Although mothers of children with intellectual and developmental disabilities (IDD) experience high levels of stress and schools constitute an important resource, the relation remains unknown between maternal stress and educational services. Responding to a national, web-based survey, 965 mothers of students with disabilities completed a 163-item questionnaire about parent stress. We examined which child, parent, and parent–school characteristics correlated with maternal stress. Mothers with lower stress levels reported better parent–school relationships and low levels of parent advocacy. However, lower stress levels were predominantly shown by mothers with good-to-excellent parent–school relationships (vs. poor-to-fair partnerships) and who engaged in virtually no (vs. any) advocacy activities. Lower maternal stress levels were also noted when children had fewer behavior problems, Down syndrome, and did not have autism. Less stress was also reported by mothers who had not enacted procedural safeguards, were minorities, and rated themselves lower on neuroticism and were more extroverted, dependable, and open to new experiences. This study has important implications for practitioners and researchers.
Compared to mothers of children without disabilities, mothers of children with disabilities experience increased levels of stress (Dyson, 1991; Song & Singer, 2006). Such high stress levels can lead these mothers to decreased physical health (Miodrag & Hodapp, 2011) and increased risks of depression (Brehaut et al., 2004) and fatigue (Lach et al., 2009). To effectively address the stress of mothers of children with developmental disabilities, a better understanding is necessary of the correlates of maternal stress.
Perhaps the most popular framework for understanding maternal stress is the ABCX model (Lazarus & Folkman, 1984; McCubbin & Patterson, 1983) and its subsequent disability-related adaptation, the Double ABCX model (Minnes, 1988). Within this model, the stress of parenting a child with disabilities is related to three main variables: child characteristics, perceptions of the child by the family, and the family's internal and external resources. To date, most research about maternal stress has focused on child characteristics. Parents of children with (vs. without) maladaptive behaviors report increased stress (Baker, Blacher, Crnic, & Edelbrock, 2002); indeed, maladaptive behaviors may constitute one of the best predictors of parent psychological health (Abbeduto et al., 2004). In addition, compared to mothers of children with other disabilities, mothers of children with Down syndrome may experience lower stress levels (Hodapp, 2007), and those of children with autism higher levels (Hayes & Watson, 2013).
Other research has focused on aspects of family resources and their relation to maternal stress. Parents who rely on coping strategies that are problem-focused (vs. emotion-focused) have significantly decreased parental stress (Sloper & Turner, 1993). Such resources may also partially relate to cultural differences. Thus, compared to groups of non-Hispanic White families, Hispanic families experience increased depression (Magaña, Seltzer, & Krauss, 2004), while African American families experience less depression (Magaña, 2004).
Although child characteristics and certain familial characteristics and resources have received research attention, a major gap concerns the effects of parent–school relationships on parental stress. Potentially, schools should be among the most important external resources experienced by families, and strong parent–school partnerships might lessen stress among mothers of children with disabilities.
In this vein, prior work paints a generally negative picture of these relationships. Specifically, parents are often uncomfortable with the entire special education process. They feel overwhelmed with the jargon (Hammond, Ingalls, & Trussell, 2008), marginalized by professionals (Turnbull, Turnbull, Erwin, Soodak, & Shogren, 2011), and disrespected more generally (Wang, Mannan, Poston, Turnbull, & Summers, 2004). At the same time, however, certain factors may help. When parents are satisfied with school services and experience quality family–school partnerships, they may enjoy a better family quality of life (Summers et al., 2007). Conversely, poor family–school partnerships may elicit increased stress. In the extreme form of poor parent–school relationships—when parents file for due process or mediation—they encounter a frustrating process of excessive paperwork, technical detail and jargon, and the high financial costs of retaining an attorney (Schrag & Schrag, 2004).
Other aspects of family–school partnerships may also relate to maternal stress. Given that parents generally want frequent and high-quality parent–school communication (Angell, Stoner, & Shelden, 2009; Esquivel, Ryan, & Bonner, 2008), those parents who have frequent communication with the school might experience less stress. Conversely, maternal stress may be negatively related to parent advocacy activities and knowledge. Many parents of children with disabilities spend time and resources learning their special education rights and advocating for their children (Fish, 2008). Although such efforts may involve adversarial struggle and lead to increased stress (Wang et al., 2004), few studies relate maternal stress to parental advocacy. Finally, the educational placement of the child may relate to maternal stress. Ryndak and Downing (1996) examined 13 parents of students with disabilities to understand their feelings toward placement in regular education versus self-contained classrooms. Families of students in self-contained classrooms felt that their students received repetitive, meaningless, and nonfunctional work. Such studies, then, highlight aspects of the educational system that may relate to stress for mothers of children with disabilities.
Though schools may be important in lessening maternal stress, few studies systematically relate maternal stress to educational partnerships. In this study, we examined which child, parent, and parent–school characteristics correlate with maternal stress. Beyond relating stress levels to educational services, we also tried to identify which aspects of such services might be most important, including the quality of the parent–school partnership, amount of contact with the school, or parents' feelings that they did not need to assert their due process protections. We predicted that increased maternal stress might occur in instances in which there were poorer family–school partnerships, little parent–school communication, when parents enacted procedural safeguards and engaged in increased parent advocacy, and when their offspring were educated in more restrictive placements. Given the deleterious consequences of high levels of chronic stress on mothers and the potential of better parental relationships with the schools to lessen such stress levels, our goal was to determine whether school characteristics might be tied to differing levels of maternal stress.
A total of 965 mothers of students with disabilities participated in this study. Ranging in age from 24 to 67 years, respondents averaged 43.34 (SD = 7.29) years. Over two-thirds of respondents had received their bachelor's degrees, were married, and were White. Children of these respondents were predominantly male (69.5%) and averaged 10.89 (SD = 4.39) years (range from 3 to 22 years). Students had a variety of disabilities including autism, developmental delays, Down syndrome, learning disabilities, and intellectual disabilities. Except for Delaware, Hawaii, and Wyoming, 47 of the 50 states were represented among this study's participants (see Table 1).
To attain a diverse national sample, participants were recruited in multiple ways. First, e-mails and flyers were sent to local, state, and national parent support groups, the Association of University Centers in Disabilities (AUCD), and all 106 parent training and information centers (PTIs) throughout the United States. Next, e-mails were sent to each agency listed in Wrightslaw Yellow Pages for Kids (http://www.yellowpagesforkids.com/help/az.htm). From this website, over 7,843 agencies were contacted. These agencies included: local parent support groups (436); nonprofit disability agencies (1,271), including chapters of The Arc; schools and educational agencies (1,386); attorney, advocacy, and legal agencies (1,725); and private service providers (3,025).
In collaboration with other disability researchers and community advocates, the questionnaire was developed and revised. The lead author first drafted the survey by examining the literature about maternal stress and family–school partnerships. This survey, hereafter referred to as the parent–school collaboration survey, included both established scales as well as novel questions. Before piloting this survey with five parents of students with disabilities we shared and received feedback about the survey from 10 professors, parents of individuals with disabilities, students, and community advocates. Upon revision, the Institutional Review Board (IRB) approved the study. The questionnaire was then put onto a secure website of the University. As survey responses accumulated, they were stored on the REDCap program (Harris et al., 2009) and were downloaded periodically to guard against computer malfunctions. The study was posted on the Internet from December 2010 until June 2011.
Although the large majority of surveys were completed electronically, respondents could also complete and return paper versions. Responding to both phone and e-mail requests, over 50 paper surveys were distributed. Only five paper-based responses were returned (these returned surveys were then entered onto the website).
The parent–school collaboration survey took approximately 20–25 minutes to complete. Upon going to the address of the website, respondents first saw a screen describing the study. Once agreeing to participate, the respondent was then directed to the second screen, which contained a brief description of the survey and the survey itself. Upon completing the survey, the respondent was thanked. Respondents then submitted their survey responses. By pushing the completion button, the survey was submitted to the website.
Parent–School Collaboration Survey
The parent–school collaboration survey was comprised of distinct sections related to the parent, student, and parent–school relationship. Within the survey, we included the following scales: the Parenting Stress Index—Short Form (Abidin, 1990); the Family–Professional Partnership Scale (Summers et al., 2005); Advocacy Activities and Knowledge Scale (Burke & Hodapp, submitted); the Scales of Independent Behavior—Revised (Bruininks, Woodcock, Weatherman, & Hill, 1996); and the Big Five Personality Inventory (Rammstedt & John, 2007). Responses generally involved clicking on the appropriate response; the responses were either categorical (e.g., gender) or on a Likert scale. We begin by describing our measure of maternal stress, which in the survey fell within the parent section. We then discuss three sets of correlates: parent–school, child, and parent characteristics.
Parenting Stress Index—Short Form
Developed as a short, easy-to-administer measure of maternal stress, the Parenting Stress Index—Short Form (PSI-SF; Abidin, 1990) involves a subset of the longer PSI measure. Mothers rated themselves on 36 items involving three, 12-item domains. Each item was rated on a 5-point Likert scale from (1) strongly disagree to (5) strongly agree. Domains included Parental Distress, or the “distress that a parent is experiencing in his or her role as a parent” (Abidin, 1990, p. 55); Parent–Child Dysfunctional Interaction, the “parent's perception that his or her child does not meet the parent's expectations, and that interactions with his or her child are not reinforcing to him or her as a parent” (p. 56); and Difficult Child, “behavioral characteristics of children that make them either easy or difficult to manage” (p. 56). Higher scores denoted increased levels of overall maternal stress (PSI-SF Total) and of Parental Distress, Parent–Child Dysfunction, and Difficult Child domains. We also examined the percentage of this sample whose scores fell within the clinical range, or raw scores at or above 90% on the PSI-SF's norming sample. All items load (above .40) on their respective domains, and the PSI's long and short forms correlate highly (r = .82, Abidin, 1990).
Parent–School Relationship and Educational Placement
Family–Professional Partnership Scale
Designed to gauge the family's satisfaction with their relationship with the school, the Family–Professional Partnership Scale is comprised of two factors, with 9 items apiece (Summers et al., 2005). In the Child–Professional Relationship subscale, parents rated items concerning the degree to which they felt satisfied that their child's school personnel had the skills to help the child succeed, build on the child's strengths, treat the child with dignity, keep the child safe, and speak up for the child's best interests when working with other service providers (Cronbach's α = .95). In the Family-Focused Relationship subscale, sample items concern the degree to which respondents feel that school staff members are available to them, are friendly, are honest (even when imparting bad news), use words that parents understand, listen nonjudgmentally, and respect the family's values and beliefs (Cronbach's α = .94). For each item, participants responded on a 5-point Likert scale from (1) never to (5) very often.
Mothers answered the question “How often do you communicate with the school?” The three options were: (1) daily, (2) weekly, and (3) monthly.
Filing for Mediation/Due Process
This was a dichotomous question: “Have you ever filed for mediation or due process?” The response options were: (1) no or (2) yes.
One question was asked about placement: “How much time does your child spend in the regular education classroom?” Responses included: (1) 0–20%, (2) 21–40%, (3) 41–60%, (4) 61–80%, and (5) 81–100%.
Developed for this project from parental behaviors identified in prior studies of parental educational advocacy for students with disabilities (Burke, 2012), the Advocacy Activities and Knowledge Scale (Burke & Hodapp, submitted) was comprised of 8 questions pertaining to whether mothers had used an advocate/attorney at an IEP meeting; had ever had someone else (i.e., not an attorney) attend such meetings; attended a workshop on parental special education rights; had difficulty reading or understanding such rights; or called an agency, read a copy, or talked with another parent about special education rights and/or procedural safeguards. Respondents answered questions on a 5-point Likert scale from (1) never to (5) very often, with total scores ranging from 8 to 40. Items comprised a single factor (48.7% of the variance), Cronbach's α = .82.
Type of Disability
Many studies show that mothers of children with autism experience increased stress levels (Hayes & Watson, 2013) and those of children with Down syndrome decreased levels (Hodapp, 2007). Given these findings and the fact that we had sufficient numbers of respondents in these two subgroups (N = 442 for autism and N = 88 for Down syndrome), we examined separately whether mothers of children with autism or with Down syndrome differed from the remaining respondents in their levels of maternal stress. For each analysis, we compared stress levels of mothers of children with the disability of interest (either autism or Down syndrome) to the stress levels of all other respondents.
Scales of Independent Behavior—Revised
In alignment with the Double ABCX model, the child's behavior problems may relate to maternal stress. We used the 8-item Scales of Independent Behavior—Revised (SIB-R), which provides a General Maladaptive Index (GMI), with higher scores inferring more serious maladaptive behaviors (Bruininks, Woodcock, Weatherman, & Hill, 1996). For this sample, Cronbach's α = .75.
Combining answers to two questions (“What ethnicity are you?” and “Are you Hispanic?”) we created a dichotomous variable of White, non-Hispanic versus Minority.
The Big Five Personality Inventory (BFI-10)
Respondents rated 10 items from the Big Five Inventory—10 (Rammstedt & John, 2007). Principal components analyses (with varimax rotation) revealed 4 factors accounting for 63.4% of the variance. These included extroversion, dependable, neuroticism, and openness to experience (agreeableness did not emerge as a separate factor). Parents responded on a 7-point Likert scale: (1) disagree strongly, (2) disagree moderately, (3) disagree a little, (4) neutral, (5) agree a little, (6) agree moderately, and (7) agree strongly.
We first examined existing scales to ensure that each scale replicated its original factors. We conducted Cronbach's alphas for each scale. Next, we imputed missing values for scales, with mean scores substituted for missing values following the guidelines of Harrell (2001). As initial analyses revealed that the PSI-SF was normally distributed, we proceeded with parametric statistics (e.g., Pearson correlations, t tests, and ANOVAs).
Compared to PSI-SF norming samples, overall stress levels were especially high for mothers of children with disabilities. Compared to the 50% line of the PSI-SF, which equates to a PSI-SF raw score total of 69, mothers in this sample averaged 91.57 (SD = 23.70), one sample t test, t(965) = 29.58, p < .0001. Over half of all respondents (53.2% = 513/965) reported that their PSI-SF stress levels exceeded the 90% cutoff score (i.e., a total raw score of 91 or above), denoting clinical levels of maternal stress, binomial distribution, p < .0001.
Similarly, mothers in this sample reported high levels of stress for each of the three PSI-SF subdomains. Compared to 50% lines of 25, 19, and 25 for Parental Distress, Parent–Child Dysfunction, and Difficult Child domains, scores for these mothers were 29.69 (SD = 9.47), 26.22 (SD = 7.87), and 35.66 (SD = 10.57), respectively, all p values < .0001. Disproportionately higher percentages of mothers also had scores over the 90% level for Parent Distress (27.9%), Parent-Child Dysfunction (44.2%), and Difficult Child (52.4%), all p values < .0001.
Maternal Stress and Child and Parent–Family Characteristics
As found in previous studies, mothers whose children had greater levels of behavior problems showed higher overall stress levels, r(965) = .57, p < .0001. In addition, mothers of children with (vs. without) Down syndrome experienced less stress. Conversely, autism related to increased stress levels. As stress scores may be overinflated in autism due to the child's problem behaviors raising the Difficult Child domain (and, subsequently, the overall PSI-SF score; Zaidman-Zait et al., 2010), we reran analyses separately for each of the three PSI-SF domains. Mothers of children with autism (vs. those without) showed higher scores on all three PSI-SF domains: Parental Distress (means of 31.17 vs. 28.80), Dysfunctional Interaction (27.81 vs. 25.35), and Difficult Child (38.40 vs. 32.88), all p values < .001 (Table 2).
There were also correlations between certain maternal-familial characteristics and maternal stress. Specifically, lower stress levels were reported by mothers who were less neurotic and more dependable, extroverted, and open to experience. Parental race was also significant: African Americans had significantly less stress than White mothers. Compared to White mothers, mothers of Asian, Hispanic, or other minority backgrounds did not differ.
Maternal Stress and Parent–School Characteristics
Both the Child and Family-Focused Relationship subscales of the Family–Professional Partnership scale related to increased parent stress, r(965) = −.202 and −.155, p < .001, respectively. Increased maternal stress was also noted among families who filed (vs. did not file) for mediation or due process, t(965) = −1.92, p < .055, ES = .15, and for respondents who engaged in greater amounts of parental advocacy activities/knowledge, r(965) = .128, p < .001. Frequency of parent–school communication and placement in a regular education class were not significant.
Because we were interested in the ways in which specific parent–school characteristics related to maternal stress, we used linear and curvilinear regression to determine the best fitting models for the relationships between parent–school characteristics and maternal stress. Regarding advocacy, a stronger curvilinear (vs. linear) relation existed with stress. With a linear regression, advocacy only explained 1.3% of the variance of maternal stress, F(1, 963) = 16.01, p < .001. However, using advocacy as a quadratic variable resulted in explaining 2.5% of the variance of maternal stress, F(2, 962) = 12.09, p < .001. A curvilinear versus linear relation thus explained more of the variance, F(2, 962) = 8.05, p = .005.
Equally pronounced nonlinear relationships occurred for the family–school partnership scores. Using a linear regression, the Child subscale explained 1.6% of the variance, F(1, 963) = 16.01, p < .001. However, using a curvilinear regression, the Child subscale explained 4.7% of the variance, F(2, 962) = 15.91, p < .001. This difference is significant, F(2, 961) = 15.62, p < .0001. Similarly, for the Family subscale, the linear regression explained 1.6% of the variance, F(1, 963) = 16.01, p < .001, whereas the curvilinear regression explained 3.3% of the variance, F(2, 962) = 10.80, p < .001. Again, the difference in variance accounted for was significant, F(2, 961) = 8.07, p < .001. For both subscales, when parents experienced stronger partnerships with the school, stress decreased.
As we had over 960 data points, both linear and nonlinear best-fit regression lines are difficult to examine visually. Thus, for illustrative purposes we present the mean PSI-SF scores by quartiles of the Child subscale, the Family subscale, and the levels of parental advocacy. As Figure 1 shows, mothers with poor (i.e., 1st quartile) and mediocre (2nd quartile) parent–school relationships with the school experienced similar, exceptionally high levels of parent stress. Only mothers who experience good (3rd quartile) or excellent (4th quartile) relationships showed decreased stress levels. A similar, albeit reversed, pattern is apparent for parental educational advocacy: compared to mothers who engaged in advocacy at fairly low, fairly high, or very high levels, mothers showing the very lowest levels of advocacy (1st quartile) showed relatively low levels of parent stress (Figure 1).
The same nonlinear patterns occur when examining clinical levels of maternal stress by parent–school characteristic quartile scores. For the Child–Professional Relationship subscale, 38.6% of mothers with excellent (4th quartile) partnerships experienced clinically significantly levels of stress, compared to 49.7%, 55.0%, and 64.2% of mothers with good (3rd), mediocre (2nd), and poor (1st) partnership scores, χ2(3; N = 961) = 52.29, p < .001. Similarly, for the Family-Focused Relationship subscale, 42.1% and 44.3% of mothers with excellent and good partnerships experienced clinically significant levels of stress compared to 55.5% of mothers with mediocre, and 63.8% of mothers with poor partnerships, χ2(3; N = 961) = 48.08, p < .001. In the opposite way, of the mothers who advocated least frequently (1st quartile), 43.2% experienced clinically significant levels of stress compared to 59.7%, 51.8%, and 59.9% of mothers who advocated more frequently, χ2(3; N = 961) = 22.23, p < .001.
This pattern is further illustrated by individual items of each subscale. For all 9 items of the Child subdomain, maternal stress levels were lower only when mothers reported that their child's school often or very often provided specific child-related services or treated the child respectfully. For example, in rating the degree to which the school “has the skills to help your child succeed,” mothers who rated this item as occurring very often (M = 82.98; SD = 25.64) or often (M = 86.98; SD = 21.28) showed lower stress levels than mothers who said the school sometimes (M = 93.66, SD = 22.93), occasionally (M = 95.19, SD = 23.59), or never (M = 96.80, SD = 25.27) had these skills, F(4, 961) = 5.62, p < .001. Although the pattern was not quite as consistent across items of the Family-Focused Relationship subscale, mothers experienced lower stress levels when they reported that school personnel often or very often listened without judging the child or family, were people that mothers could depend on and trust, paid attention to what mothers had to say, were friendly, were honest even when they have bad news, and were available when mothers needed them.
It has long been established that mothers of children with disabilities experience high levels of stress, and that such high levels relate to maternal health and other problems. The Double ABCX model has shed light on the issue of maternal stress, resulting in much research about the influence of stressors and family resources. Few studies, though, have examined how parent–school characteristics relate to maternal stress. Using a large-scale, national survey that highlighted parent–school relationships, we connected maternal stress to educational partnerships.
This study produced three main findings. First, we demonstrated several parent and child correlates of maternal stress. As shown in prior studies (Abbeduto et al., 2004; Baker et al., 2002), the child's level of behavior problems was an especially strong correlate of maternal stress levels. Other correlates of lower maternal stress included the presence of Down syndrome (Hodapp, 2007), the absence of autism (Hayes & Watson, 2013), and parents who were of minority backgrounds (Magaña, 2004). This study thus replicates many child and parent correlates of maternal stress. Even the connection of maternal stress to maternal personality—less stress among mothers who rated themselves as less neurotic and more extroverted, dependable, and open to new experiences—has sometimes been reported in the parenting literature (e.g., Glidden, Billings, & Jobe, 2006).
Second, the quality of family–school partnerships significantly related to maternal stress. For both child and family relations, when mothers enjoyed positive family–school partnerships, they experienced significantly less stress. In addition, mothers also experienced less stress when they engaged in less advocacy activities and (marginally) when they did not enact their procedural safeguards. Parents who do not engage in advocacy or enact their safeguards may be satisfied with services and, subsequently, have lower stress levels. Taken together, better family–school partnerships and less advocacy did relate to lower levels of maternal stress.
Third, this relation often did not occur in a straight-line fashion. Specifically, only the lowest levels of advocacy activities related to lower maternal stress. Anything more than few or no advocacy activities related to increased stress levels. Although Fish (2008) noted the ways in which advocacy activities draw on parents' time and resources, the main distinction for parental stress levels may involve engaging versus not engaging in any advocacy activities.
Even more striking nonlinear connections occurred between maternal stress and the child and family domains of parent–school partnerships. As demonstrated by the regressions and figures, only the best family–school partnership related to lower maternal stress. In contrast, overall maternal stress levels were not significantly different when school personnel never, occasionally, or sometimes performed such activities. Again, this finding, while somewhat unexpected, may highlight the importance of very good—not simply of adequate—parent–school relationships as a factor in promoting less stress for these mothers.
These findings reinforce the general sense that parent–school relationships are important. For most children—those with or without disabilities—the school system constitutes the social service that children and their families most frequently encounter. For students with disabilities and their mothers, school services may be especially important. Given that mothers of children with disabilities also constitute a highly stressed group, it may be important to have strong partnerships between these mothers and the child's teachers (and other school personnel). We also re-emphasize the nonlinear nature of this relationship, as mothers with poor or even adequate parent–school partnerships still reported exceptionally high levels of maternal stress.
Given the relation between maternal stress and family–school partnerships, this study has implications for both policy and practice. The Individuals with Disabilities Education Act (IDEA) embeds many rights for parents and recognizes them as equal partners (Turnbull et al., 2011). This study reinforces the importance of these family–school partnerships. IDEA, however, recognizes the importance of family–school partnerships almost exclusively with respect to student—not parent or family—outcomes. This study expands the focus, highlighting the importance of family–school partnerships for outcomes that include the psychological health of the mother.
In a related way, practitioners may need to expand their focus as to the goals of—family–school partnerships. Within the regular education literature, parents are generally seen as crucial to student achievement (Burke, 2012). Special educators too may need to pay greater attention to the impact of their relationships with families. By cultivating stronger family–professional relationships, practitioners may be able to produce better achievement outcomes for their students and to improve the psychological health of their mothers. As parent–school researchers have long appreciated, the circle then comes around again, as better-supported parents are increasingly able to help their offspring to achieve academic and social success (Epstein, 2001).
These findings also highlight the many unexplored issues concerning parent–school connections and maternal stress. One important issue, for example, involves the conceptualization of stress as a maternal trait that remains constant over time. This conceptualization might be inaccurate. Specifically, does parent stress increase immediately before an IEP meeting? For many families, the jargon used in IEP meetings can be overwhelming and intimidating (Hammond et al., 2008). Anticipating these meetings, mothers may experience increased stress in the days before. Still further, the developmental stage of the child and various social and ecological factors may all affect maternal stress. Additional studies are needed that longitudinally examine parent stress over time as it relates to various aspects of parent–school relations.
Future research may also more closely examine the relation between the type of disability, family–school characteristics, and maternal stress. As demonstrated in this survey, mothers of children with autism (vs. other developmental disabilities) experience greater stress. Child characteristics and, more specifically, the severity of autism symptoms may mediate the amount of maternal stress (Hayes & Watson, 2013). Interventions may be needed to reduce the maternal stress of these families. Also, additional research should examine certain disability groups to refine the relation between maternal stress and family–school partnerships.
It is also important to acknowledge this study's limitations. Although featuring a large, national sample of parents of students with disabilities, this study's web-based format may have precluded participation from parents from low-income or minority backgrounds. Furthermore, while we contacted 7,843 agencies about the study, only 965 mothers participated. Why certain parents participated and others did not remains unknown. We also focused exclusively on maternal—as opposed to paternal—stress, the idea being that mothers and fathers may differ in regard to the amount and correlates of their stress (Krauss, 1993). Limitations of the PSI-SF based on age (Abidin, 1990) and autism (Zaidman-Zait et al., 2010) should also be noted, although we attempted to check for both with our follow-up analyses. Our main limitation, however, most likely entailed this study's cross-sectional nature. As a result, we could not decipher the direction of effects. Do strong family–school partnerships lessen maternal stress, or are less stressed mothers better able to cultivate strong family–school relationships? Without longitudinal data, we cannot determine which explanation (or others) might best explain these connections between parent–school relationships and maternal stress.
Still, even considering these limitations, this study linked parent–school relations to lessened levels of maternal stress. Beyond showing that such connections exist, we also noted that especially strong partnerships—but not poor or even adequate partnerships—seemed related to lower stress levels. Although much more remains to be discovered, this study begins the process of detailing which aspects of parents' relationships with the school are related to lessened levels of maternal stress, thereby tying the functioning of a critically important person in the students' lives—their mothers—to the school environment in which most children spend the majority of their days.
Support for this research was provided by the National Institute on Disability and Rehabilitation Research (H133P1100004) and the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P30 HD15052).
Meghan M. Burke (e-mail: firstname.lastname@example.org), University of Illinois at Chicago, Disability and Human Development, 1640 W Roosevelt Road, Chicago, IL 60608, USA; Robert M. Hodapp, (email@example.com) Vanderbilt University, Vanderbilt Kennedy Center and Department of Special Education, 230 Appleton Place, Nashville, TN 37203-5701.