likely to report that they or someone in their family had been a victim of physical or sexual abuse over the past 6 months. This result was expected given that children with SEDs are, by definition, experiencing emotional distress (e.g., depression, anxiety, post-traumatic stress). Furthermore, caregivers of children with SEDs have also been shown to be at greater risk of emotional strain that can take the form of depression and anxiety (Smith et al. 2001).

In addition to emotional stress, families of children with SEDs were also more likely to report risk related to the family’s well-being. Although they reported greater stress on Family/Community Stressors, when this scale was separated into its subscales of Family Stressors and Community Stressors, Family Stressors was the only subscale that was significantly different between groups. Stressors in this subscale included internal family factors such as difficulty related to a family member’s physical or mental illness, fights between family members, and concern about a family member getting into trouble. Again, given the fact that the identified child is suffering from a mental illness, it is not surprising that these families would report greater concern about their family members’ well-being and experience more conflict in their households as a result of their attempts to manage the stress of caring for a mentally ill child (Crawford and Manassis 2001).

Given that the families in this study came from similar neighborhoods in a large city and, as such, were presumably dealing with a comparable level of community-related stress, it follows that risk factors in the Community subscale were not significantly different between the two groups. Despite the intuitive nature of this finding, it is an important reminder for professionals who may inadvertently attribute some community risk factors (e.g., having friends who use drugs or get in trouble or difficulties because of racism) to having a child with an SED rather than to the influence of living in a particular neighborhood. However, this finding should be interpreted with caution given that the Community subscale did not demonstrate adequate internal consistency (Table 2).

In contrast to stressors, the results from the Resource scales were more surprising and lend greater support to the importance of conducting in-depth and comprehensive assessments of these families’ perceived strengths and resiliency factors. In general, the two groups reported similar levels of overall resources. When this factor was divided into its scales, the same pattern of results emerged for each scale with the exception of Child Education Resources. The identified children in the comparison group had significantly higher scores on this scale than the children with SEDs. This result was expected based on the fact that the children in the clinical group had been taken out of public education to receive special academic and psychologic support from day-treatment and residential programs.

The other resource scales add to an interesting pattern of perceived resiliency for families of children with SEDs. Basic life and financial resources did not differ between groups, and the clinical group actually rated their adult employment resources higher after controlling for income and education level. Parent–child and personal resources were similar across groups. This finding lends support to the idea that, despite the strain that occurs between family members when one member has a mental illness, these families feel they have significant personal resiliency, and, perhaps, even greater reserves of resilience than treatment providers might assume given their level of stressors. Although treatment professionals may perceive problems in family members’ ability to relate to and connect with each other, the members themselves feel that they are as competent as families who are dealing with less stress overall.

The examination of the risk and resilience patterns of families of a child with an SED in day versus residential treatment revealed no differences. The families of children with SEDs in residential care did not report fewer resources or greater risk factors than those of children in day treatment as expected. While there may not be any significant differences in normative risk and resilience patterns for these two populations, one would expect a relationship between the degree of assistance needed and stress level. However, it may be that parents of children receiving residential care actually experience a decrease in stress after enrolling their children in a facility that assumes full responsibility for monitoring and providing care for their children.

Finally, as expected, families in the clinical group who reported greater social support also reported less overall risk and less risk on the Family/Community Stressor scale. Families in the clinical group who reported greater parent–child resources (i.e., more confidence in their parenting skills), also reported less overall risk and less risk on both the Emotional and Family/Community Stressor scales. Similar to past research that considers social support and parenting competence to be resiliency factors for at-risk families (Beresford 1994; Murry et al. 2001; Walsh 1996), the current results support the conclusion that access to a network of supportive friends and family members and the ability of parents to provide structure and connection for their children is associated with less perceived risk across several scales.

On the other hand, the lack of a hypothesized correlation between Emotional Stressors and Social Support resources was surprising. On further reflection, this is likely related to the time at which the families in this study were assessed. Given that their FIRST responses were collected at placement intake, the families were concerned enough about their children’s functioning to place them in a more specialized school setting (i.e., day or residential treatment), and, thus, experiencing a level of emotional and academic distress, that had become more than their social support system could buffer.

Implications

This research offers a compelling rationale for pursuing more in-depth inquiries into the resiliency and risk factors of families during intake and treatment planning for children with SEDs. When these families are given the opportunity to assess their own family functioning, they report more risk than families of children without SEDs but a similar level of resiliency. For these families, elevated stress levels do not necessarily indicate diminished resources. In fact, they appear to have a pattern of resiliency similar to what is normative for families of children without SEDs. Given these results, a treatment provider’s knowledge of specific aspects of family functioning that the families perceive as a resource could help treatment providers better focus on using these areas of resilience to address areas of need. For example, if during the assessment a family reports significant arguments with one another, but also identifies that they are good at problem solving, the treatment provider can focus the treatment towards activating their problem-solving skills to defuse arguments. Knowledge of the families’ perceptions of their own resiliency enables treatment providers to engage them more positively as they develop strategies to address their risk factors.

Limitations

Although this study provides a preliminary understanding of what might be normative family resiliency and risk factors as perceived by families of children with SEDs, it has several limitations. One limitation is that the families of children with SEDs in this study were interviewed at placement intake and comprised a heterogeneous population whose children differed in age and illness severity. In order to identify truly normative patterns of risk and resiliency for families with children with SEDs, it would be important to repeat the study with more homogeneous SED populations and follow them over time. Practitioners’ knowledge of the typical trajectory for the development of risk and resiliency factors for these families over the course of their children’s treatment would enable the practitioners to develop prevention programs that capitalize on family strengths as well as address family risks. For example, it would be important to determine if parents of children with SEDs begin to perceive their children’s education as a resource once his or her educational needs are better met. Access to specific information about each child’s diagnosis would help to determine whether patterns of family resources and stressors differ based on the diagnosis or severity of the child’s illness. Finally, the families in this study were, in general, financially disadvantaged, which may contribute to higher stress levels than would be reported by more privileged families. In particular, families of children with SEDs and medium to high incomes may have a different pattern of risk and resiliency than those families who are also dealing with limited financial resources.

Another limitation of the study is that clinical and comparison families completed the FIRST using a different format. Although it is unlikely that families of children without SEDs who completed the survey anonymously on a computer responded defensively, it is possible that the families of children with SEDs who answered questions in the context of an intake meeting with a treatment provider may have exaggerated their strengths and underemphasized their stress levels. On the other hand, that these families reported significantly greater stress levels than those of children without SEDs suggests that they did not answer defensively and that the pattern of resources and stressors reported is valid.

Future Research

While the results of this study support the recommendations of professionals in the field who suggest that using a collaborative and strengths-based assessment approach with families of children with SEDs provides important treatment information, further research is needed to determine how best to use this information. For example, it would be important to examine how treatment professionals’ ratings of family functioning on an assessment tool such as the FIRST compare to ratings by parents. Do professionals as a group report fewer resiliency factors, and is this because they are more objective reporters or because they have been trained to focus on deficits rather than strengths? More importantly, which perspective serves as a better guide for treatment planning? Are parents who are exposed to a collaborative assessment approach more likely to be active participants in treatment (e.g., come to appointments, follow through on reinforcement schedules at home) than those who are not? In a similar vein, do children of families who receive a comprehensive assessment and collaborative treatment have better outcomes (e.g., shorter treatment stays, improved academic performance, and less conflict at home) than those who do not? The answers to these questions would lend further support to the assertion that, for families of children with SEDs, willingness to champion the families’ own perceptions of their risk and resiliency factors can lead to greater collaboration and treatment gains for children with SEDs (Table 4).

Table 4 Clinical group: correlation matrix of social support parent–child with stressors
Full size table