Abstract
Purpose:
Expectations held by parents of youth with disabilities contribute to decision making and planning for adulthood. This study investigated longitudinal stability of parent expectations about the transition to adulthood for youth with disabilities (i.e., likelihood of obtaining postsecondary education, living independently, becoming financially self-sufficient) and how the stability of parent expectations was related to adult outcomes.
Materials and Methods:
Participants were parents of 3,640 youth who participated in the National Longitudinal Transition Study-2 (NLTS2) Waves 1 and 2 and had available data related to parent expectations.
Results:
Parent expectations were moderately correlated across waves; however, over one-third of parents changed their expectations across waves, including directional shifts (e.g., changing from thinking they probably will achieve the outcome to probably will not achieve it). Directional shifts in parent expectations between Waves 1 and 2 significantly predicted young adult outcomes at Wave 5. The children of parents who demonstrated uncertainty, as indicated by directional expectation shifts, were less likely live independently as young adults. Family involvement in the transition process is critical and should continue to be advocated for in both policy and practice. Improved supports may be warranted for families who experience uncertainty.
Keywords: adolescence, transition to adulthood, parent expectations, youth with disabilities, special education
Introduction
In the United States, young people transitioning to adulthood navigate several developmental tasks, including the completion of schooling, entrance into the labor force, and the establishment of a household separate from parents [1, p.7]. Youth with disabilities face an even more complex process, as they navigate educational, social, and healthcare systems to support their transition [2]. Postsecondary experiences such as attending higher education, maintaining paid employment, and living independently are achieved less often by youth with disabilities than typically developing youth [3–4]. Over the past several decades, researchers and policymakers have examined family, community, and policy variables that increase the capacity of people with disabilities to experience productive engagement in their communities (e.g., postsecondary education or employment), be economically self-sufficient (e.g., financial independence from parents, not receiving benefits like Supplemental Security Income [SSI]), and live independently [3]. Adult outcomes for youth with disabilities are highly variable; some adults with disabilities complete postsecondary education, find jobs, and establish their own households, while others remain mostly dependent on parents and experience difficulty in obtaining education or employment [4].
Available evidence suggests a number of factors contribute to differences in postsecondary outcomes for youth with disabilities, including the type of disability, socioeconomic status, and family factors [4]. For example, in a national study, youth with multiple disabilities were less likely to live independently than peers with numerous other types of disabilities [4]. Youth from households with more financial security were more likely to have postsecondary employment than those from households below the poverty line [4]. In terms of family life, outcomes may be influenced by activities at home such as the provision of chores [5]. Expectations for youth held by their parents (i.e., parent expectations) are a family factor that research has repeatedly demonstrated to be associated with youth outcomes [5–7]. Despite a growing body of evidence indicating the importance of parent expectations for youth with disabilities during the transition to adulthood [5–7], research about the nature of these expectations among parents of youth with disabilities is lacking.
The expectancy-value theory of achievement motivation emphasizes the importance of expectations as predictors of later outcomes [8] and provides a basis for examining the association between parent expectations and youth outcomes. In this theory, expectations “refer to beliefs about how one will do on different tasks and activities” [8, p.110] and these beliefs contribute to the subsequent outcomes. The theory posits that what individuals expect to achieve and what they value have a direct influence on their effort, persistence, choices about achievement, and performance. These expectations are also considered to be influenced by a number of personal and societal factors, including family demographics, the beliefs and behaviors of others, gender role stereotypes, and self-concept of abilities [8]. This theory focuses on an individual’s own expectations about their future; however, because of the prominent role of parents in the preparation of youth with disabilities for adult life [9], it has been extended to inform understanding of parent postsecondary expectations as well [7]. In research on parents’ expectations for youth postsecondary outcomes, expectations are operationalized as parents’ ratings of the likelihood that youth will attain culturally relevant benchmarks of adulthood (e.g., be financially self-sufficient, have a healthy romantic relationship [10–11]).
Parent expectations serve as predictors of outcomes (e.g., academic achievement) for youth with and without disabilities [12–14]. Among youth with disabilities, parent expectations have consistently been identified as predictors of postsecondary outcomes (e.g., employment, autonomy) [5–7], with some differences across disability classification [14–15]. Although the mechanisms remain poorly understood, there is evidence for both direct and indirect effects of parent expectations on youth outcomes. In terms of direct effects, studies focused on youth with autism spectrum disorder (ASD) suggest that parent expectations may affect choices made about transition preparation activities (e.g., talking to youth about jobs or careers) [10]. Consistent with expectancy-value theory, parent expectations can also influence youth expectations for themselves [17–18] and consequently influence youth effort, performance, and achievement choices related to outcomes.
One challenge to this research is that parent expectations and youth achievement are dynamic and mutually responsive over time (e.g., [18]), meaning that it is difficult to use cross-sectional data to identify mechanisms that contribute to the effect of expectations on outcomes [19]. Thus, longitudinal analyses are needed to unpack these constructs and understand ways that parent expectations change over time, factors that contribute to those changes, and whether those changes predict youth outcomes. However, most research to date has focused on expectations in the general population rather than for youth with disabilities. For example, Englund and colleagues [20] measured mothers’ expectations of future educational attainment (e.g., will not complete high school, will go to graduate school) at two time points during elementary school, and found they were correlated at r=0.52. Although significant at p<0.5, this correlation coefficient is weaker than would be expected considering that the same variable was measured with the same individuals across a two-year timeframe. Similarly, Mistry and colleagues [19] measured mothers’ expectations of future educational attainment in high school and found a significant but weak correlation (r=0.38) between parent expectations across time. Both studies found a slight increase in mothers’ expectations across time. These studies showed that parent expectations for general population students show both stability and change over time.
While studies have not yet examined the dynamic nature of parent expectations for youth with disabilities, there is available literature on predictors of stability and change identified in the general population. In a cross-sectional survey of 598 parents of youth in 8th through 10th grades, Jacob [21] found a small but significant (r=0.21, p<0.001) correlation between parent expectations about future educational attainment and parent-reported stability of their expectations over time. Parents with higher expectations perceived themselves as having more stable expectations over time [21]. Using data from the National Education Longitudinal Study of 1988, Zhang and colleagues [22] examined associations between parent socioeconomic resources and their expectations for their child’s academic achievement. Parents of males (compared with females) and parents of White students (compared with Asian, Hispanic, and African American students) had more stable academic expectations across 8th to 12th grades, suggesting that systemic and/or cultural variables play a role in stability of expectations. In this study, parent expectations for educational attainment were comparable for boys and girls in 8th grade but slightly higher for girls by 12th grade [22]. Although both studies conceptualized parent expectations as dynamic, neither study tested whether stability or change in parent expectations were related to student outcomes. Further, though these studies are informative about parent expectations, there is no literature examining change in parent expectations among youth with disabilities specifically.
In summary, existing research has identified that parent expectations relate to youth postsecondary outcomes among youth with disabilities. However, although we know from literature with the general population that expectations are dynamic constructs, little is known about factors that predict stability or change in parent expectations for youth with disabilities, or about how stability or change in expectations relate to youth postsecondary outcomes. Research in the general population suggests parent expectations correlate only weakly to moderately over time and that factors such as child gender and family culture can influence expectation stability. To understand if and how the postsecondary expectations of parents of youth with disabilities change over time and to examine associations with changes in parent expectations, we addressed the following research aims: (1) determine stability of parent postsecondary expectations (i.e., how similar expectations are at two time points); (2) identify and describe instances of positive and negative changes (i.e., changing belief that an outcome is more or less likely over time) in parent postsecondary expectations; (3) determine associations between youth characteristics and directional parent expectation shifts (i.e., changing from believing an outcome is likely to unlikely, or vice versa); and (4) explore if directional shifts in parent expectations are associated with youth postsecondary outcomes. Based on existing literature, we hypothesized that parent expectations would be moderately stable over time and that parents of males and those with higher socioeconomic status would have more stable expectations. There was not sufficient literature to support hypotheses around differences related to disability type or if shifts in expectations would be associated with youth outcomes, but we did anticipate negative shifts in expectations to predict less independent outcomes.
Materials and Methods
NLTS2 Dataset
The National Longitudinal Transition Study-2 (NLTS2 [23]) provides longitudinal nationally-representative data on students who were enrolled in U.S. special education services in 2000 (beginning of the sampling period). Data collection from parents, youth, and schools occurred across five waves approximately two years apart, covering a time period when youth were 13–16 through 21–25 years of age. Although it is now several years old, the NLTS2 is the best available resource to examine our longitudinal research questions.
The NLTS2 team randomly sampled school districts and students, resulting in over 11,000 youth representing the broader population in terms of geography, district/school size, demographic characteristics, age, and disability classification [23]. At entry into the study, students had to be at least 13 years of age and in seventh grade. The NLTS2 team then developed a sample of students in special education by randomly selecting from a nationally representative pool of local education agencies (LEAs) and a sample of state supported special schools. They stratified LEAs and special schools by region, district enrollment, and district wealth. Student samples included at least 1,250 students in most disability categories except for those which are rarer including autism, traumatic brain injury, and deaf-blindness. They tested the response sample to assess for bias with regard to the whole population to which the results are generalized; researchers determined that differences were extremely minimal [24]. The Institute of Education Sciences (IES), who oversees the NLTS2 data, provides variable weights to represent the population for which the data was designed for generalization; they strongly recommend use of these sample weights for analysis. We had an IES Restricted-use Dataset Agreement in place for the access and use of the NLTS2; accordingly, we upheld IES data security procedures and herein report sample sizes to the nearest 10.
Sample and Variables.
For the present analyses, we included individuals who had available parent expectation data (at least one key variable) at both Waves 1 and 2 (approximate N=3,640). NLTS2 expectation data is part of the parent telephone survey data, which is the data type with the most complete available data [24]. However, the sample size available for our current analyses represents only about a third of the total NLTS2 sample. Exclusion from our analyses was related to youth age and socioeconomic factors; a post-hoc un-weighted logistic regression analysis revealed that youth who were older (p<0.001), had received free/reduced meals (p<0.01), and were not White (p<0.05) were significantly less likely to have data necessary for inclusion. Neither youth gender nor Wave 1 parent expectations significantly predicted exclusion.
We selected key variables for these analyses from Waves 1, 2, and 5; the data were provided by the school district, the parent, or the youth (if parent reported they were capable of participating in data collection). Additionally, we re-coded and created new variables through calculation as necessary to address the research aims (e.g., merging response options or creation of a new variable with data from multiple existing variables). Table 1 presents descriptive information on the sample for all key variables included in these analyses. We included variable names from the NLTS2 dataset to increase transparency and reproducibility of our analyses.
Table 1.
Sample Description on Study Variables
| Variable | Weighted Distribution | Wave, Variable name (source) | |
|---|---|---|---|
| Age in years—Mean (SD) | 14.33 (0.82) | Wave 1, Age071501 (school district) | |
| Male gender | 66.7% | Wave 1, np1A1 (parent survey) | |
| Main disability | Wave 1, w1_Dis12 (school district) | ||
| Learning disability | 61.7% | ||
| Speech impairment | 4.9% | ||
| Intellectual disability | 11.6% | ||
| Emotional disturbance | 10.8% | ||
| Hearing impairment | 1.4% | ||
| Visual impairment | 0.5% | ||
| Orthopedic impairment | 1.2% | ||
| Other health impairment | 4.7% | ||
| Autism | 0.9% | ||
| Traumatic brain injury | 0.3% | ||
| Multiple disabilities | 1.8% | ||
| Deaf-blindness | 0.2% | ||
| Race/ethnicity | Wave 1, np1A3b (school district) | ||
| White | 64.2% | ||
| African American | 15.6% | ||
| Hispanic | 18.4% | ||
| Asian/Pacific Islander | 1.1% | ||
| American Indian/Alaska Native | 0.5% | ||
| Other or multiple | 0.2% | ||
| Eligible for free/reduced meals | 39.2% | Wave 1, np1St_Meal (school district) | |
| Parent expects youth will attend school after high school (postsecondary education) | Wave 1, np1J2 reverse scored [not reported here: Wave 2, np2G6] (parent survey) | ||
| Definitely will not (1) | 7.1% | ||
| Probably will not (2) | 27.9% | ||
| Probably will (3) | 38.8% | ||
| Definitely will (4) | 26.3% | ||
| Parent expects youth will live away without supervision (independent living) | Wave 1, np1J7 reverse scored [not reported here: Wave 2, np2G10] (parent survey) | ||
| Definitely will not (1) | 5.1% | ||
| Probably will not (2) | 7.5% | ||
| Probably will (3) | 28.8% | ||
| Definitely will (4) | 58.6% | ||
| Parent expects youth will earn enough to support self (financial independence) | Wave 1, np1J10 reverse scored [not reported here: Wave 2, np2G12b] (parent survey) | ||
| Definitely will not (1) | 3.8% | ||
| Probably will not (2) | 12.3% | ||
| Probably will (3) | 34.3% | ||
| Definitely will (4) | 49.6% | ||
| Ever attended a postsecondary institution since leaving high school | 54.7% | Wave 5, np5A3a_A3e_A3i_ever (parent/youth survey) | |
| Ever worked for pay outside the house since leaving high school | 85.3% | Wave 5, np5A4a (parent/youth survey) | |
| Does not currently receive money from Supplemental Security Income | 84.4% | Wave 5, np5A4g reverse scored (parent/youth survey) | |
| Lives on own (not with parents, other guardian, or in other supervised situation) | 39.6% | Wave 5; see note below np5A4g (parent/youth survey) | |
Notes. Percentages were calculated based on available data for each variable (excluded if missing). Final listed variable was created by linking multiple variables to separate those living on their own (including with roommate, partner/spouse, and in college dorm) from those reported to live with parents or in other supervised settings.
Our primary variables of interest in this study relate to questions about parent expectations. The NLTS2 survey repeated parent expectation questions at Waves 1 and 2, if relevant, (i.e., expectation questions were not asked if the youth had already accomplished the outcome). The same parent respondent was invited to participate at each wave, providing consistency across waves. The survey asked respondents about the likelihood that their youth would attain up to 10 different outcomes (e.g., graduate high school, attend postsecondary school, live away from home without supervision, get a paid job, earn enough to be self-sufficient). The four response options included: definitely will not, probably will not, probably will, definitely will. We selected three expectation questions (i.e., attend postsecondary school, live away from home without supervision, and earn enough to be financially self-sufficient) for this study because they addressed expectations about three key postsecondary outcome areas and were the most normally distributed (i.e., others were highly skewed). We also calculated new variables to summarize magnitudes of change in expectations (i.e., how many points the expectation changed on the four-point scale) and directional shifts (i.e., if expectations shifted from probably/definitely will not to probably/definitely will or vice versa).
In addition to parent expectation data, we used descriptive variables and postsecondary outcome variables. Descriptive variables from Wave 1 included age, gender, disability classification, race/ethnicity, and whether or not the youth was eligible to receive free or reduced meals in school (i.e., proxy for socioeconomic status). We also selected variables representing relevant postsecondary outcomes (i.e., ever attended postsecondary institution, ever worked for pay outside the house, currently receives money from SSI, does not live with parents or in another supervised situation) from Wave 5.
Data Analysis
We conducted all analyses using the Statistical Package for Social Sciences, Version 25. We only included the cases for whom there was valid available data (i.e., pairwise removal of missing data). Pairwise deletion allowed us to answer each research question with the maximum data available in the NLTS2. We weighted all our analyses using the sampling weight designed for cases with parent interview data at any wave (wt_AnyPY) and we adjusted our analyses using Bonferroni corrections. To determine stability of parent expectations between Waves 1 and 2 (Aim 1), we calculated Pearson correlation coefficients to examine general consistency in score patterns and paired t-tests to examine if there were significant changes on average in parent expectations across waves. We generated descriptive statistics to examine and describe changes in the expectations of individual cases (i.e., direction and magnitude of changes) and to describe instances of directional (i.e., will/will not) shifts in expectations across waves (Aim 2). We applied Chi-square and logistic regression techniques to determine associations between youth characteristics (i.e., age, gender, race/ethnicity, free/reduced meal status, disability classification) and directional (will/will not) parent expectation shifts from Wave 1 to 2 (Aim 3). We then conducted logistic regression analyses with four binary postsecondary outcome variables (yes or no: ever attended postsecondary education, lives on own, ever worked for pay outside the house since leaving high school, does not currently receive money from SSI) to test if directional expectation shifts (will/will not; positive shift, negative shift, or no shift [reference group]) for relevant parent expectations predicted postsecondary outcomes at Wave 5 when controlling for age, gender, race/ethnicity, receipt of free/reduced meals, and disability classification (Aim 4). For Aim 4, expectation and outcome pairings included: (1) postsecondary education expectation predicting postsecondary education outcome, (2) independent living expectation predicting lives on own outcome, (3) financial independence outcome predicting ever worked for pay outcome, and (4) financial independence outcome predicting no current receipt of SSI outcome.
Results
Aim 1: Stability in Expectations
Table 2 presents the means and standard deviations of expectation ratings at each wave along with Pearson correlation coefficients and t-statistics for comparisons between Waves 1 and 2. We calculated significant correlation coefficients between waves for expectation ratings in all three areas (i.e., postsecondary education, independent living, financial independence). Given that the same respondents answered the same expectation questions across both waves, the magnitudes of the correlation coefficients are of particular interest. Consistent with our hypotheses, postsecondary expectation correlation coefficients were moderate in size (0.43 – 0.59). Next, we conducted t-tests to determine whether there were any significant differences in parent expectations across waves. We observed significant differences on all three expectation variables, with postsecondary education expectations increasing on average while independent living and financial independence expectations decreased on average.
Table 2.
Comparison of Parent Expectations Across Waves
| Parent Expectations |
Wave 1 M (SD) |
Wave 2 M (SD) |
|
|---|---|---|---|
| Postsecondary Education | 2.78 (0.9) | 3.10 (0.8) | |
| Pearson correlation | 0.44*** | ||
| Partial correlationa (youth age) | 0.43*** | ||
| T-test | −74.1*** | ||
| Observed sample size | 2630 | ||
| Independent Living | 3.35 (0.9) | 3.33 (0.9) | |
| Pearson correlation | 0.59*** | ||
| Partial correlationa (youth age) | 0.59*** | ||
| T-test | 76.1*** | ||
| Observed sample size | 3640 | ||
| Financial Independence | 3.25 (0.8) | 3.19 (0.9) | |
| Pearson correlation | 0.52*** | ||
| Partial correlationa (youth age) | 0.52*** | ||
| T-test | 79.6*** | ||
| Observed sample size | 3360 | ||
Notes.
partial correlations accounted for youth age. Bonferroni correction: adjusted for 6 tests—correlation and t-test for each variable set:
p < 0.008;
p < 0.001;
p < 0.0002.
Observed sample sizes are rounded to the nearest 10 in accordance with the data-use agreement; analyses are weighted.
Aim 2: Describing Changes in Expectations
Approximately half to two-thirds of parents rated their expectations exactly the same across the waves (49.1% postsecondary education; 62.7% independent living; 57% financial independence). Of parents who did change their expectations, the majority did so by only one rating level on the 1 to 4 scale. We also examined directional expectation shifts (shifting from probably/definitely will to probably/definitely will not, or vice versa); between 10.4% and 26.1% of all parents made a directional shift on the expectation variables between Waves 1 and 2. Figure 1 displays directional shift percentages across waves on each of the three expectation questions. More parents made a positive shift (i.e., shifted from probably/definitely will not to probably/definitely will) than a negative shift on postsecondary education expectations, whereas we observed the opposite pattern for independent living and financial independence expectations.
Figure 1.
Weighted Percentages of Stability and Change in Parent Expectations
Notes. W1, Wave 1; W2, Wave 2. Horizontal arrows indicate the percentage of parents with stable expectations across waves. Arrows pointing up or down indicate the percentage of parents who reported that their expectations changed in the indicated direction.
Aim 3: Characteristics and Experiences Associated with Directional Shifts in Expectations
To identify potential factors that may contribute to directional (will/will not) shifts in parent expectations, we computed binary variables indicating whether or not a directional shift was observed for each expectation between Waves 1 and 2. We entered youth characteristics (age, gender, race/ethnicity, and receipt of free or reduced meals) in logistic regression analyses to determine predictors of directional shifts in parent expectations. Table 3 displays results of these analyses. All entered variables were found to be significant across each expectation model, with the exception of gender predicting directional shift in independent living expectations and non-significance with some of the smaller racial/ethnic groups. Across expectation types, parents of older youth and youth who received free/reduced meals through school were more likely to have directional expectation shifts. Parents of males were more likely to have a directional postsecondary education expectation shift, while parents of females were more likely to have a directional financial independence expectation shift. Parents of African American youth were less likely than those of White youth to change postsecondary education expectations, yet more likely to change their independent living and financial independence expectations. Parents of Hispanic youth were more likely than those of White youth to change postsecondary education expectations, yet less likely to change their independent living and financial independence expectations. Parents of youth in other racial/ethnic groups were generally less likely to change their expectations than parents of White youth or showed no significant difference.
Table 3.
Logistic Regression Analyses Examining Youth Characteristics as Predictors of Directional Shifts in Parent Expectations
| Youth Characteristics | Directional Shifts in Parent Expectations | ||
|---|---|---|---|
| Postsecondary Education OR (CI*) |
Independent Living OR (CI*) |
Financial Independence OR (CI*) |
|
| Age | 1.48 (1.45 – 1.50) | 1.05 (1.03 – 1.08) | 1.43 (1.40 – 1.46) |
| Male gender | 1.15 (1.12 – 1.19) | 1.02 (0.99 – 1.06) | 0.70 (0.67 – 0.72) |
| Race/Ethnicity | |||
| African American | 0.66 (0.63 – 0.69) | 1.19 (1.13 – 1.25) | 1.32 (1.26 – 1.37) |
| Hispanic | 1.08 (1.04 – 1.19) | 2.14 (2.06 – 2.23) | 0.74 (0.71 – 0.77) |
| Asian/Pacific Islander | 0.18 (0.14 – 0.23) | 0.95 (0.81 – 1.19) | 0.29 (0.23 – 0.37) |
| American Indian/Alaska Native | 1.03 (0.83 – 1.29) | 0.28 (0.18 – 0.44) | 0.54 (0.41 – 0.72) |
| Other or multiple | 0.11 (0.03 – 0.52) | - | - |
| Eligibility for free/reduced meals | 1.31 (1.27 – 1.35) | 2.87 (2.80 – 2.95) | 2.69 (2.59 – 2.79) |
| Observed sample size | 1910 | 2610 | 2400 |
Notes. OR, odds ratio; CI, confidence interval.
Bonferroni correction, adjusted for 3 tests, resulted in a CI of 99.983%; ORs with CIs not spanning 1.0 are bolded. Dashes (−) indicate insufficient data to analyze. Observed sample sizes are rounded to the nearest 10 in accordance with the data-use agreement; analyses are weighted.
Using chi-square analyses with a Bonferonni correction, we compared rates of directional (will/will not) shifts in parent expectations across youth disability categories. Results are presented in table 4. We observed differences across groups in the rates of directional expectation shifts, with variability across expectation types. Across expectation types, parents of youth with emotional disturbance were consistently among the most likely to have a directional shift in expectations, and parents of youth with speech impairment were consistently among the least likely to have a directional shift in expectations.
Table 4.
Comparisons of Directional Expectation Change by Disability Classification
| Percent Directional Expectation Change | |||
|---|---|---|---|
| Disability Category | Postsecondary Education | Independent Living | Financial Independence |
| Learning Disability | 26.2%a | 6.7%a | 12.2%a |
| Speech Impairment | 13.6%b | 10.0%b | 8.0%b |
| Mental Retardation | 33.3%c | 18.5%c,d | 17.9%c |
| Emotional Disturbance | 32.3%d | 20.1%e | 18.6%c |
| Hearing Impairment | 16.0%e | 7.9%f | 5.9%d |
| Visual Impairment | 13.5%b,e | 21.4%e | 12.2%a,e,f |
| Orthopedic Impairment | 10.6%f | 13.0%g | 18.1%c |
| Other Health Impairment | 19.5%g | 10.1%b | 12.2%a |
| Autism | 24.4%a | 16.5%h | 11.0%a,f |
| Traumatic Brain Injury | 28.6%a,d | 21.5%d,e | 9.2%b,f |
| Multiple Disabilities | 32.7%c,d | 16.7%h | 13.6%e |
| Deaf-blindness | 17.4%b,e,g | 15.1%c,g,h | 18.9%c |
| Observed sample size | 2630 | 3640 | 3360 |
Notes. Directional change indicates parent changed ‘probably/definitely will’ to/from ‘probably/definitely will not’ between Waves 1–2. Values in each column not sharing the same subscript letter(s) are significantly different in the two-sided test of equality using Bonferroni correction. Observed sample sizes are rounded to the nearest 10 in accordance with the data-use agreement; analyses are weighted.
Aim 4. Outcomes Associated with Directional Shifts in Expectations
Table 5 presents the results of the logistic regression analyses examining a relationship between directional shifts in expectations (i.e., will/will not; positive shift, negative shift, or no shift [reference group]) and postsecondary outcomes (i.e., ever attended postsecondary education, lives on own, ever worked for pay outside the house since leaving high school, currently not receiving money from SSI) while controlling for age, gender, race/ethnicity, receipt of free or reduced meals, and disability classification. Negative shifts in expectations predicted outcomes in the expected direction (i.e., negative shifts predicted less independence in adulthood). Specifically, parents who had negative directional shifts in their expectations were less likely to have youth who participated in postsecondary education, lived on their own, worked for pay, and who did not receive SSI payment. The difference was most profound for the postsecondary education outcome, wherein youth whose parents experienced a negative expectation shift were less than a fifth as likely to attend postsecondary education as those who did not shift their expectations. The prediction of positive directional shifts in expectations were less consistent, however. As expected, parents who had positive directional shifts in financial independence expectations were more likely to have youth who worked for pay. Unexpectedly, parents who had positive shifts in independent living expectations were less likely to have youth who lived on their own. Youth whose parents demonstrated positive directional shifts in their expectations for postsecondary education and financial independence did not significantly differ from those whose parents did not shift their expectations for postsecondary education participation or receipt of SSI benefits, respectively.
Table 5.
Logistic Regression Models Testing if Directional Shifts in Parent Expectations Predict Postsecondary Outcomes
| Postsecondary Education OR (CI*) |
Lives on Own OR (CI*) |
Worked for Pay OR (CI*) |
Not Receiving SSI OR (CI*) |
|
|---|---|---|---|---|
| Directional Shifts in Parent Expectations | ||||
| Positive change | 1.01 (0.97 – 1.06) | 0.83 (0.76 – 0.92) | 2.90 (2.50 – 3.36) | 1.14 (0.98 – 1.32) |
| Negative change | 0.18 (0.17 – 0.19) | 0.55 (0.51 – 0.58) | 0.27 (0.25 – 0.29) | 0.63 (0.58 – 0.70) |
| Observed sample size | 1260 | 1660 | 1510 | 1330 |
Notes. SSI, Supplemental Security Income; OR, odds ratio; CI, confidence interval. Relevant expectation variables used for each outcome (see text). Covariates included in these models were youth age, gender, race/ethnicity, disability classification, and receipt of free/reduced meals. The reference group for directional shifts variables was the group of parents with no directional change in expectations, the reference group for race/ethnicity was White, and the reference group for the disability classification was Multiple Disabilities. Bonferroni correction, adjusted for 4 tests, resulted in a CI of 99.9785%; ORs with CIs not spanning 1.0 are bolded. Observed sample sizes are rounded to the nearest 10 in accordance with the data-use agreement; analyses are weighted.
Discussion
In the current study, we examined the longitudinal stability of parent expectations for postsecondary outcomes among parents of youth with disabilities in the largest and most recent dataset available for this purpose. The expectancy-value theory of achievement motivation theorizes that beliefs about whether an individual will achieve a goal and how much the individual or their family values a particular goal both contribute to whether they achieve that goal or outcome [8]. This theory has been proposed to also extend to understanding how parent expectations influence the transition process for youth with disabilities [7]. The current study investigated change over time in parent expectations for the adult outcomes of their youth with disabilities, predictors of shifts in parent expectations, and how stability or uncertainty in parents’ expectations predicted adult postsecondary outcomes.
We found evidence for both consistency and change in parent expectations for youth’s postsecondary outcomes. Parent expectations at Wave 1 and Wave 2 were moderately correlated within each domain, with significantly different mean scores between waves. One-half to two-thirds of parents reported expectations that remained stable after two years (i.e., endorsed the same response option). However, we also found evidence for change in parental expectations, including a subset of parents with positive or negative directional shifts (i.e., changing from thinking the outcome is unlikely to likely or vice versa, respectively) in their expectations. The findings of moderate, but significant, correlations for educational expectations over time are consistent with previous literature on educational expectations among parents of youth without disabilities [19–20], demonstrating that parent expectations for youth with disabilities may follow similar patterns. We also identified that age, gender, race/ethnicity, socioeconomic status, and disability classification relate to changes in expectations, some of which has been identified in prior literature (e.g., differences by race [22]) while other findings expand on prior literature. This study also contributed to the literature by examining directional expectation shifts, which we argue signify the most meaningful changes in expectations.
Directional expectation shifts were most common in the domain of postsecondary education, suggesting that parents may be particularly uncertain regarding whether youth with disabilities will have the opportunity to access postsecondary education. The costs of attending college are substantial and rising, and parents may be uncertain whether youth with disabilities will have the skills and support needed to graduate with a degree, or whether investing in a college degree will provide a significant increase in earning potential or life satisfaction [25]. The effects of a college degree on lifetime earnings differ by gender, race/ethnicity, and occupation, and it is possible for an individual with only a high school diploma to make more money in some occupations than an individual with an Associate’s degree might make in others [25]. Notably, young adults with disabilities who complete postsecondary education have higher average hourly wages ($12.50) than those who complete high school ($9.50) or some postsecondary education ($9.80), but are not more likely to receive benefits like paid vacation, sick leave, or health insurance [4]. Research suggests that parents of youth with some disabilities (e.g., intellectual disability) may have limited knowledge and limited access to information about postsecondary education because of low expectations on the part of school employees [26]. This points to a need for dissemination of information about postsecondary education as a transition option among parents, teachers, and others who engage in transition planning.
Directional shifts in parent expectations may occur for myriad reasons. For students with disabilities in the NLTS2 sample, we posit that parents may have received new information from schools or other providers and/or have had new experiences that altered their perceptions. A recent study examined the formation of parent expectations using qualitative semi-structured interviews with parents of adolescents with ASD at a single time point [27]. That study suggested that there were three main types of influences on parent expectation formation: youth factors (e.g., interests, strengths, challenges, past experiences), parent factors (e.g., uncertainty, hope, fear), and social and societal influences (e.g., presence or lack of exemplars/models, and services and resources) [27]. Future studies examining reasons for directional shifts in expectations for parents of youth with disabilities using both quantitative and qualitative methods could further elucidate this phenomenon and offer additional implications for practice.
To understand potential impacts of parent expectation changes, we also examined the relationship between directional parent expectation shifts and postsecondary youth outcomes. As expected, children of parents who raised their financial independence expectations over time were more likely to work for pay; conversely, children of parents who lowered their expectations were less likely to work for pay. Interestingly, however, we identified that both negative and positive shifts in parent expectations were predictive of decreased likelihood of youth achieving independent living outcomes, though the effect is smaller for those with positive shifts. One potential explanation for this is that parental uncertainty about the future could affect youth’s postsecondary education and independent living outcomes.
Uncertainty is a common experience for parents of youth with disabilities (e.g.,[28]). When parents receive conflicting information or are generally unsure of what their child’s future will look like, it is possible that they could experience an inability to make goal-oriented, effective decisions [29] when it comes to making transition-related decisions and helping prepare youth for adulthood. Our results suggest that parents experiencing uncertainty (i.e., displaying directional changes in expectations across time—positive or negative) may be in particular need of support during their youths’ transition to adulthood to ensure youth are given ample opportunities to prepare for postsecondary outcomes that are well-suited to them. In our analyses, the parents who experienced directional expectation shifts were more likely to have youth who were older and who were receiving free or reduced meals at school. Thus, parents of older children and those of lower socioeconomic status may be particularly in need of guidance to support goal-oriented decision-making and maximize youth potential for independence in adulthood. Furthermore, a recent publication by Wigfield and Gladstone examined the expectancy-value theory of achievement motivation in the context of change and uncertainty [30]. They were not specifically looking at parent uncertainty, but suggested that maintenance of positive expectations can support better abilities for children to cope with uncertainty [30]. This may be related to our finding that those who demonstrated uncertainty were less likely to live independently, and may translate to parent uncertainty as well. Thus, promoting parents to maintain more positive expectations may better support families to cope with change and uncertainty.
Rehabilitation professionals, such as occupational therapists and rehabilitation counselors, may be involved in transition planning for youth with disabilities in the community [34–35]. Occupational therapists can support the development of life skills, identifying needed environmental adaptations, and meeting personal goals [34]. Rehabilitation counselors offering vocational rehabilitation services can support youth with disabilities to develop employment skills and to enter the workforce [35]. Unlike educational services, which are mandated to involve parents [36], rehabilitation services in the U.S. are not necessarily guided by specific policies requiring the involvement of parents in the planning process. However, theory and guidelines about the adolescent-to-adult healthcare transition point to the need for coordination and communication between youth, parents, and professionals to promote optimal psychosocial outcomes [2,37–38]. This is the foundation of the family-centered care model that is widely embraced in pediatric rehabilitation practice. Based on the results of this study and the existing literature pointing to the important role of parent expectations during the transition process, there is a strong need for rehabilitation professionals to collaborate with parents of youth with disabilities to reduce uncertainty and promote positive outcomes for youth. Furthermore, the passage of the Workforce Innovation and Opportunity Act (WIOA) in 2014 supports inter-agency collaboration to support employment for youth with disabilities. As distinct agencies work together under WIOA to achieve employment for individuals with disabilities, there are increased opportunities for initiatives to provide education and consistent communication to families about opportunities and outcomes and to encourage providers to take the expectations of parents and youth as important considerations in service planning.
Finally, our findings relating to gender warrant consideration. In the present study, parents of males were more likely to have a directional shift in postsecondary education expectations, while parents of females were more likely to have a directional shift in financial independence expectations, suggesting more uncertainty about whether males will attend college or whether females will be able to support themselves financially. Previous work has found that the relationship between parental expectations and postsecondary outcomes is important regardless of gender [6], but that parents of males may have more stable expectations [22]. Our findings expand upon this to identify gender-related influences on expectation changes over time may differ by expectation type. The finding may be related to broad patterns for postsecondary outcomes among males and females. Specifically, national statistics indicate that females are more likely to attend postsecondary education than males [31], which may lead to less uncertainty about postsecondary education among parents of females, while males are more likely to earn more money [32], leading to less uncertainty about financial independence among parents of males.
Limitations and Future Directions
General limitations of this study relate to the nature of secondary data analysis, including lack of control over recruitment, the high prevalence of missing data, potential for bias, and the design of survey questions. To address the aims of this study, parent expectation data from both Wave 1 and Wave 2 were needed. Additionally, at Wave 2, parents of youth who had already achieved an outcome (e.g., postsecondary education) were not presented with a question about their expectations for whether that outcome would occur in the future. Therefore, the particular questions selected for this study compounded the typical issues of missing data in longitudinal research associated with attrition (e.g., change of contact information). Despite the large, generalizable nature of the NLTS2 dataset broadly, missing data over subsequent longitudinal waves means that our study may not generalize to the broader population of youth with disabilities in the U.S. Since we found age, race/ethnicity, and receipt of free/reduced meals to be significant factors related to expectation changes as well as to study inclusion, our findings should be considered with the caveat that our included sample was less generalizable on those variables than at Wave 1.
A further consideration is that the NLTS2 is dated, and may not reflect the most current experiences of parents and youth. Some prior research on parent expectations has focused specifically on the effects that mothers or fathers have on youth, and on the differences between the expectations of mothers and fathers (e.g., [18]). Therefore, an additional limitation of this study was the generic use of parent expectations without examining differences in expectations by parent/caregiver gender or marital status. However, there are numerous benefits to being able to use the NLTS2’s national sample, including the ability to examine participant responses to the same questions longitudinally among a large and diverse sample.
Finally, the current study examined some predictors and outcomes associated with changes in parent expectations for transition outcomes, yet there are likely many other factors that could contribute to stability or instability of parent expectations for youth with disabilities. For example, prior research related to parent expectations has suggested that parent beliefs can contribute to the formation of expectations [27] and can alter the stability of expectations [33]. Specifically, Yamamoto and Holloway [33] hypothesized that parents who believe effort determines the success of their child may be more likely to alter their expectations based on their perception of the child’s efforts, whereas parents who believe success is based on innate abilities (e.g., intellect) will be more likely to report stable expectations over time. Future research could incorporate more assessment of parent achievement-related beliefs in order to better understand stability and instability in parent expectations for youth outcomes.
Conclusion
This study examined stability and change in parent expectations for postsecondary outcomes and explored how directional shifts in expectations were related to later postsecondary outcomes. There is evidence that parents of older children and those from lower SES backgrounds were more likely to experience an impactful change in their expectations. For some postsecondary outcomes, positive shifts in parent expectations (i.e., believing their youth could be more independent than they thought previously) were associated with more independent outcomes; however, for other outcomes, any shifts in expectations were associated with less independence in adulthood, suggesting parental uncertainty should be taken into consideration during the transition planning process. This study provided further evidence to support the important role parents play in the transition process and the need for policy and practice involving rehabilitation professionals to involve families in the transition planning and preparation processes. Based on the results of this study, additional supports may be warranted for families who experience uncertainty.
Implications for Rehabilitation.
Practitioners should recognize that parents of youth with disabilities have expectations for their child’s transition to adulthood that may change over time and are influenced by factors such as youth age, disability classification, gender, and socioeconomic status.
The ways parents’ expectations change over time may have lasting implications for their child; specifically, we identified that youth of parents who change their expectations negatively (by shifting to expect less independence) are less likely to attain certain adult milestones.
Parents experiencing uncertainty about their youths’ postsecondary potential may require additional support and resources to understand their child’s strengths and challenges, consider available services, and ultimately to promote more independent outcomes for the youth.
Acknowledgements
The data used in this study was acquired through a data-use agreement with the Institute for Education Sciences. The first author’s research efforts were supported by the [university] Program in Personalized Health and the National Center for Advancing Translational Sciences of the National Institutes of Health under Award Number KL2TR001065. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Declaration of Interest Statement
The authors report no conflicts of interest.
Contributor Information
Anne V. Kirby, University of Utah.
Laura Graham Holmes, Drexel University.
Andrew C. Persch, Colorado State University.
References
- [1].Furstenberg FF, Rumbaut RC, Settersten RA. On the Frontier of Adulthood: Emerging Themes and New Directions. In Settersten RA Jr., Furstenberg FF Jr., & Rumbaut RG (Eds.), The John D. and Catherine T. MacArthur foundation series on mental health and development. Research network on transitions to adulthood and public policy). On the frontier of adulthood: Theory, research, and public policy (pp. 3–25). 2005. Chicago, IL, US: University of Chicago Press. [Google Scholar]
- [2].Gorter JW, Stewart D, Woodbury-Smith M. Youth in transition: care, health and development. Child Care Health Dev. 2011;37:757–63. [DOI] [PubMed] [Google Scholar]
- [3].Newman L, Wagner M, Cameto R, Knokey A-M. The post-high school outcomes of youth with disabilities up to 4 years after high school: A report of findings from the National Longitudinal Transition Study-2 (NLTS2) (NCSER 2009–3017). 2009. Menlo Park, CA: SRI International. [Google Scholar]
- [4].Newman L, Wagner M, Knokey A-M, Marder C, Nagle K, Shaver D, Wei X,…Schwarting M The Post-High School Outcomes of Young Adults With Disabilities up to 8 Years After High School. A Report From the National Longitudinal Transition Study-2 (NLTS2) (NCSER 2011–3005). 2011. Menlo Park, CA: SRI International. [Google Scholar]
- [5].Carter EW, Austin D, Trainor AA. (2012). Predictors of postschool employment outcomes for young adults with severe disabilities. J Disabil Policy Stud. 2012;23:50–63. [Google Scholar]
- [6].Doren B, Gau JM, Lindstrom LE. (2012). The relationship between parent expectations and postschool outcomes of adolescents with disabilities. Except Child. 2012;79,7–23. [Google Scholar]
- [7].Kirby AV. Parent expectations mediate outcomes for young adults with autism spectrum disorder. J Autism Dev Disord. 2016;46:1643–1655. [DOI] [PubMed] [Google Scholar]
- [8].Eccles JS, Wigfield A. Motivational beliefs, values, and goals. Ann Rev Psychol. 2002;53:109–132. [DOI] [PubMed] [Google Scholar]
- [9].Ankeny EM, Wilkins J, Spain J. Mothers’ experiences of transition planning for their children with disabilities. Teach Except Child. 2009;41:28–36. [Google Scholar]
- [10].Holmes LG, Kirby A, Himle MB, Strassberg DS. Parent expectations and preparatory activities as adolescents with ASD transition to adulthood. J Autism Dev Disord. 2018;48:2925–2937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Holmes LG, Himle MB, Strassberg DS. Parental romantic expectations and parent-child sexuality communication in autism spectrum disorders. Autism. 2016;20(6):687–99. [DOI] [PubMed] [Google Scholar]
- [12].Jeynes WH. The relationship between parental involvement and urban secondary school student academic achievement: A meta-analysis. Urban Educ. 2007;42:82–110. [Google Scholar]
- [13].Davis-Kean PE, Sexton HR. Race Differences in Parental Influences on Child Achievement: Multiple Pathways to Success. Merrill-Palmer Quart. 2009;55(3):285–318. [Google Scholar]
- [14].Neuenschwander MP, Vida M, Garrett JL, Eccles JS. Parents’ expectations and students’ achievement in two western nations. Int J Behav Dev. 2007;31:594–602. [Google Scholar]
- [15].Blacher J, Kraemer BR, Howell EJ. Family expectations and transition experiences for young adults with severe disabilities: Does syndrome matter? Adv Ment Health Learn Disabil. 2009;4:3–16. [Google Scholar]
- [16].Grigal M, Neubert DA. Parents’ in-school values and post-school expectations for transition-aged youth with disabilities. Career Dev Transition Except Individ. 2004;27:65–85. [Google Scholar]
- [17].Bowen GL, Hopson LM, Rose RA, Glennie EJ. Students’ perceived parental school behavior expectations and their academic performance: A longitudinal analysis. Fam Relat. 2012;61:175–191. [Google Scholar]
- [18].Fredricks JA., Eccles JS. Children’s competence and value beliefs from childhood through adolescence: Growth trajectories in two male-sex-typed domains. Dev Psychol. 2002;38:519–533. [PubMed] [Google Scholar]
- [19].Mistry RS, White ES, Benner AD, Huynh VW. A longitudinal study of the simultaneous influence of mothers’ and teachers’ educational expectations on low-income youth’s academic achievement. J Youth Adol. 2009;38:826–838. [DOI] [PubMed] [Google Scholar]
- [20].Englund MM, Luckner AE, Whaley GJL, Egeland B. Children’s achievement in early elementary school: Longitudinal effects of parental involvement, expectations, and quality of assistance. J Educ Psychol. 2004;96:723–730. [Google Scholar]
- [21].Jacob MJ. Parental expectations and aspirations for their children’s educational attainment: An examination of the college-going mindset among parents (Doctoral dissertation). 2010. Retrieved from http://hdl.handle.net/11299/93924.
- [22].Zhang Y, Haddad E, Torres B, Chen C. The reciprocal relationships among parents’ expectations, adolescents’ expectations, and adolescents’ achievement: A two-wave longitudinal analysis of the NELS Data. J Youth Adol. 2011;40:479–489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Institute of Education Sciences. Welcome to NLTS2. Retrieved from www.nlts2.org
- [24].Javitz H, Wagner M. Analysis of potential bias in the sample of local education agencies (leas) in the national longitudinal transition study-2 (nlts2) sample. 2003. Menlo Park, CA: SRI International. [Google Scholar]
- [25].Carnevale AP, Rose SJ, Cheah B. The college payoff: Education, occupation, and lifetime earnings. 2011; Georgetown University Center on Education and the Workforce. Retrieved from https://repository.library.georgetown.edu/bitstream/handle/10822/559300/collegepayoff-complete.pdf?sequence=1
- [26].Martinez DC, Conroy JW, Cerreto MC. Parent involvement in the transition process of children with intellectual disabilities: The influence of inclusion on parent desires and expectations for postsecondary education. J Policy Pract Intel Disabil. 2012;9:279–288. [Google Scholar]
- [27].Kirby AV, Bagatell N, & Baranek GT The formation of postsecondary expectations among parents of youth with autism spectrum disorder. Focus Autism Other Dev Disabil. 2019; Early Online: 10.1177/1088357619881221 [DOI] [Google Scholar]
- [28].Schneider J, Wedgewood N, Llewellyn G, McConnell D. Families challenged by and accommodating to the adolescent years. J Intel Disabil Res. 2006;50:926–936. [DOI] [PubMed] [Google Scholar]
- [29].Carpenter DM, Geryk LL, Chen AT, Nagler RH, Dieckmann NF, Han PKJ. Conflicting health information: A critical research need. Health Expect. 2015;19:1173–1182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Wigfield A, Gladstone J. What Does Expectancy-value Theory Have to Say about Motivation and Achievement in Times of Change and Uncertainty? Motivation in Education at a Time of Global Change (Advances in Motivation and Achievement, Vol. 20), Emerald Publishing Limited; 2019. [Google Scholar]
- [31].National Center for Education Statistics. Table 302.10. Recent high school completers and their enrollment in 2-year and 4-year colleges, by sex: 1960 through 2015. 2016; Digest of Education Statistics. Retrieved from: https://nces.ed.gov/programs/digest/d16/tables/dt16_302.10.asp?current=yes
- [32].Fontenot K, Semega J, Kollar M. Income and Poverty in the United States: 2017. United States Census; 2018. [Google Scholar]
- [33].Yamamoto Y, Holloway SD. Parental expectations and children’s academic performance in sociocultural context. Educ Psychol Rev. 2010;22:189–214. [Google Scholar]
- [34].Eismann MM, Weisshaar R, Capretta C, Cleary DS, Kirby AV, Persch AC. Characteristics of students receiving occupational therapy services in transition and factors related to postsecondary success. Am J Occup Ther. 2017;71,7103100010. [DOI] [PubMed] [Google Scholar]
- [35].Plotner AJ, Trach JS, Strauser DR. Vocational rehabilitation counselors’ identified transition competencies: Perceived importance, frequency, and preparedness. Rehab Counsel Bull. 2012;55:135–143. [Google Scholar]
- [36].Individuals with Disabilities Education Improvement Act. H.R. 1350, 108th Congress;2004
- [37].Schwartz LA, Tuchman LK, Hobbie WL, Ginsberg JP. A socio-ecological model of readiness for transition to adult-oriented care for adolescents and young adults with chronic health conditions. Child Care Health Dev. 2011;37:883–895. [DOI] [PubMed] [Google Scholar]
- [38].Gorter JW, Stewart D, Woodbury-Smith M, King G, Wright M, Nguyen T, Freeman M, Swinton M. Pathways toward positive psychosocial outcomes and mental health for youth with disabilities: A knowledge synthesis of developmental trajectories. Can J Commun Ment Health. 2014;33:45–61. [Google Scholar]