Family involvement in planning and choosing services has become a key intervention concept in developmental disability services. This study (N = 547) modeled patterns of family decision making and assessed benefits to persons with developmental disabilities (DDs) and their family members. A latent profile analysis identified 4 classes that were highly involved in decision making (n = 118), involved only in planning (n = 166), involved only in financial decisions (n = 75), and uninvolved (n = 188). Multiple regression analysis indicated that consumers with DD whose family members were highly involved received more services than consumers in other families. A multivariate analysis of covariance indicated that the family members in the highly involved and planning classes experienced more family member satisfaction than others. Findings have implications for practice.
Families of adults with developmental disabilities (DDs) are often a critical source of care and support, as evidenced by the fact that 61% of individuals with DD live with their family (Fujiura, 1998). Thus, understanding the role the family plays in planning and accessing services for their member with DD is a key to services research for this population. Research on the experience of families providing care for a member with DD demonstrates wide-ranging impacts. Several studies have reported that family members experience fulfillment and joy when they provide care to persons with DD (Hastings & Taunt, 2002; Scorgie & Sobsey, 2002; Stainton & Besser, 1998). Others document family stress and burden (Hayden & Goldman, 1996), particularly when the person with DD has high caregiving needs and the family lacks support (Caldwell & Heller, 2003; Meyers & Marcenko, 1989; Seltzer & Krauss, 1989). Another study has noted the economic implications of caring for an adult with DD, particularly for mothers who have to give up work opportunities (Parish, Seltzer, Greenberg, & Floyd, 2004).
Much of the DD literature has focused on increasing consumer choice and self-determination (Bambara, 2004; Halle, 1995; Neely-Barnes, 2005). Interventions such as person-centered planning put the consumer in the decision-making role (Kincaid & Fox, 2002; O'Brien & O'Brien, 2002). Researchers have reported an association between consumer direction and increased consumer satisfaction (Kim, White, & Fox, 2006; Kroll, Beatty, & Bingham, 2003). Although much of the DD literature places an emphasis on consumer direction, family members continue to play an important role in the lives of persons with DD. Family members often know the person with DD best and can assist in making decisions that reflect the person's interests, goals, and dreams (Everson & Zhang, 2000; O'Brien, O'Brien, & Mount, 1997). Thus, we designed this study to examine the role of the family member in the decision-making process.
DD programs that support families have become increasingly empowerment based (Heller, Miller, & Hsieh, 1999). Empowerment-based practice assumes that families are competent and can take action to affect positive changes in their lives (Dunst, Trivette, & Deal, 1988). According to Dunst and Trivette (1993), to increase empowerment, professionals should place emphasis on family priorities, encourage family participation, use collaboration, provide families with information to make informed decisions, and accept and support family decisions. Based on empowerment theory, we expected to find that family members involved in decision making would get needed services and be more satisfied with the outcomes of the intervention.
Consistent with empowerment theory, researchers have found that the family caregiving experience is mitigated by the degree to which family members determine the support they receive (Hodges, Burwell, & Ortega, 1998; Law et al., 2003), the degree to which they possess strategies and competencies for seeking help (Lustig, 2002), and the family's coping style (Essex, Seltzer, & Krauss, 1999; Kim, Greenberg, Seltzer, & Krauss, 2003). In a study by Lustig (2002), families had better adjustment when they could reframe adverse events to make them less stressful and mobilize sources of help. Kim et al. (2003) found that families had better psychological well being when they were oriented to problem-focused strategies and had an increased sense of mastery and control over their environment.
A few studies have examined the benefits of family decision making. Consumer-directed, family support programs have been associated with increased satisfaction with services (Law et al., 2003), fewer unmet needs (Heller et al., 1999), greater caregiver self-efficacy (Heller et al., 1999), better access to health care (Caldwell, 2006), and more social activities (Caldwell, 2006). Likewise, in a study of residential placement, Schwartz (2005) found that parents who were involved in choosing the residential placement were more satisfied with the placement. In a study of respite care, Caldwell and Heller (2003) found that families who exercised more control over the hiring and management of their respite care workers were more satisfied with their services.
However, not all researchers have found a link between increased family decision making and increased family satisfaction. Other studies of respite care have noted that families tend to be satisfied with the services they receive regardless of whether they have control over the services and whether the services are meeting the family's need. These high levels of family satisfaction may be due to the scarcity of resources available to support families of persons with DD (Freedman, Griffiths, Krauss, & Seltzer, 1999; Ptacek et al., 1982).
Although prior research indicates there are benefits to family decision making, less is known about the patterns of family involvement in making decisions. Herman (1991, 1994) found that when a cash subsidy is available, parents will choose a wide range of services to meet the family's needs related to both the family income level and the type of disability. In a review of literature, Heller (2000) found that aging parents differed in their legal and financial planning for their child with a disability. Some parents did a great deal of planning and others did not plan. In a study of consumer-directed family support, Caldwell and Heller (2003) found that families differed in their level of control and management of services. Yet, we are not aware of any study that has specifically examined patterns of family decision making and the relationship to services received, family satisfaction, and family perception of case manager competence.
Based on the literature review, this study has three hypotheses. The first hypothesis is that there will be subgroups within the population of family caregivers who have differential patterns of involvement in decision making about services. Consistent with the findings of Herman (1991, 1994) and Caldwell and Heller (2003), we expect to find that some family members are highly involved in decision making and that others are uninvolved. The second hypothesis is that, consistent with the findings of Caldwell (2006), consumers receive more services when their family members are more involved in decision making. The final hypothesis is that families who are more involved in decision making will be more satisfied and perceive their case managers as more competent. This finding would support the literature that suggests a link between family decision making and increased family satisfaction (Caldwell & Heller, 2003; Law et al., 2003; Schwartz, 2005).
This study used data from Washington state's 2004 National Core Indicators (NCI) Adult Family Survey. NCI conducts surveys of consumers, family members, and providers annually, with the purpose of developing performance and outcome indicators that state developmental disability authorities can use to evaluate their service system. The NCI survey includes items that address (a) case management, (b) choice and control, (c) community connections, (d) family outcomes, (e) functioning and disability, (f) informal support, (g) information and planning, (h) satisfaction, and (i) services and supports. The 61 core indicators were developed by a task force consisting of staff of developmental disability authorities nationwide and were tested during Phase I of the NCI project (Human Services Research Institute, 2005). Eleven states and 1 local DD authority participated in the 2004 (Phase VI) adult family survey. Only the data from Washington state are used in the present study. In addition to the NCI survey items, the Washington state Division of Developmental Disabilities (WSDDD) added items to the survey on informal supports, case manager competence, and the level of supervision required by the person with DD. These items are analyzed in the present study.
The survey was reviewed and approved by the Washington state Institutional Review Board. The survey instrument was mailed to the parents along with a cover letter and self-addressed stamped envelope. Parents were asked to complete and return the survey. They were told that the WSDDD was interested in family members' perceptions of services. Family members received no compensation for completing the survey. Approximately 4 weeks after the survey was mailed, family members who had not responded were sent a follow-up letter. Informed consent was implied by the return of the survey.
Family members were selected for participation in the study if they met the following criteria: (a) They were caring for a person with a DD in the family home, (b) the person with DD was over the age of 18, and (c) the person with DD received at least one service (other than case management) from WSDDD. From the approximately 3,500 family members who met the criteria above, a total of 1,520 participants were randomly selected. If there was more than one person with DD living in the home, the family member was asked to respond regarding the person named in the mailing. A total of 580 family members returned the mailed survey, yielding a response rate of 38%. This is a typical rate for a mailed survey (Grinnell, 1997) and close to the NCI anticipated participation rate of 40% (Health Services Research Institute, 2005).
Of the 580 family members who completed the survey, 26 reported that the person with DD did not live at home and were removed from the study. Five family members who reported the person with a DD was less than 18 years of age were also removed. Finally, 2 surveys were removed because more than 50% of the data was missing. Data from the remaining 547 family members were included in this study. Gender of the consumer, χ2(1, N = 540) = 1.15, did not differ between respondents and nonrespondents. Race/ethnicity could not be compared because of differences in definitions of the variable between the WSDDD administrative data and the survey. Respondents differed from nonrespondents with respect to age of the consumer, t(546) = −2.84, p = .005. Consumers in nonrespondent families (M = 31.22) were on average 1 year older than those in the respondent families (M = 29.96).
Almost 91% of the family members reported that they were the parent of the person with DD. Other family members responding included siblings, aunts, grandmothers, and sisters-in-law. Most indicated that they were the primary caregiver for the person with DD (97%) and the legal guardian (63%). Ninety-five percent of the consumers had intellectual disabilities (IDs) and 27% had a communication disorder. See Table 1 for more demographic information.1
Respondents were asked whether the person with DD had mental retardation, with five possible response categories: no MR = 1, mild = 2, moderate = 3, severe = 4, and profound = 5. This variable will be referred to as intellectual disability (ID) throughout the remainder of the article. Family members were also asked whether the consumer had a diagnosis of autism, cerebral palsy, communication disorder, mental illness, physical disabilities, seizure disorder, and/or vision or hearing impairment. Diagnostic categories were coded as no = 0 and yes = 1.
Family members were asked to rate the level of support or supervision the person with DD needs using six categories. The categories ranged from “none—the person with DD can be left unattended” to “line of sight—a support person must be in the same room as the person with DD at all times.” Responses were: none = 1, remote = 2, monitoring = 3, close proximity = 4, onsite = 5, and line of sight = 6. The support needs item was reviewed by 40 members of the WSDDD administrative staff, who judged it to have good content validity. In addition, prior testing by WSDDD has indicated that this item has an interrater reliability of .72 and that it is a strong predictor of service use rates when other variables are controlled.
Seven items measured the family member's involvement in making decisions. Items included whether the family member (a) helps develop the service plan, (b) helps decide what is in the plan, (c) chooses agencies to work with family, (d) chooses support workers to work with family, (e) controls hiring and management of support workers, (f) knows how much money is spent by WSDDD, and (g) gets to decide how much money is spent by WSDDD. Items were coded as seldom or never = 0, sometimes = 1, and always or usually = 2.
Case manager contact
Family members were asked how frequently they had contact with their case manager in the past year. Responses to this open-ended question were recoded as none = 0, once = 1, more than once but less than quarterly = 2, quarterly to less than once a month = 3, monthly to less than once a week = 4, and more frequently than weekly = 5.
Family members were asked three questions about informal support. They were asked whether friends, extended family, or neighbors (a) assist with care, (b) are available to talk about problems, and (c) are available in an emergency. The three items were coded yes = 1 and no = 0 and were combined to create a measure of informal support, with a range of 0 to 3.
Family members were asked whether they received eight types of services from the WSDDD: (a) financial support, (b) in-home support, (c) respite care, (d) day–employment supports, (e) transportation, (f) therapy services, (g) community guide (a mentor who connects the family to informal supports), or (h) a personal agent (an assistant who negotiates the use of funding or finds ways to meet needs). Services were coded as yes = 1 and no = 0 and summed to create a measure of the number of services received, with a range of 0 to 8.
Perception of case manager competence
Thirteen variables were used to create a scale of family perception of case manager competence, with a range of 0 to 26. Family members were asked whether their case manager (a) returns phone calls promptly, (b) seems knowledgeable about WSDDD services, (c) seems knowledgeable about community services, (d) seems knowledgeable about the person's disability, (e) listens attentively, (f) understands the needs of the family, (g) understands the needs of the person with DD, (h) regularly asks how services are going, (i) regularly asks if needs have changed, (j) monitors the quality of services, and (k) supports the family members' suggestions. In addition, family members were asked (l) whether they feel they work as a team with the case manager and (m) whether they are satisfied overall with case management. All items were coded as seldom or never = 0, sometimes = 1, and always or usually = 2. The scale had a reliability of α = .96.
Satisfaction with services
Six items were used to measure satisfaction with services. Family members were asked (a) whether their family receives needed services and supports, (b) whether supports and services meet the family's needs, (c) whether supports and services are available when the family needs them, and (d) whether supports and services have made a positive difference. In addition, family members were asked (e) whether they were satisfied overall with services and (f) the family's perception of whether overall the person with DD was happy. All items were coded as seldom or never = 0, sometimes = 1, and always or usually = 2. Responses were combined to create a satisfaction with services scale, with a range of 0 to 12 and a reliability of α = .89.
The division of the satisfaction with services and perception of provider competence scale was supported by an exploratory factor analysis using a principal components extraction and varimax rotation. Loadings of the satisfaction items on Factor 1 ranged from .54 to .84, whereas loadings of the perception of case manager competence items on Factor 2 ranged from .66 to .84. The two factors explained 68% of the variance.
Data were analyzed using SPSS 13.0 (SPSS, 2004) and Mplus version 3.1 (Muthén & Muthén, 2004). SPSS Missing Value Analysis was used to analyze patterns of missing data and to impute values for continuous variables. This software package uses a maximum likelihood method based on Little and Rubin's (1987) work on expectation-maximization (EM) algorithms. Missing data were not imputed for categorical variables. Cases that had imputed values outside the variable range were recoded to either highest or lowest meaningful value as appropriate. Patterns of missing data ranged from 0% to 34.2% missing. Maximum likelihood estimation was the chosen method because prior research has demonstrated that the method will produce nonbiased estimates on both independent and dependent variables when the rate of missing values is as high as 80% (Schafer & Graham, 2002).
To identify variation in patterns of family decision making, a latent profile analysis (LPA) was conducted. As a form of mixture modeling, LPA is a person-centered approach that assesses whether there is an underlying typology that fits the data. This approach is useful when a variation in respondents is believed to be due to an unmeasured latent variable (McCutcheon, 1987). The LPA was completed using Mplus version 3.1 (Muthén & Muthén, 2004). Seven variables were used to model family decision making: (a) develops service plan, (b) decides what is in the plan, (c) chooses agencies–providers, (d) chooses support workers, (e) controls hiring–management of support workers, (f) knows amount of money spent by WSDDD, and (g) decides amount of money spent by WSDDD. The number of classes was determined through an iterative process that involved specifying the number of classes and then using indicators of model fit to interpret the results. The Bayesian information criterion (BIC) and the Lo-Mendell-Rubin adjusted likelihood ratio test (LRT) were used to determine the fit of the model. The Lo-Mendell-Rubin test compares the estimated model to the model with one less class. Significant findings indicate that the fit of the estimated model is better. In addition, a random set of multiple start values was tested. Using multiple starting points in the analysis increases the likelihood that one will find the best solution for the model instead of a local minimum (Muthén & Muthén, 2004).
After class memberships were estimated, characteristics of persons with DD and of family members in each class were examined. Chi-square tests of independence were used to compare classes on gender of person with DD, diagnoses, race, and type of caregiver. To avoid the problem of inflated alpha levels when using multiple statistical tests, a multivariate analysis of variance (MANOVA) was used to examine the association between class membership and (a) age of person with DD, (b) age of family member, (c) family income, (d) frequency of case manager contact, (e) informal support, (f) level of mental retardation, (g) level of support needs, (h) number of services received, (i) overall satisfaction, and (j) perception of case manager competence. Box's M test of homogeneity of covariance matrices was significant (Box's M = 213.34, p = .018), thus the Pillai's trace statistic is reported. A Scheffé test was used for post hoc comparisons.
A multiple regression analysis was used to examine the influence of class on the number of services' controlling the following variables (because these variables are also associated with receipt of services): (a) age of family member, (b) age of person with DD, (c) gender of person with DD, (d) family income, (e) frequency of case manager contact, (f) informal support, (g) level of mental retardation, and (h) support needs of person with DD. Dummy variables were created to represent class membership. All variables were entered in one block.
A multivariate analysis of covariance (MANCOVA) was used to examine the influence of class membership on satisfaction with case management and services. The influences of (a) age of family member, (b) age of person with DD, (c) family income, (d) frequency of case manager contact, (e) informal support, (f) number of services received, and (g) support needs of the person with DD were controlled. Box's M test indicated that equal variances could not be assumed; thus, Pillai's trace test was the multivariate test statistic used. Using a Bonferroni correction, a series of post hoc tests were conducted to identify class differences.
Means and standard deviations of continuous variables are reported in Table 2. Using LPA, we assessed for variations in the sample using the indicators of family involvement in decision making. Comparisons among the Bayesian information criterion (BIC) values for the one-class model (BIC = 9,978.49), the two-class model (BIC = 8,712.09), the three-class model (BIC = 8,480.19), and the four-class model (BIC = 7,892.80) indicated that the four-class model had the lowest BIC value and fit the data best. In addition, the Lo-Mendell-Rubin likelihood ratio test for the three-class model versus the four-class model was significant, indicating that the four-class model was superior. Although the five-class model had an even lower BIC (7,750.77), the Lo-Mendell-Rubin likelihood ratio test for the four-class model versus the five-class model was nonsignificant (349.11, p = .07). Thus, we decided that the four-class model was the best fit. Average class probabilities for the four classes were .96 for the first class, .94 for the second class, .97 for the third class, and .97 for the fourth class, indicating that the classes were cleanly defined.
Percentages for the family decision-making indicators are reported in Table 3. The first class (n = 118) had high involvement across all items. Thus, this class was named the highly involved class. The second class (n = 166) had high involvement in annual planning and selecting agencies but low involvement in day-to-day management. This class was named the planning class. A third class (n = 75) had low involvement in planning and choosing support workers but high involvement in choosing agencies and knowing and deciding how much money is spent. This class was named the financial class. The final and largest class (n = 188) had low involvement across all items. This class was named the uninvolved class.
Classes did not differ on the basis of race of the person with DD, type of caregiver relationship, or diagnoses. Classes differed on gender of the person with DD, with the planning class having a much higher percentage of men with DD (67%) than the financial class (51%). However, independent-samples t tests did not reveal significant relationships between gender and the dependent variables in the study: number of services, t(538) = −0.128, p = .90; case manager competence, t(538) = 0.30, p = .20; and satisfaction with services, t(538) = −1.09, p = .28.
We used a MANOVA to assess the relationships between class and (a) age of the person with DD, (b) age of family member, (c) income, (d) level of ID, (e) level of support, (f) case manager contact, (g) informal support, (h) number of services, (i) case manager competence, and (j) satisfaction with services, F(10, 533) = 2,889.75, p < .001, Pillai's trace = .983. Univariate F tests and post hoc tests were completed to examine the relationship between class and each variable.
The MANOVA indicated that younger family members were more involved in decision making. Higher income families were more involved specifically in financial decisions. Family members were more involved in decision making when the person had more severe IDs and higher support needs. Families involved in planning received more informal support. Families were more involved in decision making when they had more case manager contact. The results are reported in Table 2.
We completed a multiple regression analysis (Table 4) to examine the relationship between class and the number of services received. Because the MANOVA indicated significant relationships between class and level of mental retardation, level of support needs, age of the person with DD, age of family member, informal support, and case manager contact, these variables were controlled in the multiple regression. The analysis revealed that families received more services when persons with DD were younger and had less severe mental retardation. However, it should be noted that the bivariate correlation between level of MR and number of services received was nonsignificant. In addition, families received more services when they had greater need for support, lower incomes, and more frequent case manager contact. Last, the planning, financial, and uninvolved classes received significantly fewer services than the highly involved class. In other words, the families that were highly involved in decision making received more services even when other variables related to receipt of services were taken into account.
A MANCOVA was undertaken to assess the influence of class membership on perception of case manager competency and satisfaction with services. Age of the person with DD, F(1, 534) = 8.67, p =. 003; case manager contact, F(1, 534) = 11.82, p = .001; informal support, F(1, 534) = 35.11, p < .001; and number of services received, F(1, 534) = 7.10, p = .008, significantly adjusted the relationship between class and case manager competency. Family members of older persons with DD who received more services perceived their case managers as more competent. In addition, family members who had more case manager contact and more informal support perceived their case manager as more competent. Age of the person with DD, F(1, 534) = 5.86, p = .016; level of mental retardation, F(1, 534) = 4.36, p = .037; informal support, F(1, 534) = 36.03, p < .001; and number of services received, F(1, 534) = 34.98, p < .001, significantly adjusted the relationship between class and overall satisfaction with services. Family members of older persons with DD who had more severe mental retardation were more satisfied. Family members who received more services and more informal support were also more satisfied.
Pillai's trace indicated significant differences in perception of case manager competence and satisfaction with services with respect to class (Pillai's trace = .10), F(3, 543) = 9.19, p < .001, partial η2 = .049. Univariate ANOVA results indicated that classes differed significantly in both perception of case manager competence and satisfaction with services (see Table 5). A comparison of adjusted means indicated that the highly involved and planning classes viewed case managers more positively than the financial and uninvolved classes. In other words, the families that were highly involved and involved in planning perceived their case managers as more competent than the financially involved and uninvolved families. The uninvolved group was significantly less satisfied with services than all other groups (see Table 5).
Discussion and Implications
Although we expected to find patterns of high and low involvement in decision making, it was an interesting and surprising finding that some families only had high involvement in one of two specific areas: planning and financial decisions. From this study, it is not possible to determine what processes led parents to choose these types of involvement. Additional research could provide insight into whether family members are aware of the various opportunities for decision making and why family members choose particular types of involvement.
The finding that family decision making did not differ on the basis of race or type of family member is surprising because prior researchers have found differences by race and type of family member for variables such as burden, stress, and impact on the family of disability (Dykens, 2005; Neely-Barnes & Marcenko, 2004). It may be that the sample in this study contained too few families of color and siblings to detect differences along these dimensions, or it may indicate that there is truly no difference in family decision making related to these variables.
Results of this study indicated that high family involvement in decision making is associated with receiving more services, increased family member satisfaction, and increased family member perception of provider competence. These findings have important implications in that they support the current developmental disability policy movement toward increasing family decision making. Current service delivery models such as person-centered and family-centered planning movements put either the individual with the disability or the family member in the central role of determining what supports are needed and who should provide them (Caldwell, 2006; O'Brien & O'Brien, 2002). The results of this study provide support for both continuing with these current models and creating opportunities for more families to be actively involved in decision making.
It is important to note that the direction of the relationships among family decision making, number of services received, and family satisfaction cannot be determined by this study because all data were collected at one time. This study cannot rule out the possibility that families who were more satisfied were choosing to be more involved in decision making. Despite this limitation, this study did indicate that family decision making, family satisfaction, and receipt of services are positively related.
Results of this study indicated that family members involved only in annual planning and selection of provider agencies perceive greater provider competence and experience more satisfaction than the families with no involvement. In addition, these families received more informal support than family members in other classes. These findings are consistent with prior research confirming that families who participate in problem solving and managing aspects of a stressful situation have more positive outcomes (Essex, Seltzer, & Krauss, 1999; Kim et al., 2003). This finding also has an important implication for DD practitioners. For family members who do not have input into the day-to-day aspects of service delivery, results suggested that even limited involvement can have benefits for families. This increased involvement is positively related to family satisfaction and better family perception of services provided. DD practitioners might seek to engage families on a limited or ad hoc basis around major decisions as a mechanism for promoting greater family satisfaction.
Although the findings of this study support the policy movement toward giving families more control over planning and provider selection, it provides limited support for the current policy movements toward giving family members more control over how money is spent (Ashbaugh, 2002). Financially involved families did not receive as many services as the highly involved families, nor did they perceive as much provider competence as the highly involved or planning-involved families. Yet, financially involved families were more satisfied with services than families without involvement. The members of this financially involved class had the highest income of any group, significantly higher than all other classes. Thus, this study also raises an important question as to whether service models that encourage family control over how money is spent are more easily accessed by higher income families.
The finding that family members in the uninvolved group were older than family members in the other three groups may reflect changes in service models. DD services have moved over the last 40 years from a care model that emphasized professional intervention to a support model that emphasizes choice (Mary, 1998). Older family members, who have been accustomed to having professionals make decisions, may have difficulty adjusting to the newer service models. In addition, this finding is consistent with prior research that has indicated that older families use fewer services and have fewer service system expectations (Hayden & Heller, 1997).
The uninvolved group was also the largest, constituting 34% of the family members in this study. This group received fewer services, received less informal support, and was less satisfied than the other groups. However, 27% of family members in this group reported that the consumer needed at least onsite supervision and another 56% reported that the consumer needed supervision at least once a day. These findings highlight the need to pay close attention to older families who do not have input into the care of their family member. This class may represent families who are at risk for burnout or crisis, which can lead to out-of-home placement. Thus, it is particularly important for practitioners to consider the use of long-term planning models with these families (Heller, 2000; Heller, Miller, Hsieh, & Sterns, 2000).
Limitations and Future Directions
There are several limitations to this study that should be noted. First, this study only addressed the decision making of family members who lived with the person with DD. Future research could examine the decision making of family members who do not reside with the consumer. NCI data could be used again to complete this follow-up study. Second, the study surveyed only family members. It did not directly survey the consumers themselves. Prior research has indicated that responses by proxies may differ from responses given by the consumers themselves (Stancliffe, 1995, 2000). Some questions on the survey (e.g., questions about whether the consumer is happy) might have been better answered by the consumer, and some of the consumers in this study would have been able to answer questions.
Other limitations to this study include that we used cross-sectional data. Because all data were collected at one time, directional relationships between the variables cannot be confirmed. A future longitudinal study on family decision making is needed. In addition, there was a low response rate, due to the use of a mailed survey. We cannot know how the 62% of families who did not participate differed from the 38% who did. A future study that involves telephone or face-to-face interviewing is needed to capture the families who do not respond to mailed surveys.
Another limitation of the study was the limited number of variables available. The survey did not include the family member's education nor did it ask the family member's gender. Prior studies have noted differences in family members related to gender (Essex et al., 1999; Orsmond & Seltzer, 2000). There were also a limited number of dependent variables available in the data set. The only dependent variable related to consumers that could be analyzed was the number of services received. This was a weak indicator of consumer benefit that was related to consumer age and level of intellectual disability (although these variables were controlled in the analysis). Future research could examine the relationship between family decision making and other indicators of consumer benefits, such as the consumer's quality of life or self-determination. The only family dependent variables available for analysis were overall satisfaction and perception of case manager competence. Prior research has noted that family members tend to be highly satisfied with services even when those services are inadequate (Freedman et al., 1999; Ptacek et al., 1982). Future research could examine the relationship of family decision making to other variables such as family quality of life and family member's well being.
In conclusion, the present analysis adds to researchers' understanding of the differences in families in their involvement in choosing and planning developmental disability services. Family decision making has been considered important by service models that are based in empowerment theory, yet researchers have known little about the relationship between family decision making and other variables. This study indicates that family decision making is associated with increased family satisfaction and perception of provider competence, which has important implications for state DD agencies and other providers of DD services.
The descriptive statistics reported in this study may differ from those reported by WSDDD to the Human Services Research Institute due to differences in inclusion criteria between the two studies.
Authors: Susan Neely-Barnes, PhD (email@example.com), Assistant Professor, College of Social Work, University of Tennessee, 711 Jefferson Ave., Room 607W, Memphis, TN 38163. J. Carolyn Graff, PhD, Associate Professor, College of Nursing, University of Tennessee Health Science Center, 711 Jefferson Ave., Memphis, TN 38163. Maureen Marcenko, PhD, Associate Professor, School of Social Work, University of Washington, Seattle, WA 98105. Lisa Weber, PhD, Senior Research and Data Manager, Washington State Division of Developmental Disabilities, Olympia, WA 98504-5310