View in EDS HTML Full Text PDF Full Text

Characterizing the Special Education Goals of Autistic Students: Latent Class Analysis with Demographic and Developmental Covariates

Saved in:
Bibliographic Details
Title: Characterizing the Special Education Goals of Autistic Students: Latent Class Analysis with Demographic and Developmental Covariates
Language: English
Authors: Matthew C. Zajic (ORCID 0000-0002-9806-4832), Juliette Gudknecht (ORCID 0000-0002-0352-1907), Nancy S. McIntyre (ORCID 0000-0001-5085-2672)
Source: Exceptional Children. 2025 91(4):359-376.
Availability: SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com
Peer Reviewed: Y
Page Count: 18
Publication Date: 2025
Sponsoring Agency: Institute of Education Sciences (ED)
Contract Number: R305B220021
Document Type: Journal Articles
Reports - Research
Education Level: Elementary Secondary Education
Descriptors: Autism Spectrum Disorders, Students with Disabilities, Special Education, Educational Objectives, Student Needs, Elementary Secondary Education, Student Characteristics, Individualized Education Programs, Symptoms (Individual Disorders), Language Skills, Literacy, Communication Skills, Academic Achievement
Assessment and Survey Identifiers: Vineland Adaptive Behavior Scales, Social Responsiveness Scale
DOI: 10.1177/00144029251331861
ISSN: 0014-4029
2163-5560
Abstract: Many autistic children receive special education services via Individualized Education Programs (IEPs) that include specific educational goals and needs. Prior research has examined programming options available to support autistic students, but less is known about their educational needs across academic and developmental educational goals. Additionally, existing approaches have often relied on small studies focused on describing individual goal areas. This study uses data from 551 families from across the United States with an autistic child in grades K-12 and latent class analysis to (a) identify latent, or underlying, subgroups based off multivariate response patterns across educational goals endorsed in six domains (reading, writing, math, language and communication, social skills, behavior), and (b) examine if demographic and developmental covariates predict latent class membership. We identified five latent classes: "All Goals" (40.49%); "Autistic Characteristics" (21.63%); "Language, Literacy, and Autistic Characteristics" (18.99%); "Academic" (13.94%); and "Language and Communication" (4.95%). Two covariates--percentage of time spent in general education and adaptive behavior--predicted differences in latent class membership. Findings offer a comprehensive examination into the heterogeneous educational needs of autistic school-age children. Our results emphasize the need for researchers and educators to understand the educational needs of autistic students beyond the presence of an IEP.
Abstractor: As Provided
IES Funded: Yes
Entry Date: 2025
Accession Number: EJ1475373
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwHJmCQSqwCyMwIWFJsFxxcNAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDNNIAI0hMsxDk1fWlgIBEICBm6eIJnTswTMtUlN0D_l-5Rgm2l2SK2y71suA63DNRdCdR-Hg-p0oZDS-M65dR5W4kPjCchzTNpDVHPIlXZztrOylhV5ETlJtiS49vFYLCJYiKK1kU0rRKNrfJsT4AoBTNHykIqsH7pAo3x7uSI2XM_T9pfMH5o2H8etuATmSGPcuTTqEeEFRz2jheljjC2MtIw4ijLlwy_fGMRq1
Text:
  Availability: 1
  Value: <anid>AN0186128908;exc01jul.25;2025Jun26.00:55;v2.2.500</anid> <title id="AN0186128908-1">Characterizing the Special Education Goals of Autistic Students: Latent Class Analysis With Demographic and Developmental Covariates </title> <p>Many autistic children receive special education services via Individualized Education Programs (IEPs) that include specific educational goals and needs. Prior research has examined programming options available to support autistic students, but less is known about their educational needs across academic and developmental educational goals. Additionally, existing approaches have often relied on small studies focused on describing individual goal areas. This study uses data from 551 families from across the United States with an autistic child in grades K-12 and latent class analysis to (a) identify latent, or underlying, subgroups based off multivariate response patterns across educational goals endorsed in six domains (reading, writing, math, language and communication, social skills, behavior), and (b) examine if demographic and developmental covariates predict latent class membership. We identified five latent classes: All Goals (40.49%); Autistic Characteristics (21.63%); Language, Literacy, and Autistic Characteristics (18.99%); Academic (13.94%); and Language and Communication (4.95%). Two covariates—percentage of time spent in general education and adaptive behavior—predicted differences in latent class membership. Findings offer a comprehensive examination into the heterogeneous educational needs of autistic school-age children. Our results emphasize the need for researchers and educators to understand the educational needs of autistic students beyond the presence of an IEP.</p> <p>Keywords: autism; education; educational goals; individualized education programs; latent class analysis; literacy</p> <p>Autism is a neurodevelopmental condition that is understood diagnostically as affecting social communication and behavioral development across the lifespan ([<reflink idref="bib1" id="ref1">1</reflink>]). The services and supports required for many autistic individuals vary across individuals and contexts, and better understanding the needs of autistic individuals can lead to developing more effective programs throughout the school-age years ([<reflink idref="bib14" id="ref2">14</reflink>]). A large proportion of autistic individuals require formalized educational supports and services; approximately 12% of children (∼876,000) served under the Individuals with Disabilities Education Act (IDEA) receive educational services for an autism disability classification ([<reflink idref="bib12" id="ref3">12</reflink>]). Current prevalence rates in the United States estimate that 1 in 36 school-age children have an autism diagnosis ([<reflink idref="bib19" id="ref4">19</reflink>]), and [<reflink idref="bib2" id="ref5">2</reflink>] estimated an IDEA prevalence of 1 in 91. Access to effective educational services that align with child needs is critical to ensure that the academic and functional needs of autistic individuals are being met.</p> <p>A medical diagnosis of autism does not determine eligibility for an autism educational classification. Students who have an autism diagnosis must participate in the school evaluation process to qualify for special education services ([<reflink idref="bib32" id="ref6">32</reflink>]). Importantly, an educational classification determines if a child is eligible for receiving services, not the type of services they will receive. Determination of services is done through the Individualized Education Program (IEP) process, which requires a collaborative team (e.g., educational professionals, the child's caregiver, and oftentimes the child) to interpret a child's evaluation report and recommend appropriate services and goals for the child. As part of this process, the collaborative team must specify operationalized, ambitious, measurable, and individualized short- and long-term annual goals that detail student achievement expectations within the next year ([<reflink idref="bib32" id="ref7">32</reflink>]; [<reflink idref="bib41" id="ref8">41</reflink>]). These goals can include both functional and academic goals that are designed to help the student be involved and make progress in the general education curriculum ([<reflink idref="bib11" id="ref9">11</reflink>]; [<reflink idref="bib39" id="ref10">39</reflink>]). Knowing that a child has an autism educational classification provides limited insight into that child's educational needs ([<reflink idref="bib40" id="ref11">40</reflink>]).</p> <p>Existing work examining the IEPs and educational needs of autistic children remains a growing area of inquiry. Prior studies have focused generally on the prevalence of IEPs and related special education services among autistic students. For example, [<reflink idref="bib17" id="ref12">17</reflink>] reported that 75.7% of their sample of school-age autistic children had an IEP in place but made no mention of documented goals or services. [<reflink idref="bib21" id="ref13">21</reflink>] reviewed the IEPs of 89 autistic children with and without intellectual disability to find most children received speech and occupational therapies (70.8% and 56.2%, respectively) relative to children receiving social skills instruction or having behavior plans in place (16.9% and 15.7%, respectively). While services were examined, the authors did not mention specific IEP goals. [<reflink idref="bib24" id="ref14">24</reflink>] leveraged a large pre-existing dataset from the National Survey of Children's Health to examine demographic factors associated with special education services among autistic youth. Approximately 73.26% of their sample (<emph>n </emph>= 2,379) received special education services, and they found that receipt of services was positively associated with parent-reported autism severity levels (odds ratio [<emph>OR</emph>] = 1.75) and a co-occurring intellectual disability (<emph>OR</emph> = 2.76) and negatively associated with gender identity (female; <emph>OR</emph> = 0.74), age (<emph>OR</emph> = 0.96), and parent education level (less than high school relative to college or higher; <emph>OR</emph> = 0.52). However, data regarding specific services or goals were unavailable.</p> <p>Further characterizing the educational goals present among autistic students may help educators better understand the variable academic and developmental needs present among autistic individuals. Autistic children demonstrate highly heterogeneous academic skill profiles, particularly in the areas of reading, writing, and mathematics (e.g., [<reflink idref="bib3" id="ref15">3</reflink>]; [<reflink idref="bib22" id="ref16">22</reflink>]; [<reflink idref="bib43" id="ref17">43</reflink>]). Findings from academic achievement research complement existing research related to language and communication, social, and behavioral development that highlight variable strengths and challenges that can impact learning in educational settings (Ibrahim, [<reflink idref="bib10" id="ref18">10</reflink>]; [<reflink idref="bib13" id="ref19">13</reflink>]). Better characterizing the heterogeneity present among these domains may further knowledge about how underlying heterogeneous factors ([<reflink idref="bib18" id="ref20">18</reflink>]) impact autistic individuals' educational needs. This is particularly important given that language, social communication, and behavior are foundational to developing more advanced academic skills (e.g., [<reflink idref="bib15" id="ref21">15</reflink>]; [<reflink idref="bib36" id="ref22">36</reflink>]; [<reflink idref="bib43" id="ref23">43</reflink>]). While educators may think of the common diagnostic areas associated with autism (i.e., social skills and behavior), autistic children may require supports across a wide range of goal areas, including communication and academics ([<reflink idref="bib40" id="ref24">40</reflink>]).</p> <p>Characterizing educational needs also requires considering how needs may be differentiated by demographic and developmental factors. Existing research has found that such factors may impact the receipt of special education services (e.g., [<reflink idref="bib21" id="ref25">21</reflink>]; [<reflink idref="bib24" id="ref26">24</reflink>]; [<reflink idref="bib38" id="ref27">38</reflink>]). [<reflink idref="bib38" id="ref28">38</reflink>] systematically reviewed 36 articles to characterize the presence of differences and disparities present in school-based services provided to autistic students. For child-level demographic factors, they found a paucity of studies focusing on differences among older students (e.g., secondary and transition-age youth) as well as mixed findings for race/ethnicity (with differences in receipt of school-based service types and quality not consistently associated with race or ethnicity). For family-level demographic factors, they identified socioeconomic status to be a significant factor impacting service characteristics (noting a positive association between income and service receipt). Another important demographic variable at the school level is the amount of time children spend in general education contexts. According to recent estimates from the 2019–2021 school years, autistic children are distributed across multiple educational settings, with about one-third spending limited time (< 40%) in general education settings (33.2%–34.1%), one-fifth spending some of their time (40%–79%) in general education settings (17.2%–18.4%), and the remainder spending most of their time (≥ 80%) in general education settings (40.1%–40.8%; [<reflink idref="bib27" id="ref29">27</reflink>]). Placement decisions are made based on numerous child-, school-, and state-level factors that may impact further educational decision making ([<reflink idref="bib16" id="ref30">16</reflink>]).</p> <p>Developmental factors include characteristics specific to autism as well as the presence of co-occurring diagnostic conditions. The presence of co-occurring difficulties in cognition, language, and attention may contribute to identified academic difficulties and the types of services autistic children receive ([<reflink idref="bib6" id="ref31">6</reflink>]; [<reflink idref="bib13" id="ref32">13</reflink>]). Co-occurring conditions may be associated with academic achievement difficulties as well, particularly concerning the presence of attentional difficulties (e.g., [<reflink idref="bib3" id="ref33">3</reflink>]; [<reflink idref="bib42" id="ref34">42</reflink>]). Executive function challenges have been noted to contribute to the educational services received by middle school autistic students ([<reflink idref="bib5" id="ref35">5</reflink>]). The presence of a significant co-occurring cognitive difficulty like intellectual disability has been associated with a generally higher rate of special education service receipt (e.g., [<reflink idref="bib21" id="ref36">21</reflink>]; [<reflink idref="bib24" id="ref37">24</reflink>]), although the specific types of services are less established. Services and needs may also be impacted by the presence of a language or communication disorder or variable adaptive behavior skills, as difficulties across these domains may impact educational access ([<reflink idref="bib6" id="ref38">6</reflink>]; [<reflink idref="bib13" id="ref39">13</reflink>]). Given the expectation for heterogeneity among autistic individuals (e.g., [<reflink idref="bib18" id="ref40">18</reflink>]), educators need to take into consideration how developmental factors may differ among autistic children based on their educational needs.</p> <hd id="AN0186128908-2">Current Study</hd> <p>The current study characterizes the educational goals of autistic children who receive special education services (i.e., who have an IEP) and examines if demographic and developmental factors differentiate patterns of educational needs. To capitalize on the expected heterogeneity present in autistic children's educational goals, we draw on mixture modeling, a person-centered statistical approach useful for identifying and analyzing homogeneous subgroups present within a heterogeneous population. Existing studies have typically used variable-centered approaches that assume individuals are drawn from a single population and examine relationships within that population ([<reflink idref="bib9" id="ref41">9</reflink>]). In contrast, mixture modeling is a person-centered approach that identifies underlying subgroups based on response patterns in observed data ([<reflink idref="bib20" id="ref42">20</reflink>]; [<reflink idref="bib28" id="ref43">28</reflink>]). Individuals within the same identified subgroup tend to have similar response patterns, while those in different subgroups typically show distinct response patterns from one another. Mixture modeling also allows for examining how variables not directly used to identify the subgroups (e.g., auxiliary variables or covariates) may predict subgroup membership (e.g., [<reflink idref="bib29" id="ref44">29</reflink>]).</p> <p>Our study is guided by two broad aims. Our first aim is to characterize the educational goals in the domains of academics (reading, writing, and mathematics), language and communication, social skills, and behavior by using latent class analysis (a type of mixture modeling where all observed variables used to estimate the underlying subgroups are dichotomous). Our second aim is to examine if demographic (age, gender, race/ethnicity, mother's level of education, time in general education classrooms) and developmental (autistic characteristics, adaptive behavior, and presence of co-occurring intellectual disability, language disorder, or attention-deficit/hyperactivity disorder [ADHD]) covariates predict subgroup membership.</p> <hd id="AN0186128908-3">Methods</hd> <p>Data come from an online survey that examined education, literacy development, and instructional practices between 2019 and 2021. All participating families lived in the United States and reported on a school-age child with a professional autism diagnosis. The survey was cross sectional, and all the data analyzed in this paper asked caregivers (parents or legal guardians) to reflect on the 2019–2020 school year before the COVID-19 pandemic. As a result, the data do not account for any impact of COVID-19 on children's special education services. The full survey took families approximately 1–2 hr to finish, and families could complete the survey over multiple sessions. All families received a $25 gift card upon completion of the survey.</p> <p>Participating families were recruited in collaboration with the Simons Foundation Powering Autism Research for Knowledge (SPARK) Research Match program offered by the Simons Foundation Autism Research Initiative (SFARI). SPARK Research Match facilitates matching researchers with families with an autistic individual who live within the United States with relevant, approved research projects ([<reflink idref="bib33" id="ref45">33</reflink>]). All families enrolled in SPARK have a family member with a professional autism diagnosis. The research team sought Institutional Review Board approval and SPARK study team approval before beginning study recruitment.</p> <p>Recruitment occurred separately for two age cohorts throughout 2021 and 2022. Caregivers for the elementary-age cohort completed the survey between September 2021 and December 2021. The SPARK study team sent 7,718 emails to eligible families concerning interest in participating in the study. A total of 1,569 families expressed interest, and 1,177 families provided contact information to the research study team. We contacted 935 families, with 563 families consenting to participate and 402 completing the study. Caregivers for the secondary-age cohort completed the survey between January 2022 and April 2022. The SPARK study team sent 10,896 emails to eligible families concerning interest in participating in the study. A total of 1,837 families expressed interest, and 1,445 families provided contact information to the research study team. We contacted 1,393 families, with 585 families consenting to participate and 409 completing the study.</p> <hd id="AN0186128908-4">Study Sample</hd> <p>From the 811 completed surveys, we excluded participants for the following reasons. First, we excluded children who were educated in a non-public school setting (<emph>n</emph> = 117). Second, we excluded children who were enrolled in preschool or who were reported as being 5 or younger during the 2019–2020 school year (<emph>n </emph>= 19). Third, we excluded children who did not have an IEP during the 2019–2020 school year (<emph>n </emph>= 105). Fourth, we excluded participants who participated in the secondary cohort after having already participated in the elementary cohort for the same child (<emph>n</emph> = 3). Fifth, we excluded children who had missing data on our selected covariate variables; this included missing data for race or ethnicity (<emph>n</emph> = 2), classroom placement (<emph>n</emph> = 3), and adaptive behavior (<emph>n </emph>= 4). We also excluded seven children who identified as non-binary due to the limited sample size. In total, we excluded 260 children, but only nine children were excluded due to missing data.</p> <p>Our analytical sample included information about 551 autistic children who were receiving special education services (Table 1). Our sample was predominantly male (79.67%) with about two-thirds identifying as White (66.61%). All children were in the K-12 grades (K-3<sups>rd </sups>= 33.76%, 4–6<sups>th </sups>= 31.76%, 7–8<sups>th </sups>= 18.15%, 9–12<sups>th </sups>= 16.33%), were approximately equally distributed for free or reduced-price lunch (FRPL) eligibility (47.91% eligible), and spent a mixed amount of time in general education contexts (< 40% = 39.02%, 40–79% = 13.97%, ≥ 80% = 47.01%). Families lived across the United States (Northeast = 20.33%, Midwest = 25.05%, South = 33.03%, West = 21.60%, based on U.S. Census categories), and mothers ranged in their levels of education (< College = 28.86%, College = 43.92%, > College = 27.22%). A small subset of participants had a parent-reported co-occurring diagnosis of intellectual disability (19.78%) or a co-occurring language or communication disorder (10.89%), while just over half (56.99%) had a parent-reported co-occurring ADHD diagnosis.</p> <p>Table 1. Full Sample (n = 551) Descriptive Statistics Based on Covariate Coding.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /></colgroup><thead><tr><th align="left">Variable and Coding</th><th align="left">M (<italic>SD</italic>)/<italic>n</italic> (%)<sup>a</sup></th></tr></thead><tbody><tr><td colspan="2"><italic>IEP Goal Domain</italic> (0 = no goals; 1 = ≥ 1 goal)</td></tr><tr><td> Reading</td><td>172 (31.22%), 379 (68.78%)</td></tr><tr><td> Writing</td><td>142 (25.77%), 409 (74.23%)</td></tr><tr><td> Math</td><td>214 (38.84%), 337 (61.16%)</td></tr><tr><td> Language and Communication</td><td>142 (25.77%), 409 (74.23%)</td></tr><tr><td> Social Skills</td><td>104 (18.87%), 447 (81.13%)</td></tr><tr><td> Behavior</td><td>235 (42.65%), 316 (57.35%)</td></tr><tr><td colspan="2"><italic>Demographics</italic></td></tr><tr><td> Age (Years)</td><td>10.53 (3.10)</td></tr><tr><td> Gender Identity (0 = male, 1 = female)</td><td>439 (79.67%), 112 (20.33%)</td></tr><tr><td> Grade (0 = elementary, 1 = secondary)</td><td>361 (65.52%), 190 (34.48%)</td></tr><tr><td> Race/Ethnicityb (0 = White, 1 = POC, 2 = multiracial [White + POC])</td><td>367 (66.61%), 138 (25.05%), 46 (8.35%)</td></tr><tr><td> %Time Gen Ed (0 = < 40%, 1 = 40%–79%, 2 = ≥ 80%)</td><td>215 (39.02%), 77 (13.97%), 259 (47.01%)</td></tr><tr><td> Free/Reduced-price Lunch (0 = not eligible, 1 = eligible)</td><td>287 (52.09%), 264 (47.91%)</td></tr><tr><td> Mother's Level of Education (0 = ≤ college, 1 = > college)</td><td>401 (72.78%), 150 (27.22%)</td></tr><tr><td colspan="2"><italic>Developmental</italic></td></tr><tr><td> Intellectual Disability (0 = no, 1 = yes)</td><td>442 (80.22%), 109 (19.78%)</td></tr><tr><td> Language/Communication Disorder (0 = no, 1 = yes)</td><td>491 (89.11%), 60 (10.89%)</td></tr><tr><td> ADHD (0 = no, 1 = yes)</td><td>237 (43.01%), 314 (56.99%)</td></tr><tr><td> Autism Characteristics (T Score, SRS-2 Total Score)</td><td>77.77 (11.44)</td></tr><tr><td> Adaptive Behavior (Standard Score, Vineland-3 ABC)</td><td>71.39 (15.01)</td></tr><tr><td> Socialization (Standard Score, Vineland-3)</td><td>66.76 (14.37)</td></tr><tr><td> Daily Living Skills (Standard Score, Vineland-3)</td><td>74.05 (20.28)</td></tr><tr><td> Communication (Standard Score, Vineland-3)</td><td>75.13 (18.74)</td></tr></tbody></table> </ephtml> </p> <p>1 <emph>Note.</emph> POC = person of color. SRS-2 = Social Responsiveness Scale, 2<sups>nd</sups> Edition. Vineland-3 ABC = Adaptive Behavior Composite, Vineland, 3<sups>rd</sups> Edition. ADHD = attention-deficit/hyperactivity disorder. IEP = Individualized Education Program. %Time Gen Ed = Percentage of time a child spends in general education.</p> <p>2 Mean and standard deviations provided for continuous variables. Frequencies and percentages provided for categorical variables. <sups>b</sups>Non-exclusive categories for race/ethnicity are as follows: White (<emph>n</emph> = 493, 89%), Hispanic/Latinx (<emph>n</emph> = 91, 16%), Black (<emph>n</emph> = 55, 10%), Asian (<emph>n</emph> = 43, 8%), American Indian or Alaskan Native (<emph>n</emph> = 12, 2%), Native Hawaiian or Pacific Islander (<emph>n</emph> = 2, 0.3%). Multiracial represents families who endorsed White with a non-White racial/ethnic identity.</p> <hd id="AN0186128908-5">Study Variables</hd> <p>Study variables included information related to IEP goals as well as demographic and developmental factors. All data were collected from caregivers from the online survey.</p> <hd id="AN0186128908-6">IEP Goal Domains</hd> <p> <emph>IEP goal domains</emph> were collected by asking caregivers if their child (a) had an IEP in place and (b) what specific goal areas were covered by that IEP. Caregivers were provided with a multiple-option checklist with choices for reading, writing, mathematics, language and communication, social skills, and behavior. An "Other" option was also provided, and relevant answers were cleaned for inclusion. Caregivers endorsed their child as having an IEP goal in a skill domain if their child had at least one IEP goal; therefore, the data are dichotomous as either having no IEP goals (0) or having at least one IEP goal (<reflink idref="bib1" id="ref46">1</reflink>) in that skill domain.</p> <hd id="AN0186128908-7">Demographic Variables</hd> <p>Caregivers provided information about the following demographic variables. <emph>Age</emph> represented the child's age in 2019. <emph>Gender identity</emph> asked respondents to select male, female, or other/non-binary options for their child. <emph>Race/Ethnicity</emph> asked respondents to check all potential racial and ethnic identities that applied to their child (White, Hispanic/Latinx, Black, Asian, American Indian or Alaskan Native, and Native Hawaiian or Pacific Islander). Respondents were provided with blank answer boxes to provide additional identities. <emph>Grade</emph> asked respondents to designate their child's grade level (kindergarten through 12<sups>th</sups> grade). <emph>FRPL</emph> asked respondents to designate whether their child qualified for FRPL during the 2019–2020 school year. <emph>Mother's level of education (MLE)</emph> asked respondents to designate the current level of education for the child's mother on the following scale: (a) less than high school completion, (b) high school graduate, (c) some college, (d) associate degree, (e) bachelor's degree, (f) master's degree, (g) doctoral degree, and (h) professional degree (e.g., JD, MD). If a family had multiple mothers, then the highest level of education was used. <emph>Percentage of time in general education</emph> (% <emph>Time Gen Ed</emph>) required respondents to indicate on a scale from 0% to 100% the amount of time their child spent in general education settings, special education settings, and other educational settings during a typical period (either day or week, whichever was easier for the caregiver to consider). Percentages across settings were required to sum to 100%. For this study, we are using the percentage of time the caregiver reported their child was in a general education setting.</p> <hd id="AN0186128908-8">Developmental Variables</hd> <p>Caregivers provided information about the following developmental characteristics. Children's adaptive behavior was assessed using the parent/caregiver form of the Vineland Adaptive Behavior Scales, Third Edition (Vineland-3; [<reflink idref="bib37" id="ref47">37</reflink>]). Caregivers completed the Domain-Level forms for Socialization, Daily Living Skills, and Communication. Most items are scored on a 0–2 rating scale with some items forced to a dichotomous 0/2 rating. Socialization contains items related to interpersonal relationships, play and leisure, and coping skills subdomains. Daily Living Skills contains items related to personal, domestic, numeric, community, and school community subdomains. Communication contains items related to receptive, expressive, and written subdomains. All Domain-Level forms are reported as standard scores (<emph>M </emph>= 100, <emph>SD </emph>= 15). We generated an Adaptive Behavior Composite (ABC) standard score based on responses across all three forms. Sample-specific internal consistencies across domain and ABC scores were excellent (αs =.95–.98) and consistent with publisher-reported estimates (αs =.94–.98; [<reflink idref="bib37" id="ref48">37</reflink>]). Children's autism characteristics were collected using the Social Responsiveness Scale, 2<sups>nd</sups> Edition (SRS-2; [<reflink idref="bib4" id="ref49">4</reflink>]). The SRS-2 is a 65-item scale that provides a Total Score on a T-score scale (<emph>M </emph>= 50, <emph>SD </emph>= 10) that broadly measures characteristics associated with autism. Internal consistency reliability was great (.87) and consistent with publisher estimates (.95; [<reflink idref="bib4" id="ref50">4</reflink>]). Co-occurring diagnostic information was collected via caregiver report.</p> <hd id="AN0186128908-9">Data Analysis</hd> <p>Research aims were addressed by using latent class analysis. We used the <emph>MplusAutomation</emph> package ([<reflink idref="bib8" id="ref51">8</reflink>]) in R ([<reflink idref="bib30" id="ref52">30</reflink>]) to fit all latent class models and examined all fit statistics via R Studio using M<emph>plus</emph> (Version 8.10; [<reflink idref="bib25" id="ref53">25</reflink>]). The general procedure for latent class analysis involves class enumeration (deciding on the appropriate number of classes), classification quality (accuracy and precision), and auxiliary variable testing ([<reflink idref="bib28" id="ref54">28</reflink>]).</p> <hd id="AN0186128908-10">Class Enumeration and Classification Accuracy</hd> <p>We examined class enumeration by fitting a series of latent class models with the same indicator variables that began with a 1-class model and continued with subsequent models that added one new class per model. This continued until either a lack of meaningful increase in model fit or a convergence error occurred. To select a final model, we drew on empirical findings from fit indices and theoretical justification for the selected model (as statistical evidence alone cannot justify an appropriate final model). As there is no single fit index that determines a best-fitting latent class model (see [<reflink idref="bib28" id="ref55">28</reflink>]), we examined a range of information criteria—Bayesian Information Criterion (BIC), Sample-size adjusted BIC (aBIC), Consistent Akaike Information Criterion (CAIC), and Approximate Weight of Evidence Criterion (AWE)—and likelihood-based tests—Vuong-Lo-Mendell-Rubin adjusted likelihood ratio test (VLMR-LRT), Bootstrapped likelihood ratio test (BLRT), Bayes Factor (<emph>BF</emph>), and the correct model probability (<emph>cmP<subs>k</subs></emph>). Both the VLMR-LRT and BLRT compare a current model with a prior model, with a non-significant <emph>p</emph> value (>.05) endorsing the model with the fewer number of classes. The <emph>BF</emph> conducts a pairwise comparison between neighboring class models and provides a ratio of the probability that each model being compared is the "true" model; <emph>BFs</emph> can be interpreted as weak support (1 < <emph>BF </emph>< 3), moderate support (3 < <emph>BF </emph>< 10), and strong support (<emph>BF </emph>> 10) for the model with fewer classes. The <emph>cmP<subs>k</subs></emph> estimates the correctness for each model out of all tested models; the model with the largest value is endorsed as the "true" model. (See [<reflink idref="bib20" id="ref56">20</reflink>] and [<reflink idref="bib28" id="ref57">28</reflink>] for more information about fit statistics.)</p> <p>We then examined classification accuracy and precision for all models under potential consideration ([<reflink idref="bib26" id="ref58">26</reflink>]). The average posterior probability of correct classification for each class (<emph>AvePP<subs>k</subs></emph>) provides a classification metric that describes the proportion that is correctly classified in each class. Perfect classification is when an average is equal to 1, and values greater than.70 are deemed as adequate. The odds of correct classification ratio (<emph>OCC<subs>k</subs></emph>) provides a class-specific classification accuracy measure that shows the odds of being in a class relative to classification based on random assignment. Larger values of <emph>OCC<subs>k</subs></emph> (> 5) indicate high classification accuracy. We also compared the class proportions to the modal class assignment proportions (<emph>mcaP<subs>k</subs></emph>) to estimate classification error (as <emph>mcaP<subs>k</subs></emph> should fall within 90% of the bootstrapped confidence interval for the class proportions). Modal class membership refers to assigning an individual to only their most likely latent class. We examined entropy as a measure of overall classification precision for the model; larger values (>.80) indicate high classification accuracy. Lastly, we calculated frequently occurring response patterns across classes to examine how well-aligned response patterns were for estimated posterior class probabilities (e.g., the probability that an individual belongs to each latent class) in relation to modal class assignments. Tables for classification accuracy and precision estimates are provided in the supplemental files.</p> <hd id="AN0186128908-11">Including and Testing Auxiliary Variables</hd> <p>After a final model was selected, we examined if demographic and developmental variables predicted membership differences between latent classes. Estimating auxiliary variable relations with latent class variables can be complicated depending on several involved factors, including the scale of the auxiliary variables and the proposed analytical design (see [<reflink idref="bib29" id="ref59">29</reflink>]). To assist with interpretation, we dichotomized multi-categorical variables (see Table 1) except for % Time Gen Ed. Demographic auxiliary variables included age, gender, race/ethnicity, % Time Gen Ed, FRPL, and MLE. Developmental auxiliary variables included co-occurring intellectual disability, language or communication disorder, or ADHD; and Vineland-3 ABC. We removed grade level and SRS-2 Total Score from covariate testing due to strong intercorrelations with age (<emph>r </emph>=.98) and Vineland-3 ABC (<emph>r </emph>= −.60), respectively. Continuous variables (age, ABC) were grand-mean centered.</p> <p>We used the manual ML three-step approach to fit all covariates to the selected final model. The first step involves saving the posterior probabilities and modal class assignment from the best-fitting unconditional model. The second step computes estimated conditional probabilities for modal class assignments given their true latent class memberships. This allows for quantifying the estimated average classification errors for the modal class assignment, which provides fixed parameter values that describe the relationship between the latent class variable and the multinomial class assignment variable. The third step specifies a new analytic model that uses the modal class assignment from the first step with classification error (class-specific parameters) fixed based on the second step (see [<reflink idref="bib29" id="ref60">29</reflink>]). We then regressed the latent class variable onto the predictors, producing a multinomial logistic regression that included all covariate variables simultaneously to identify class-specific covariate differences. We tested the effect of each covariate variable by using a Wald test before examining differences between classes based on their logit (and accompanying <emph>OR</emph>) for each statistically significant covariate (<emph>p </emph><.05). M<emph>plus</emph> reparametrizes multinomial regression results based on different reference classes, and we provide results for all comparisons.</p> <hd id="AN0186128908-12">Results</hd> <p>Model fit indices are provided in Table 2. The BIC, aBIC, CAIC, and <emph>cmP<subs>k</subs></emph> endorsed a 3-class solution. The AWE suggested a 2-class solution. Most information criteria was minimized between the 3-, 4-, and 5-class solutions. The BLRT endorsed a 5-class solution. The VLMR-LRT endorsed a 4-class solution. The <emph>BF</emph> supported models from a 3-class through a 5-class solution. We selected the 3-, 4-, and 5-class solutions as candidate models.</p> <p>Table 2. Model Fit Summary Table.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /></colgroup><thead><tr><th align="left">Classes</th><th align="left">Par</th><th align="left">LL</th><th align="left">BIC</th><th align="left">aBIC</th><th align="left">CAIC</th><th align="left">AWE</th><th align="left">BLRT</th><th align="left">VLMR</th><th align="left">BF</th><th align="left">cmP<italic><sub>k</sub></italic></th></tr></thead><tbody><tr><td>1-Class</td><td>6</td><td>−1,981.86</td><td>4,001.59</td><td>3,982.55</td><td>4,007.59</td><td>4,057.46</td><td>-</td><td>-</td><td>0.00</td><td><.001</td></tr><tr><td>2-Class</td><td>13</td><td>−1,743.14</td><td>3,568.33</td><td>3,527.06</td><td>3,581.33</td><td><bold>3,689.38</bold></td><td><.001</td><td><.001</td><td>0.00</td><td><.001</td></tr><tr><td>3-Class</td><td>20</td><td>−1,704.70</td><td><bold>3,535.64</bold></td><td><bold>3,472.16</bold></td><td><bold>3,555.64</bold></td><td>3,721.88</td><td><.001</td><td><.001</td><td><bold>> 10</bold></td><td><bold>1.00</bold></td></tr><tr><td>4-Class</td><td>27</td><td>−1,696.54</td><td>3,563.49</td><td>3,477.78</td><td>3,590.49</td><td>3,814.91</td><td>.05</td><td><bold>.02</bold></td><td><bold>> 10</bold></td><td><.001</td></tr><tr><td>5-Class</td><td>34</td><td>−1,687.85</td><td>3,590.31</td><td>3,482.38</td><td>3,624.30</td><td>3,906.90</td><td>.04</td><td>.17</td><td><bold>> 10</bold></td><td><.001</td></tr></tbody></table> </ephtml> </p> <p>3 <emph>Note.</emph> Par = parameters. LL = model log likelihood. BIC = Bayesian information criterion. aBIC = sample-size adjusted BIC. CAIC = consistent Akaike information criterion. AWE = approximate weight of evidence criterion. BLRT = bootstrapped likelihood ratio test <emph>p</emph> value. VLMR = Vuong-Lo-Mendell-Rubin adjusted likelihood ratio test <emph>p</emph> value. BF = Bayes factor. cmP<emph><subs>k </subs></emph>= approximate correct model probability. Best fitting statistical evidence bolded for each fit statistic.</p> <p>To select a final model, we considered the above-mentioned fit indices as well as model parsimony and interpretability. Model classification diagnostics for all three candidate models appeared overall similar and provided limited information to differentiate between the models (see supplemental files). All three models had <emph>AvePP<subs>k</subs>s</emph> above the recommended.70 threshold (3-class =.82–.94; 4-class:.73–.94; 5-class:.83–.92). All models had <emph>mcaP<subs>k</subs></emph> estimates within the 95% CIs of the modal class proportions. All models had <emph>OCC<subs>k</subs></emph> estimates above 5 for each class. All models had similar entropy estimates (3-class = 0.77; 4-class = 0.81; 5-class = 0.78). Therefore, we examined item probability plots. All three models had a class with high probabilities for all IEP goal domain areas (termed <emph>All Goals</emph>; 3-class: 50.36%, 4-class: 51.16%; 5-class: 40.49%). All three models had a class with low probabilities in academic, heterogeneous probabilities in language and communication, and elevated probabilities in social skills and behavior IEP goal domain areas (termed <emph>Autistic Characteristics</emph> [<emph>Aut</emph>]; 3-class: 31.13%; 4-class: 28.71%; 5-class: 21.63%). All three models had a class with high probabilities in the academic IEP goal domain areas with lower probabilities in the language, social skills, and behavior IEP goal domain areas (termed <emph>Academic</emph>; 3-class: 18.51%; 4-class: 13.66%; 5-class: 13.94%). A key difference between the three models in terms of this latent class is that the 3-class solution kept language likelihood more similar to academic domains, whereas the 4- and 5-class solutions created a larger separation between language and academic skills (with language being more in line with social skills). The 4- and 5-class solutions had a class with elevated writing and language as well as variable reading IEP goal domain areas with diminished mathematics, social skills, and behavior areas (termed <emph>Language and Communication [LangComm]</emph>; 4-class: 6.47%; 5-class: 4.95%). One key difference is that the 5-class solution had a far lower endorsement likelihood for the social skills IEP goal domain area relative to the 4-class solution (.08 vs..45). A potential reason for this is the new class added in the 5-class model had elevated reading, writing, and language while also having elevated social skills and behavior IEP goal domain areas endorsement (termed <emph>Language, Literacy, and Autistic Characteristics</emph> [<emph>LangLitAut</emph>]; 5-class: 18.99%).</p> <p>Given this information, we selected the 5-class model as the final model due to the novel information added by the extra classes in the 5-class model compared to the 3- and 4-class models. The 5-class model solution included the classes observed in the 3- and 4-class models solutions, suggesting no substantial loss of information. Furthermore, the 5-class model identified an additional class that both reduced the largest, most general class (<emph>All Goals</emph>) and created a novel, substantial additional class (<emph>LangLitAut</emph>). While the additional class added in the 4-class model solution appears similar to the one added in the 5-class model solution, it is important to note the additional presence of a high probability on goals associated with autistic characteristics in the latter versus the former. This suggests two distinct classes may be present in relation to autistic children who have language and literacy goals, with the smaller group consisting of those who predominantly have at least one language goal and minimal goals associated with autism characteristics (social skills and behavior).</p> <hd id="AN0186128908-13">Model Interpretation</hd> <p>Conditional item probabilities for the 5-class solution are presented Figure 1. We reviewed class homogeneity by examining conditional item probabilities to identify indicators that characterized each class by using the heuristic of <.30 (low probability of endorsement) and >.70 (high probability of endorsement) as indicating high homogeneity ([<reflink idref="bib20" id="ref61">20</reflink>]). Item probabilities between these two extremes may be considered more heterogeneous, meaning there is endorsement variability among class members. The largest class (<emph>All Goals</emph>; 40.49%) showed high endorsement probability for all indicators. The second largest class (<emph>Aut</emph>; 21.63%) showed high endorsement probability for social skills and low endorsement probability for reading, writing, and math. This class appeared heterogeneous for language (.43) and behavior (.60) endorsement. The third largest class (<emph>LangLitAut</emph>; 18.99%) showed elevated endorsement probability for writing, language, and social skills and low endorsement probability for math. This class appeared heterogeneous for behavior (.60) and reading (.68) endorsement. The fourth largest class (<emph>Academic</emph>; 13.94%) showed high endorsement probability for reading, writing, and math, and low endorsement probability for behavior. This class appeared heterogeneous for language (.54) and social skills (.45) endorsement. The last, smallest class (<emph>LangComm</emph>; 4.95%) showed high endorsement probability for language and writing and low endorsement probability for math, social skills, and behavior. This class appeared heterogeneous for reading (.45).</p> <p>Graph: Figure 1. Conditional Item Probability Profile Plot for the 5-Class Model.</p> <hd id="AN0186128908-14">Auxiliary Variable Comparisons</hd> <p>A breakdown of demographic and developmental covariate variables based on modal class assignment (assignment based on most likely class) is provided in the supplemental files (Supplemental Table 5). (This is provided for descriptive purposes only, as covariate tests take into consideration posterior probabilities across classes and not modal assignment.) Two covariates demonstrated statistically significant omnibus effects: % Time Gen Ed, χ<sups>2</sups>(<reflink idref="bib8" id="ref62">8</reflink>) = 48.75, <emph>p </emph><.001, and Vineland-3 ABC, χ<sups>2</sups>(<reflink idref="bib4" id="ref63">4</reflink>) = 11.22, <emph>p </emph>=.02. Remaining effects were not statistically significant for age, χ<sups>2</sups>(<reflink idref="bib4" id="ref64">4</reflink>) = 2.04, <emph>p </emph>=.73; gender identity, χ<sups>2</sups>(<reflink idref="bib4" id="ref65">4</reflink>) = 3.73, <emph>p </emph>=.44; race/ethnicity, χ<sups>2</sups>(<reflink idref="bib4" id="ref66">4</reflink>) = 4.62, <emph>p </emph>=.33; FRPL, χ<sups>2</sups>(<reflink idref="bib4" id="ref67">4</reflink>) = 1.91, <emph>p </emph>=.75; MLE, χ<sups>2</sups>(<reflink idref="bib4" id="ref68">4</reflink>) = 4.03, <emph>p </emph>=.40; co-occurring intellectual disability, χ<sups>2</sups>(<reflink idref="bib4" id="ref69">4</reflink>) = 6.41, <emph>p </emph>=.17; co-occurring language or communication disorder, χ<sups>2</sups>(<reflink idref="bib4" id="ref70">4</reflink>) = 3.04, <emph>p </emph>=.55; and co-occurring ADHD, χ<sups>2</sups>(<reflink idref="bib4" id="ref71">4</reflink>) = 7.15, <emph>p </emph>=.13.</p> <p>Table 3 provides logits and <emph>OR</emph>s for covariates organized by reference class. The most consistent effect was for % Time Gen Ed that can be seen for all classes relative to <emph>All Goals.</emph> Each class was more likely (<emph>OR</emph>s = 6.59–10.95) relative to <emph>All Goals</emph> to be in general education for most of the time (≥ 80%) compared to the least amount of time (< 40%). Only <emph>LangLitAut</emph> was more likely (<emph>OR</emph> = 2.83) relative to <emph>All Goals</emph> to also be in general education for some of the time (40%–79%) compared to the least amount of time (< 40%). The effects for Vineland-3 ABC appeared more isolated. Increases in ABC above the grand-mean centered score were associated with higher odds of being in <emph>LangComm</emph> relative to <emph>All Goals</emph> (<emph>OR</emph> = 1.05), <emph>LangLitAut</emph> (<emph>OR</emph> = 1.07), and <emph>Academic</emph> (<emph>OR</emph> = 1.05). Vineland-3 ABC did not differentiate <emph>LangComm</emph> from <emph>Aut</emph> (<emph>OR</emph> = 1.03, <emph>p </emph>=.32). Further, increases in ABC were associated with higher odds of being in <emph>Aut</emph> relative to <emph>LangLitAut</emph> (<emph>OR</emph> = 1.05). No other between-class comparisons were statistically significant (<emph>ps </emph>>.05).</p> <p>Table 3. Logits and Odds Ratios for Significant Predictors of Class Membership by Reference Class.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="left" /></colgroup><thead><tr><th align="left" /><th align="left" /><th align="left" colspan="4">All Goals (40.49%)</th><th align="left" colspan="4">Autistic Characteristics (21.63%)</th><th align="left" colspan="4">Language, Literacy, and Autistic Characteristics (18.99%)</th><th align="left" colspan="4">Academic (13.94%)</th><th align="left" colspan="4">Language and Communication (4.95%)</th></tr><tr><th align="left">Class</th><th align="left">Effect</th><th align="left">Logit</th><th align="left"><italic>p</italic></th><th align="left">OR</th><th align="left">[95% CI]</th><th align="left">Logit</th><th align="left"><italic>p</italic></th><th align="left">OR</th><th align="left">[95% CI]</th><th align="left">Logit</th><th align="left"><italic>p</italic></th><th align="left">OR</th><th align="left">[95% CI]</th><th align="left">Logit</th><th align="left"><italic>p</italic></th><th align="left">OR</th><th align="left">[95% CI]</th><th align="left">Logit</th><th align="left"><italic>p</italic></th><th align="left">OR</th><th align="left">[95% CI]</th></tr></thead><tbody><tr><td><italic>All Goals (40.49%)</italic></td><td>40%–79% Gen Eda</td><td /><td /><td /><td /><td>0.28</td><td>.67</td><td>1.32</td><td>[0.37, 4.71]</td><td><bold>-1</bold>.<bold>04</bold></td><td>.<bold>04</bold></td><td><bold>0</bold>.<bold>35</bold></td><td><bold>[0.13, 0.94]</bold></td><td>-0.96</td><td>.20</td><td>0.39</td><td>[0.09, 1.67]</td><td>-0.38</td><td>.68</td><td>0.68</td><td>[0.11, 4.10]</td></tr><tr><td /><td>≥ 80% Gen Eda</td><td /><td /><td /><td /><td><bold>-2</bold>.<bold>14</bold></td><td><bold><.001</bold></td><td><bold>0</bold>.<bold>12</bold></td><td><bold>[0.05, 0.25]</bold></td><td><bold>-2</bold>.<bold>29</bold></td><td><bold><.001</bold></td><td><bold>0</bold>.<bold>10</bold></td><td><bold>[0.04, 0.26]</bold></td><td><bold>-2</bold>.<bold>39</bold></td><td><bold><.001</bold></td><td><bold>0</bold>.<bold>09</bold></td><td><bold>[0.02, 0.34]</bold></td><td><bold>-1</bold>.<bold>89</bold></td><td>.<bold>004</bold></td><td><bold>0</bold>.<bold>15</bold></td><td><bold>[0.04, 0.55]</bold></td></tr><tr><td /><td>ABC</td><td /><td /><td /><td /><td>-0.03</td><td>.09</td><td>0.97</td><td>[0.94, 1.00]</td><td>0.02</td><td>.24</td><td>1.02</td><td>[0.99, 1.05]</td><td>0.00</td><td>.92</td><td>1.00</td><td>[0.97, 1.03]</td><td><bold>-0</bold>.<bold>05</bold></td><td>.<bold>01</bold></td><td><bold>0</bold>.<bold>95</bold></td><td><bold>[0.91, 0.99]</bold></td></tr><tr><td><italic>Autistic Characteristics (21.63%)</italic></td><td>40%–79% Gen Eda</td><td>-0.28</td><td>.67</td><td>0.76</td><td>[0.21, 2.71]</td><td /><td /><td /><td /><td>-1.32</td><td>.12</td><td>0.27</td><td>[0.05, 1.41]</td><td>-1.23</td><td>.19</td><td>0.29</td><td>[0.05, 1.80]</td><td>-0.66</td><td>.56</td><td>0.52</td><td>[0.06, 4.73]</td></tr><tr><td /><td>≥ 80% Gen Eda</td><td><bold>2</bold>.<bold>14</bold></td><td><bold><.001</bold></td><td><bold>8</bold>.<bold>52</bold></td><td><bold>[3.94, 18.42]</bold></td><td /><td /><td /><td /><td>-0.15</td><td>.78</td><td>0.86</td><td>[0.31, 2.40]</td><td>-0.25</td><td>.70</td><td>0.78</td><td>[0.22, 2.75]</td><td>0.26</td><td>.72</td><td>1.29</td><td>[0.33, 5.14]</td></tr><tr><td /><td>ABC</td><td>0.03</td><td>.09</td><td>1.03</td><td>[1.00, 1.06]</td><td /><td /><td /><td /><td><bold>0</bold>.<bold>05</bold></td><td>.<bold>04</bold></td><td><bold>1</bold>.<bold>05</bold></td><td><bold>[1.00, 1.10]</bold></td><td>0.03</td><td>.19</td><td>1.03</td><td>[0.99, 1.07]</td><td>-0.03</td><td>.32</td><td>0.97</td><td>[0.93, 1.03]</td></tr><tr><td rowspan="2"><italic>Language, Literacy, and Autistic Characteristics (18.99%)</italic></td><td>40%–79% Gen Eda</td><td><bold>1</bold>.<bold>04</bold></td><td>.<bold>04</bold></td><td><bold>2</bold>.<bold>83</bold></td><td><bold>[1.06, 7.55]</bold></td><td>1.32</td><td>.12</td><td>3.74</td><td>[0.71, 19.67]</td><td /><td /><td /><td /><td>0.09</td><td>.91</td><td>1.09</td><td> [0.23, 5.19]</td><td>0.66</td><td>.51</td><td>1.94</td><td>[0.27, 13.78]</td></tr><tr><td>≥ 80% Gen Eda</td><td><bold>2</bold>.<bold>29</bold></td><td><bold><.001</bold></td><td><bold>9</bold>.<bold>86</bold></td><td><bold>[3.82, 25.42]</bold></td><td>0.15</td><td>.78</td><td>1.16</td><td>[0.42, 3.21]</td><td /><td /><td /><td /><td>-0.11</td><td>.87</td><td>0.90</td><td>[0.27, 3.05]</td><td>0.40</td><td>.57</td><td>1.50</td><td>[0.38, 5.91]</td></tr><tr><td /><td>ABC</td><td>-0.02</td><td>.24</td><td>0.98</td><td>[0.95, 1.01]</td><td><bold>-0</bold>.<bold>05</bold></td><td>.<bold>04</bold></td><td><bold>0</bold>.<bold>96</bold></td><td><bold>[0.91, 1.00]</bold></td><td /><td /><td /><td /><td>-0.02</td><td>.31</td><td>0.98</td><td>[0.94, 1.02]</td><td><bold>-0</bold>.<bold>07</bold></td><td>.<bold>003</bold></td><td><bold>0</bold>.<bold>93</bold></td><td><bold>[0.89, 0.98]</bold></td></tr><tr><td><italic>Academic (13.94%)</italic></td><td>40%–79% Gen Eda</td><td>0.96</td><td>.20</td><td>2.60</td><td>[0.60, 11.27]</td><td>1.23</td><td>.19</td><td>2.32</td><td>[0.55, 21.17]</td><td>-0.09</td><td>.91</td><td>0.92</td><td>[0.19, 4.37]</td><td /><td /><td /><td /><td>0.57</td><td>.60</td><td>1.78</td><td>[0.21, 15.15]</td></tr><tr><td /><td>≥ 80% Gen Eda</td><td><bold>2</bold>.<bold>39</bold></td><td><bold><.001</bold></td><td><bold>10</bold>.<bold>95</bold></td><td><bold>[2.93, 40.94]</bold></td><td>0.25</td><td>.70</td><td>1.29</td><td>[0.36, 4.55]</td><td>0.11</td><td>.87</td><td>1.11</td><td>[0.33, 3.76]</td><td /><td /><td /><td /><td>0.51</td><td>.53</td><td>1.66</td><td>[0.34. 8.08]</td></tr><tr><td /><td>ABC</td><td>0.00</td><td>.92</td><td>1.00</td><td>[0.97, 1.04]</td><td>-0.03</td><td>.19</td><td>0.98</td><td>[0.94, 1.01]</td><td>0.02</td><td>.31</td><td>1.02</td><td>[0.98, 1.06]</td><td /><td /><td /><td /><td><bold>-0</bold>.<bold>05</bold></td><td>.<bold>04</bold></td><td><bold>0</bold>.<bold>95</bold></td><td><bold>[0.91, 1.00]</bold></td></tr><tr><td><italic>Language and Communication (4.95%)</italic></td><td>40%–79% Gen Eda</td><td>0.38</td><td>.68</td><td>1.46</td><td>[0.24, 8.77]</td><td>0.66</td><td>.56</td><td>1.93</td><td>[0.21, 17.62]</td><td>-0.66</td><td>.51</td><td>0.52</td><td>[0.07, 3.68]</td><td>-0.57</td><td>.60</td><td>0.56</td><td>[0.07, 4.81]</td><td /><td /><td /><td /></tr><tr><td /><td>≥ 80% Gen Eda</td><td><bold>1</bold>.<bold>89</bold></td><td>.<bold>004</bold></td><td><bold>6</bold>.<bold>59</bold></td><td><bold>[1.81, 24.06]</bold></td><td>-0.26</td><td>.72</td><td>0.77</td><td>[0.20, 3.08]</td><td>-0.40</td><td>.57</td><td>0.67</td><td>[0.17, 2.64]</td><td>-0.51</td><td>.53</td><td>0.60</td><td>[0.12, 2.93]</td><td /><td /><td /><td /></tr><tr><td /><td>ABC</td><td><bold>0</bold>.<bold>05</bold></td><td>.<bold>01</bold></td><td><bold>1</bold>.<bold>05</bold></td><td><bold>[1.01, 1.10]</bold></td><td>0.03</td><td>.32</td><td>1.03</td><td>[0.98, 1.08]</td><td><bold>0</bold>.<bold>07</bold></td><td>.<bold>003</bold></td><td><bold>1</bold>.<bold>07</bold></td><td><bold>[1.02, 1.13]</bold></td><td><bold>0</bold>.<bold>05</bold></td><td>.<bold>04</bold></td><td><bold>1</bold>.<bold>05</bold></td><td><bold>[1.00, 1.11]</bold></td><td /><td /><td /><td /></tr></tbody></table> </ephtml> </p> <ulist> <item>4 <emph>Note.</emph> Statistically significant effects (<emph>p </emph><.05) are bolded for readability. Gen Ed = general education. ABC = Vineland 3 Adaptive Behavior Composite (grand mean centered). Predictor classes are shown by row with reference classes shown by column.</item> <item>5 Percent of time spent in Gen Ed is in reference to < 40% Time Gen Ed group.</item> </ulist> <hd id="AN0186128908-15">Discussion</hd> <p>The current study characterizes autistic youth's educational goals by using latent class analysis to identify subgroups based on multivariate response patterns and examine how selected demographic and developmental covariates may differentiate subgroup membership. Mixture modeling provides a person-centered approach to characterizing our sample, leveraging the expected heterogeneity present among autistic children to consider the presence of underlying subgroups across skill domain areas. Our findings support the presence of multiple subgroups based on likelihoods of having different educational goals in academics, language and communication, social skills, and behavior. We also find that select demographic (% Time Gen Ed) and developmental factors (adaptive behavior) predicted differences in subgroup membership.</p> <p>Our approach expands upon existing studies that have relied on broad characterizations of autistic children's educational goals generally focused on the presence of an IEP ([<reflink idref="bib17" id="ref72">17</reflink>]; [<reflink idref="bib21" id="ref73">21</reflink>]; [<reflink idref="bib24" id="ref74">24</reflink>]). Our findings highlight that a broad characterization of autistic subgroups is possible based on the identification of at least one IEP goal across six domain areas. Importantly, our findings reinforce that (a) the educational needs of autistic children are not defined solely by their diagnosis and (b) autistic children demonstrate a wide array of diverse educational needs.</p> <hd id="AN0186128908-16">Identified Latent Classes</hd> <p>Our identified latent classes of educational needs showed five distinct subgroups of autistic children. Our largest class—<emph>All Goals</emph>, at 40.49%—included children with a high likelihood across all six goal domains. It is important to note that high likelihood endorses goal breadth but not necessarily goal frequency, given that indicators represented having at least one goal in each area. Still, this substantial subgroup highlights that many autistic children have educational goals across all six domains.</p> <p>Two classes demonstrate elevated likelihoods for goals related to autistic characteristics but differ on the presence of language and literacy goals. Our second largest class—<emph>Aut</emph>, at 21.63%—includes autistic children who had high likelihoods in the domain area most associated with autism (social skills) and elevated likelihood in having behavior goals. With low likelihood endorsement across academic outcomes, this subgroup has educational goals that most educators might expect for autistic students (that is, goals only in social skills and behavior). However, at only one-fifth of our sample, this group is notable but by far not the majority. Our third largest class—<emph>LangLitAut</emph>, at 18.99%—shares similarities with our prior class, with the addition of elevated likelihoods for both reading and writing educational goals. As seen in Figure 1, these classes appeared similar to our first class except for the very distinct elevation in reading, writing, and language and communication goals. This is an important distinction that highlights that approximately 40% of our classes endorse belonging to a high autistic characteristics subgroup but were split on the inclusion of language and literacy, echoing extant findings of specific challenges noted in these areas for many autistic children ([<reflink idref="bib36" id="ref75">36</reflink>]; [<reflink idref="bib43" id="ref76">43</reflink>]). However, we are unsure if this difference is due to child ability or potentially stakeholder decision making. The distinction in these two groups raises an important question about the priorities and discussions about identifying meaningful IEP goals for autistic children, as our class characteristics appear similar yet distinct regarding goal representation. These findings align with broader conversations around IEP quality for autistic children and the need to understand more about educational decisions (e.g., [<reflink idref="bib7" id="ref77">7</reflink>]; [<reflink idref="bib31" id="ref78">31</reflink>]). Research examining the quality and characteristics of IEPs for children with learning disabilities shows numerous quality concerns, especially the use of empirically based decisions to identify appropriate services and supports ([<reflink idref="bib23" id="ref79">23</reflink>]), and further research is needed with autistic students.</p> <p>Our two most infrequent classes appear to be isolated in educational goals not associated with primary autistic characteristics. Our fourth class—<emph>Academic</emph>, at 13.94%—shows an elevated likelihood across only reading, writing, and math. Notably, this class appears heterogeneous in terms of both language and social skills, with only behavior showing low endorsement likelihood. Our fifth class—<emph>LangComm</emph>, at 4.95%—includes a small subgroup of children who demonstrate language and communication goals with an elevated likelihood for written language goals. Taken together, these classes represent about one fifth of the sample of children identified as autistic who did not have educational goals associated with autistic characteristics. Our findings highlight that some subgroups of autistic students have educational goals focused on broad academic or explicit spoken and written language skills rather than goals aligned to autistic characteristics.</p> <hd id="AN0186128908-17">Demographic and Developmental Covariates</hd> <p>Two covariates—% Time Gen Ed and adaptive behavior—predicted differences in latent class membership. % Time Gen Ed showed a consistent effect where those in <emph>All Goals</emph> relative to all other latent classes were most likely to spend the least amount of time in general education settings. This was most predominant for the comparison between the < 40% and ≥ 80% levels, as this comparison only held for comparing the < 40% and 40%–79% levels for <emph>LangLitAut</emph> compared to <emph>All Goals</emph>. This suggests that the children with the largest goal area breadth may correspond to the autistic students with the most significant needs who have the least access to the general education curriculum. Based on modal membership (Supplemental Table 5), approximately 58.4% of <emph>All Goals</emph> spend < 40% of their time in a general education context, which is far lower than the 19.5%–30.8% observed for the remaining classes. This does not, however, provide information about specific goal types, as there may be further substantive differences in the content of the goals themselves for children in special education versus general education settings. For example, these students may have a higher proportion of functional daily living goals as opposed to academic-oriented goals ([<reflink idref="bib39" id="ref80">39</reflink>]).</p> <p>All classes demonstrated relatively low adaptive behavior based on modal class estimates (<emph>Ms </emph>= 67.06–80.68; Supplemental Table 5). Based on these estimates, <emph>Aut</emph> and <emph>LangComm</emph> demonstrated the highest adaptive behavior estimates (<emph>Ms </emph>= 77.39 and 80.68), followed by <emph>Academic</emph> and <emph>LangLitAut</emph> (<emph>Ms </emph>= 71.77 and 72.08), and then followed by <emph>All Goals</emph> (<emph>M </emph>= 67.06). When statistically compared using posterior probabilities for class membership, however, the effect of adaptive behavior differentiated <emph>LangComm</emph> from <emph>All Goals</emph>, <emph>LangLitAut</emph>, and <emph>Academic</emph> as well as <emph>Aut</emph> from <emph>LangLitAut</emph>. The effect for <emph>LangComm</emph> on most other classes should be interpreted with caution given the low class size. The differentiation between <emph>Aut</emph> and <emph>LangLitAut</emph> suggests that a child's adaptive behavior may be related to the presence of language and literacy goals in addition to more common social and behavioral goals. As the presence of language and communication goals appeared largely heterogeneous in <emph>Aut</emph>, further clarification is needed to understand this difference and if it may be related to specific goals rather than goal domains. In general, our findings did not substantially support adaptive behavior as differentiating membership across other latent classes. It is also important to note that autistic characteristics (SRS-2) and adaptive behavior (Vineland-3 ABC) showed a strong negative association, leading us to remove autistic characteristics from our covariate testing. When both covariates are in the model, neither show a statistically significant finding (but the significant effects for the other covariates remained). This suggests that these measures may be generally associated with overall support needs in our sample.</p> <p>Importantly, we found that other demographic and developmental covariates appeared largely unrelated to latent class membership. On one hand, our findings highlight that class membership is unrelated to select demographic variables like gender, race/ethnicity, or socioeconomic status. These findings build from those offered by [<reflink idref="bib38" id="ref81">38</reflink>] in that we did not identify demographic disparities as predicting differences in the latent educational goal patterns of autistic children. However, given our findings about specific classes related to different goal domain areas, it may have been expected to see greater distinctions between latent classes on select variables aligned with co-occurring conditions like intellectual disability and language or communication disorder. We believe there are three potential reasons for this. First, we examined breadth rather than depth of educational goals. We are not comparing the presence of distinct or more intensive educational goals in these domains, so the simple nuance of having or not having goals may not be sensitive enough to detect this difference. Second, given that we relied on parent report of educational goals, parents may have conflated educational needs and goals, endorsing general educational needs of their child without a formal goal in place. Third, we might consider the lack of class differences—in factors where we expect there to be differences—to be related to the choices made around educational decision making for autistic children. Recent research has found that goal expectations for students with complex support needs like autism can be problematic across academic and social-behavioral-communication learning goals ([<reflink idref="bib35" id="ref82">35</reflink>]). In instances where developmental covariates include ranges of ability where educational goals might be expected, the choice for developing those educational goals falls to the multidisciplinary IEP team. We are not privy to those conversations, given our research design, and further information about the choices being made about the types of educational goals selected for autistic children (both in terms of type and in terms of data identified in support of such decisions) is needed.</p> <hd id="AN0186128908-18">Limitations and Future Directions</hd> <p>Our approach to characterizing the educational needs of autistic children across multiple skill domains is not without limitations. We collected data on whether children had at least one IEP goal across the six measured skill domains. This provides a perspective on breadth across all six areas but does not capture depth across these areas. With our current latent classes, we cannot identify children who have different educational goals within a given skill domain. Further research is needed to characterize the depth of educational goals possible across these skill domains. Additionally, our reliance on caregiver report means that we did not review actual IEP documents. While parents can be a knowledgeable source about their child's educational needs, it is very possible that parents mischaracterized their children's educational goals and either under- or overreported educational needs. Furthermore, given that our data were collected as part of a larger study focused on literacy and educational access during the COVID-19 pandemic, it is possible that we recruited a biased sample of families with an autistic child who specifically had a language or literacy need. While our overall proportion of children identified with an IEP is similar to other studies of autistic children, we may overrepresent specific subgroups that may not be as apparent in samples collected without a literacy focus. We also acknowledge that while we have a highly heterogeneous sample in terms of abilities, we have important racial identity representation limitations. Our sample demographics largely match those reported for the full SPARK dataset (73% White, 7% Black, 3% Asian, 11% multiracial, and 19% Hispanic; [<reflink idref="bib34" id="ref83">34</reflink>]). While our sample allowed us to estimate important aspects of heterogeneity present in autistic youth, we need additional research with more racially and ethnically diverse samples to better understand how these factors may impact latent class identification and membership. Lastly, we were unable to collect direct assessment data related to children's abilities in the IEP skills of interest; therefore, we cannot comment on the diversity of skills present beyond our participants as being identified as having educational goals in a wide array of areas.</p> <p>This study characterized the educational needs of autistic children by leveraging a large sample survey approach with a mixture modeling analytical design. Our approach endorses that multiple subgroups of educational needs exist among autistic children. Further research to understand how decisions are made about these educational needs, and better aligning intervention services to these identified needs among heterogeneous groups of autistic children, will help educators better understand the diverse needs of autistic children beyond diagnostic areas. Continued research into better characterizing and understanding the needs of autistic children will greatly benefit the educational planning for all autistic children and the ability for educators to design effective approaches to meet these diverse needs.</p> <hd id="AN0186128908-19">Supplemental Material</hd> <p>Graph: Supplemental material, sj-docx-1-ecx-10.1177_00144029251331861 for Characterizing the Special Education Goals of Autistic Students: Latent Class Analysis With Demographic and Developmental Covariates by Matthew C. Zajic, Juliette Gudknecht and Nancy S. McIntyre in Exceptional Children</p> <ref id="AN0186128908-20"> <title> References </title> <blist> <bibl id="bib1" idref="ref1" type="bt">1</bibl> <bibtext> American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders: DSM-5-TR. American Psychiatric Association Publishing.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref5" type="bt">2</bibl> <bibtext> Barnard-Brak L., Reschly A. (2019). Educational versus clinical diagnoses of autism spectrum disorder: Updated and expanded findings. School Psychology Review, 48(2), 185–189. https://doi.org/10.17105/spr-2018-0009.v48-2</bibtext> </blist> <blist> <bibl id="bib3" idref="ref15" type="bt">3</bibl> <bibtext> Bullen J., Zajic M. C., McIntyre N., Solari E., Mundy P. (2022). Patterns of math and reading achievement in children and adolescents with autism spectrum disorder. Research in Autism Spectrum Disorders, 92, Article 101933. https://doi.org/10.1016/j.rasd.2022.101933</bibtext> </blist> <blist> <bibl id="bib4" idref="ref49" type="bt">4</bibl> <bibtext> Constantino J. N., Gruber C. P. (2012). Social responsiveness scale (2<sups>nd</sups> ed.). Western Psychological Services.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref35" type="bt">5</bibl> <bibtext> Duncan A., Risley S., Combs A., Lacey H., Hamik E., Fershtman C., Kneeskern E., Patel M., Crosby L., Hood A., Zoromski A., Tamm L. (2023). School challenges and services related to executive functioning for fully included middle schoolers with autism. Focus on Autism and Other Developmental Disabilities, 38(2), 90–100. https://doi.org/10.1177/10883576221110167</bibtext> </blist> <blist> <bibl id="bib6" idref="ref31" type="bt">6</bibl> <bibtext> Gillberg C. (2019). Autism and comorbidities: Autism plus – implications for diagnosis, prognosis, and interventions. In Jordan R., Roberts J. M., Hume K. (Eds.), The SAGE handbook of autism and education (pp. 492–501). Sage. https://doi.org/10.4135/9781526470409.n39</bibtext> </blist> <blist> <bibl id="bib7" idref="ref77" type="bt">7</bibl> <bibtext> Goldstein S., Sellars T., Velez A. (2023). From eligibility assessment to intervention for students with autism spectrum disorder. Psychology in the Schools, 60(2), 364–377. https://doi.org/10.1002/pits.22795</bibtext> </blist> <blist> <bibl id="bib8" idref="ref51" type="bt">8</bibl> <bibtext> Hallquist M. N., Wiley J. F. (2018). MplusAutomation: An R package for facilitating large-scale latent variable analyes in Mplus. Structural Equation Modeling: A Multidisciplinary Journal, 25(4), 621–638. https://doi.org/10.1080/10705511.2017.1402334</bibtext> </blist> <blist> <bibl id="bib9" idref="ref41" type="bt">9</bibl> <bibtext> Howard M. C., Hoffman M. E. (2018). Variable-centered, person-centered, and person-specific approaches: Where theory meets the method. Organizational Research Methods, 21(4), 846–876. https://doi.org/10.1177/1094428117744021</bibtext> </blist> <blist> <bibtext> Ibrahim I. (2020). Specific learning disorder in children with autism spectrum disorder: Current issues and future implications. Advances in Neurodevelopmental Disorders, 4, 103–112. https://doi.org/10.1007/s41252-019-00141-x</bibtext> </blist> <blist> <bibtext> Individuals with Disabilities Education Improvement Act of 2004, 20 U.S.C. § 1400 et seq. (2004). https://sites.ed.gov/idea/statute-chapter-33/subchapter-i/1400</bibtext> </blist> <blist> <bibtext> Irwin V., Wang K., Tezil T., Zhang J., Filbey A., Jung J., Bullock Mann F., Dilig R., Parker S. (2023). Report on the Condition of Education 2023 (NCES 2023-144rev). U.S. Department of Education. Washington, DC: National Center for Education Statistics. Retrieved September 10, 2023 from https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2023144rev</bibtext> </blist> <blist> <bibtext> Jordan R. (2019). Practical effects of co-morbidities and the concept of neurodiversity. In Jordan R., Roberts J. M., Hume K. (Eds.), The SAGE handbook of autism and education (pp. 502–510). Sage. https://doi.org/10.4135/9781526470409.n40</bibtext> </blist> <blist> <bibtext> Jordan R., Roberts J. M., Hume K. (2019). The SAGE handbook of autism and education. Sage. https://doi.org/10.4135/9781526470409</bibtext> </blist> <blist> <bibtext> Keen D., Webster A., Ridley G. (2016). How well are children with autism spectrum disorder doing academically at school? An overview of the literature. Autism, 20(3), 276–294. https://doi.org/10.1177/1362361315580962</bibtext> </blist> <blist> <bibtext> Kurth J. A., Mastergeorge A. M., Paschall K. (2016). Economic and demographic factors impacting placement of students with autism. Education and Training in Autism and Developmental Disabilities, 51(1), 3–12. https://<ulink href="http://www.jstor.org/stable/26420360">www.jstor.org/stable/26420360</ulink></bibtext> </blist> <blist> <bibtext> Lavelle T. A., Weinstein M. C., Newhouse J. P., Munir K., Kuhlthau K. A., Prosser L. A. (2014). Economic burden of childhood autism spectrum disorders. Pediatrics, 133(3), e520–e529. https://doi.org/10.1542/peds.2013-0763</bibtext> </blist> <blist> <bibtext> Lord C., Charman T., Havdahl A., Carbone P., Anagnostou E., Boyd B., Carr T., de Vries P. J., Dissanayake C., Divan G., Freitag C. M., Gotelli M. M., Kasari C., Knapp M., Mundy P., Plank A., Scahill L., Servili C., Shattuck P.. McCauley J. B. (2022). The lancet commission on the future of care and clinical research in autism. The Lancet, 399(10321), 271–334. https://doi.org/10.1016/S0140-6736(21)01541-5</bibtext> </blist> <blist> <bibtext> Maenner M. J., Warren Z., Williams A. R., Amoakohene E., Bakian A. V., Bilder D. A., Durkin M. S., Fitzgerald R. T., Furnier S. M., Hughes M. M., Ladd-Acosta C. M., McArthur D., Pas E. T., Salinas A., Vehorn A., Williams S., Esler A., Grzybowski A., Hall-Lande J., Nguyen R. H. N., Shaw K. A. (2023). Prevalence and characteristics of autism Spectrum disorder among children aged 8 years - autism and developmental disabilities monitoring network, 11 sites, United States, 2020. Morbidity and Mortality Weekly Report. Surveillance Summaries (Washington, D.C.: 2002), 72(2), 1–14. https://doi.org/10.15585/mmwr.ss7202a1</bibtext> </blist> <blist> <bibtext> Masyn K. E. (2013). Latent class analysis and finite mixture modeling. In Little T. D. (Ed.), The Oxford handbook of quantitative methods in psychology: Vol. 2: Statistical analysis (pp. 551–611). Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199934898.013.0025</bibtext> </blist> <blist> <bibtext> McDonald C. A., Donnelly J. P., Feldman-Alguire A. L., Rodgers J. D., Lopata C., Thomeer M. L. (2019). Special education service use by children with autism spectrum disorder. Journal of Autism and Developmental Disorders, 49, 2437–2446. https://doi.org/10.1007/s10803-019-03997-z</bibtext> </blist> <blist> <bibtext> McIntyre N. S., Solari E. J., Grimm R. P., E Lerro L., E Gonzales J., Mundy P. C. (2017). A comprehensive examination of Reading heterogeneity in students with high functioning autism: Distinct Reading profiles and their relation to autism symptom severity. Journal of Autism and Developmental Disorders, 47, 1086–1101. https://doi.org/10.1007/s10803-017-3029-0</bibtext> </blist> <blist> <bibtext> McKenna J., Solis M., Garwood J., Parenti M. (2024). Characteristics of individualized education programs for students with learning disabilities: A systematic review. Learning Disability Quarterly, 47(3), 194–207. https://doi.org/10.1177/07319487231182697</bibtext> </blist> <blist> <bibtext> Menezes M., Pappagianopoulos J., Robinson M. F., Sadikova E., Cross R., Mazurek M. O. (2023). Disparities in special education services for youth with autism spectrum disorder in the United States. Journal of Special Education, 56(4), 219–224. https://doi.org/10.1177/00224669221096428</bibtext> </blist> <blist> <bibtext> Muthén L. K., Muthén B. O. (1998–2017). Mplus User's Guide (8th ed.). Muthén & Muthén.</bibtext> </blist> <blist> <bibtext> Nagin D. (2005). Group-based modeling of development. Harvard University Press. https://doi.org/10.4159/9780674041318</bibtext> </blist> <blist> <bibtext> National Center for Education Statistics. (2023). Table 204.60. Percentage distribution of school-age students served under Individuals with Disabilities Education Act (IDEA), Part B, by educational environment and type of disability: Selected years, fall 1989 through fall 2021. Retrieved on September 19, 2023 from https://nces.ed.gov/programs/digest/d22/tables/dt22_204.60.asp?current=yes</bibtext> </blist> <blist> <bibtext> Nylund-Gibson K., Choi A. Y. (2018). Ten frequently asked questions about latent class analysis. Translational Issues in Psychological Science, 4(4), 440–461. https://doi.org/10.1037/tps0000176</bibtext> </blist> <blist> <bibtext> Nylund-Gibson K., Grimm R. P., Masyn K. E. (2019). Prediction from latent classes: A demonstration of different approaches to include distal outcomes in mixture models. Structural Equation Modeling: A Multidisciplinary Journal, 26(6), 967–985. https://doi.org/10.1080/10705511.2019.1590146</bibtext> </blist> <blist> <bibtext> R Core Team. (2023). R: A language and environment for statistical computing. https://<ulink href="http://www.R-project.org/">www.R-project.org/</ulink></bibtext> </blist> <blist> <bibtext> Ruble L., McGrew J., Dalrymple N., Jung L. (2010). Examining the quality of IEPs for young children with autism. Journal of Autism and Developmental Disorders, 40(12), 1459–1470. https://doi.org/10.1007/s10803-010-1003-1</bibtext> </blist> <blist> <bibtext> Schwartz I. S., Bravo A., Finlayson R., Flaherty J., Luna A. (2022). Evidence-based practice in schools. In Leaf J. B., Cihon J. H., Ferguson J. L., Weiss M. J. (Eds.), Handbook of Applied Behavior Analysis Interventions for Autism. Springer. https://doi.org/10.1007/978-3-030-96478-8_6</bibtext> </blist> <blist> <bibtext> SFARI. (n.d. -a). SPARK. Retrieved on September 19, 2023 from https://<ulink href="http://www.sfari.org/resource/spark/">www.sfari.org/resource/spark/</ulink></bibtext> </blist> <blist> <bibtext> SFARI. (n.d. -b). SPARK Demographic and Clinical Information. Retrieved on September 27, 2023 from https://<ulink href="http://www.sfari.org/spark-demographic-and-clinical-information">www.sfari.org/spark-demographic-and-clinical-information</ulink></bibtext> </blist> <blist> <bibtext> Shogren K. A., Zimmerman K. N., Burke K. M., Zagona A. L., Kurth J. A., Malone E. J., Loyless R., Gerasimova D., Pace J. R., Hicks T. A. (2023). Educational placement and goal quality for students with complex support needs. Inclusion, 11(4), 222–244. https://doi.org/10.1352/2326-6988-11.4.222</bibtext> </blist> <blist> <bibtext> Solari E. J., McIntyre N. S., Dynia J. M., Henry A. (2021). Reading development for individuals with ASD: Current knowledge and a roadmap for next steps. In Cook B. G., Tankersley M., Landrum T. J. (Eds.), The next big thing in learning and behavioral disabilities (pp. 149–172). Emerald Publishing Limited. https://doi.org/10.1108/S0735-004X20210000031010</bibtext> </blist> <blist> <bibtext> Sparrow S. S., Cicchetti D. V., Saulnier C. A. (2016). Vineland adaptive behavior scales (3<sups>rd</sups> ed.). Pearson Assessments.</bibtext> </blist> <blist> <bibtext> Suhrheinrich J., Root B. V., Melgarejo M., Dickson K., Vejnoska S., Reith S. R. (2021). Exploring differences and disparities in school-based services received by students with autism: A systematic review. Research in Autism Spectrum Disorders, 85, Article 101791. https://doi.org/10.1016/j.rasd.2021.101791</bibtext> </blist> <blist> <bibtext> Vanroy K., Vermeulen P. (2019). Setting and evaluating goals in education. In Jordan R., Roberts J. M., Hume K. (Eds.), The SAGE handbook of autism and education (pp. 541–551). Sage.</bibtext> </blist> <blist> <bibtext> Wilczynski S., Menousek K., Hunter M., Hunter M., Mudgal D. (2007). Individualized education programs for youth with autism spectrum disorders. Psychology in the Schools, 44(7), 653–666. https://doi.org/10.1002/pits.20255</bibtext> </blist> <blist> <bibtext> Yell M. L., Collins J., Kumpiene G., Bateman D. (2020). The individualized education program: Procedural and substantive requirements. Teaching Exceptional Children, 52(5), 304–318. https://doi.org/10.1177/0040059920906592</bibtext> </blist> <blist> <bibtext> Zajic M. C., McIntyre N., Swain-Lerro L., Novotny S., Oswald T., Mundy P. (2018). Attention and written expression in school-age, high-functioning children with autism spectrum disorders. Autism, 22(3), 245–258. https://doi.org/10.1177/1362361316675121</bibtext> </blist> <blist> <bibtext> Zajic M. C., Wilson S. E. (2020). Writing research involving children with autism spectrum disorder without a co-occurring intellectual disability: A systematic review using a language domains and mediational systems framework. Research in Autism Spectrum Disorders, 70, Article 101471. https://doi.org/10.1016/j.rasd.2019.101471</bibtext> </blist> </ref> <ref id="AN0186128908-21"> <title> Footnotes </title> <blist> <bibtext> The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.</bibtext> </blist> <blist> <bibtext> The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported in part by a faculty grant awarded to MZ by the Provost's Office at Teachers College, Columbia University. MZ received support as a training fellow of an Institute of Education Sciences Training Grant (R305B220021) to The Regents of the University of California, Santa Barbara. We are grateful to all the families in SPARK, the SPARK clinical sites, and the SPARK staff for helping make this research possible.</bibtext> </blist> <blist> <bibtext> Matthew C. Zajic https://orcid.org/0000-0002-9806-4832 Juliette Gudknecht https://orcid.org/0000-0002-0352-1907 Nancy S. McIntyre https://orcid.org/0000-0001-5085-2672</bibtext> </blist> <blist> <bibtext> Supplemental material for this article is available online. Manuscript received January 2024; accepted March 2025.</bibtext> </blist> </ref> <aug> <p>By Matthew C. Zajic; Juliette Gudknecht and Nancy S. McIntyre</p> <p>Reported by Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib14" firstref="ref2"></nolink> <nolink nlid="nl2" bibid="bib12" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib19" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib32" firstref="ref6"></nolink> <nolink nlid="nl5" bibid="bib41" firstref="ref8"></nolink> <nolink nlid="nl6" bibid="bib11" firstref="ref9"></nolink> <nolink nlid="nl7" bibid="bib39" firstref="ref10"></nolink> <nolink nlid="nl8" bibid="bib40" firstref="ref11"></nolink> <nolink nlid="nl9" bibid="bib17" firstref="ref12"></nolink> <nolink nlid="nl10" bibid="bib21" firstref="ref13"></nolink> <nolink nlid="nl11" bibid="bib24" firstref="ref14"></nolink> <nolink nlid="nl12" bibid="bib22" firstref="ref16"></nolink> <nolink nlid="nl13" bibid="bib43" firstref="ref17"></nolink> <nolink nlid="nl14" bibid="bib10" firstref="ref18"></nolink> <nolink nlid="nl15" bibid="bib13" firstref="ref19"></nolink> <nolink nlid="nl16" bibid="bib18" firstref="ref20"></nolink> <nolink nlid="nl17" bibid="bib15" firstref="ref21"></nolink> <nolink nlid="nl18" bibid="bib36" firstref="ref22"></nolink> <nolink nlid="nl19" bibid="bib38" firstref="ref27"></nolink> <nolink nlid="nl20" bibid="bib27" firstref="ref29"></nolink> <nolink nlid="nl21" bibid="bib16" firstref="ref30"></nolink> <nolink nlid="nl22" bibid="bib42" firstref="ref34"></nolink> <nolink nlid="nl23" bibid="bib20" firstref="ref42"></nolink> <nolink nlid="nl24" bibid="bib28" firstref="ref43"></nolink> <nolink nlid="nl25" bibid="bib29" firstref="ref44"></nolink> <nolink nlid="nl26" bibid="bib33" firstref="ref45"></nolink> <nolink nlid="nl27" bibid="bib37" firstref="ref47"></nolink> <nolink nlid="nl28" bibid="bib30" firstref="ref52"></nolink> <nolink nlid="nl29" bibid="bib25" firstref="ref53"></nolink> <nolink nlid="nl30" bibid="bib26" firstref="ref58"></nolink> <nolink nlid="nl31" bibid="bib31" firstref="ref78"></nolink> <nolink nlid="nl32" bibid="bib23" firstref="ref79"></nolink> <nolink nlid="nl33" bibid="bib35" firstref="ref82"></nolink> <nolink nlid="nl34" bibid="bib34" firstref="ref83"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1475373
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Characterizing the Special Education Goals of Autistic Students: Latent Class Analysis with Demographic and Developmental Covariates
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Matthew+C%2E+Zajic%22">Matthew C. Zajic</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-9806-4832">0000-0002-9806-4832</externalLink>)<br /><searchLink fieldCode="AR" term="%22Juliette+Gudknecht%22">Juliette Gudknecht</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-0352-1907">0000-0002-0352-1907</externalLink>)<br /><searchLink fieldCode="AR" term="%22Nancy+S%2E+McIntyre%22">Nancy S. McIntyre</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-5085-2672">0000-0001-5085-2672</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Exceptional+Children%22"><i>Exceptional Children</i></searchLink>. 2025 91(4):359-376.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 18
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: SourceSuprt
  Label: Sponsoring Agency
  Group: SrcSuprt
  Data: Institute of Education Sciences (ED)
– Name: NumberContract
  Label: Contract Number
  Group: NumCntrct
  Data: R305B220021
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Elementary+Secondary+Education%22">Elementary Secondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Autism+Spectrum+Disorders%22">Autism Spectrum Disorders</searchLink><br /><searchLink fieldCode="DE" term="%22Students+with+Disabilities%22">Students with Disabilities</searchLink><br /><searchLink fieldCode="DE" term="%22Special+Education%22">Special Education</searchLink><br /><searchLink fieldCode="DE" term="%22Educational+Objectives%22">Educational Objectives</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Needs%22">Student Needs</searchLink><br /><searchLink fieldCode="DE" term="%22Elementary+Secondary+Education%22">Elementary Secondary Education</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Characteristics%22">Student Characteristics</searchLink><br /><searchLink fieldCode="DE" term="%22Individualized+Education+Programs%22">Individualized Education Programs</searchLink><br /><searchLink fieldCode="DE" term="%22Symptoms+%28Individual+Disorders%29%22">Symptoms (Individual Disorders)</searchLink><br /><searchLink fieldCode="DE" term="%22Language+Skills%22">Language Skills</searchLink><br /><searchLink fieldCode="DE" term="%22Literacy%22">Literacy</searchLink><br /><searchLink fieldCode="DE" term="%22Communication+Skills%22">Communication Skills</searchLink><br /><searchLink fieldCode="DE" term="%22Academic+Achievement%22">Academic Achievement</searchLink>
– Name: SubjectThesaurus
  Label: Assessment and Survey Identifiers
  Group: Su
  Data: <searchLink fieldCode="SU" term="%22Vineland+Adaptive+Behavior+Scales%22">Vineland Adaptive Behavior Scales</searchLink><br /><searchLink fieldCode="SU" term="%22Social+Responsiveness+Scale%22">Social Responsiveness Scale</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1177/00144029251331861
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0014-4029<br />2163-5560
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Many autistic children receive special education services via Individualized Education Programs (IEPs) that include specific educational goals and needs. Prior research has examined programming options available to support autistic students, but less is known about their educational needs across academic and developmental educational goals. Additionally, existing approaches have often relied on small studies focused on describing individual goal areas. This study uses data from 551 families from across the United States with an autistic child in grades K-12 and latent class analysis to (a) identify latent, or underlying, subgroups based off multivariate response patterns across educational goals endorsed in six domains (reading, writing, math, language and communication, social skills, behavior), and (b) examine if demographic and developmental covariates predict latent class membership. We identified five latent classes: "All Goals" (40.49%); "Autistic Characteristics" (21.63%); "Language, Literacy, and Autistic Characteristics" (18.99%); "Academic" (13.94%); and "Language and Communication" (4.95%). Two covariates--percentage of time spent in general education and adaptive behavior--predicted differences in latent class membership. Findings offer a comprehensive examination into the heterogeneous educational needs of autistic school-age children. Our results emphasize the need for researchers and educators to understand the educational needs of autistic students beyond the presence of an IEP.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: CodeSource
  Label: IES Funded
  Group: SrcInfo
  Data: Yes
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1475373
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1475373
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1177/00144029251331861
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 18
        StartPage: 359
    Subjects:
      – SubjectFull: Autism Spectrum Disorders
        Type: general
      – SubjectFull: Students with Disabilities
        Type: general
      – SubjectFull: Special Education
        Type: general
      – SubjectFull: Educational Objectives
        Type: general
      – SubjectFull: Student Needs
        Type: general
      – SubjectFull: Elementary Secondary Education
        Type: general
      – SubjectFull: Student Characteristics
        Type: general
      – SubjectFull: Individualized Education Programs
        Type: general
      – SubjectFull: Symptoms (Individual Disorders)
        Type: general
      – SubjectFull: Language Skills
        Type: general
      – SubjectFull: Literacy
        Type: general
      – SubjectFull: Communication Skills
        Type: general
      – SubjectFull: Academic Achievement
        Type: general
      – SubjectFull: Vineland Adaptive Behavior Scales
        Type: general
      – SubjectFull: Social Responsiveness Scale
        Type: general
    Titles:
      – TitleFull: Characterizing the Special Education Goals of Autistic Students: Latent Class Analysis with Demographic and Developmental Covariates
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Matthew C. Zajic
      – PersonEntity:
          Name:
            NameFull: Juliette Gudknecht
      – PersonEntity:
          Name:
            NameFull: Nancy S. McIntyre
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 07
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0014-4029
            – Type: issn-electronic
              Value: 2163-5560
          Numbering:
            – Type: volume
              Value: 91
            – Type: issue
              Value: 4
          Titles:
            – TitleFull: Exceptional Children
              Type: main
ResultId 1