An important aspect of the transition to adulthood for individuals with intellectual and developmental disabilities (IDD) is meaningful access to community-based settings and experiences including postsecondary education. The terminology IDD encompasses individuals with intellectual disability (ID) and/or autism spectrum disorder (ASD), as well as other developmental disorders that affect everyday life related to adaptive functioning, intelligence, and interactions with others (Anderson et al., 2019). Access to postsecondary educational environments such as community colleges, liberal arts colleges, universities, and vocational trade schools is important for young adults with IDD to achieve their educational goals and be fully integrated into society. Importantly, higher education experiences have been made more available to young adults with ID in recent years due in part to improved legislation and policy (Grigal et al., 2021). Young adults with IDD can be on university campuses for a variety of reasons such as pursuing a degree, earning a certificate, or participating in an 18–21 transition program. Regardless of the specific programs in which young adults with IDD are enrolled on a university campus, they will need to navigate a variety of supports and resources to be fully integrated in the postsecondary environment. Of all aspects of life on a university campus, perhaps the most important is developing friendships with same-age peers to create a sense of belonging (Björnsdóttir, 2017; Corby et al., 2022; Prohn et al., 2019).

Research indicates even when young adults with IDD are physically on university campuses, they are vulnerable to social exclusion and loneliness (Prohn et al., 2019). This is concerning because a lack of acceptance from others can contribute to lower mental health outcomes and specifically depression for young adults with ASD with no co-occurring ID (Cage et al., 2018). Relatedly, in a systematic review, Davis et al. (2021) evaluated the academic and non-academic supports provided to students with ASD in higher education as reported by the students. The most prevalent non-academic supports the students with ASD reported were social skills instruction, peer-mediated interventions, and mental health services (Davis et al., 2021). In addition, developing the self-determination skills of college students with IDD is critical to enhance their adult outcomes (Shogren et al., 2018a, c).

Young adults with IDD on university campuses are likely to benefit from social/communication supports, as well as opportunities to interact with same-age peers without disabilities. Pennington and colleagues (2020) conducted a systematic review of interventions to teach social/communication skills to individuals with IDD on university campuses. Although the social/communication intervention literature base is relatively large for learners with IDD and particularly ASD (Steinbrenner et al., 2020), the Pennington review yielded only four peer-reviewed articles. Many communication intervention studies were excluded because they were conducted in K–12 settings rather than on university campuses. The majority of included studies used multicomponent treatment packages consisting of established evidence-based practices (EBPs) for learners with ASD (e.g., technology-aided instruction, video modeling) with researchers serving as intervention agents (Pennington et al., 2020). Because of the overreliance on researchers implementing communication interventions, authors made a specific call for additional research utilizing peers as intervention agents or natural supports in postsecondary environments. Furthermore, interventions conducted in postsecondary settings appear to be less about the selected intervention and more focused on contexts to promote inclusive practices including use of peers (Pennington et al., 2020). Established EBPs in the field of special education, such as peer-mediated instruction and interventions to teach self-determination skills, appear particularly advantageous for young adults with IDD in postsecondary settings.

Peer-Mediated Interventions and Goal Setting

Peer-mediated interventions are a broad category of interventions in which peers assist their classmates in academic or social contexts (Carter et al., 2009). Peer-mediated interventions are commonly used with students with disabilities receiving instruction and support from their peers without disabilities (Carter et al., 2019a; Steinbrenner et al., 2020). The peer serving as the model receives instruction from an adult (e.g., teacher, paraprofessional) on how to best support the student who is learning the new skill. Grounded in behavior analysis and social learning theory, peer-mediated interventions have commonly been used for students with IDD and are an established evidence-based practice for this population (Bowman-Perrott et al., 2023) including individuals with ASD (Chan et al., 2009). The most common goal of peer-mediated interventions is to increase social engagement (Bowman-Perrott et al., 2023; Watkins et al., 2015). For example, several studies by Bambara and colleagues (Bambara et al., 2016, 2018a, 2021 ) used peer-mediated intervention to improve the conversational skills of secondary students with ASD during their regular lunch sessions. Use of a peer to learn conversation skills was often combined with other types of instruction such as written text cue cards to brainstorm conversational topics (Bambara et al., 2016, 2018a). Positive conversational acts included initiating conversation, making on-topic comments, and showing interest in the conversational partner (Bambara et al., 2016, 2018a, 2021). Importantly, several studies have examined the positive impact experienced by typically developing peers participating in such interventions (Carter et al., 2019b; Schaefer et al., 2016; Travers et al., 2023). Benefits of peer-mediated interventions can be reciprocal in nature, meaning beneficial for all students who participate (i.e., those with and without disabilities).

Teaching students with IDD to set goals in their lives falls under the broader category of teaching self-determination skills and is an important component of the transition to adulthood (Wehmeyer & Shogren, 2016). Research indicates students with higher levels of self-determination skills are more likely to experience positive postschool outcomes in the areas of employment and community access (Shogren et al., 2015). Included in the broader category of self-determination, goal setting is an early predictor of post-school success (Test et al., 2009). A follow-up review by Mazzotti et al. (2016) found sufficient evidence for goal setting to be included as a stand-alone predictor for post-school success. Furthermore, goal setting moved from a promising predictor to research-based predictor in an updated review (Mazzotti et al., 2021). Young adults with IDD can set long and short-term goals. For example, Burke et al. (2021) examined the type of goals set by transition-age students with ID through their participation in the Self-Determined Learning Model of Instruction (SDLMI; Shogren et al., 2018b) curricula. Across 1,546 goals analyzed, transition-age youth with ID set goals related to academics, communication, employment, finances, leisure, postsecondary education, transportation, relationships, and more (Burke et al., 2021). In short, the benefits of teaching goal setting skills are vast for young adults with IDD, and goals set by the person themselves are particularly important compared to an outside person setting goals (Carr et al., 2014).

Rationale for the Two Experiments

Providing opportunity and access to typically developing peers for young adults with IDD on university campuses can enhance their inclusive experiences. Peers can provide natural supports and instruction in social/communication skills. The majority of peer-mediated intervention research has occurred in elementary and middle schools with considerably less research occurring in secondary settings (Bowman-Perrott et al., 2023) and almost none in postsecondary settings (Pennington et al., 2020). Students in postsecondary settings have different needs than those in secondary settings as they must demonstrate increased levels of independence and advocacy while also reducing dependence on adults (e.g., parents, teachers). Peers may serve as an important catalyst for change in postsecondary settings given there may be reduced stigma resulting from peer assistance rather than adult assistance. Given the increasing number of young adults with IDD pursuing degrees, certificates, and/or participating in 18–21 transition programs on university campuses (Elias & White, 2018; Grigal et al., 2018), additional peer-mediated intervention research is needed for older learners. The existing peer-mediated intervention research for secondary students indicates positive findings (Bambara et al., 2016, 2021, 2018b); however, more research in postsecondary settings is needed to strengthen the evidence base. We evaluated the effects of structured or natural peer-mediated interventions on the conversation skills of young adults with IDD who were on a university campus as part of their enrollment in an 18–21 transition program. We selected lunchtime as the context to teach conversation skills because lunchtime is a natural time for students on a university campus to socialize with one another (regardless of disability status). The first experiment compared the effects of two structured peer-mediated interventions: peer coaching alone vs. peer coaching and the additive effects of goal setting. The second experiment compared the effects of young adults with IDD speaking with their fellow classmates with disabilities vs. peers without disabilities who provided natural supports. These two experiments represent different intensities of peer-mediated interventions that professionals on university campuses may consider when promoting the social participation of young adults with IDD.

Experiment 1

The purpose of Experiment 1 was to determine which peer-mediated intervention was most effective. Experiment 1 was guided by the following research questions: (a) What are the comparative effects of peer coaching alone vs. peer coaching + goal setting to enhance the conversation skills of young adults with IDD when speaking with typically developing peers during lunch on a university campus? and (b) What is the social validity of a peer coaching and goal setting intervention as reported by the young adults with IDD and typically developing peers who participated in the interventions?

Method

Participants

This study was approved by a university institutional review board prior to data collection and recruitment. The research team recruited young adults with IDD from a public school 18–21 transition program that was housed on a university campus. Research team members visited the classroom of the students to distribute hard copies of the recruitment flier and consent forms. Students enrolled in the 18–21 transition program received special education services as described on their individualized education programs (IEPs). The purpose of the program was to prepare students for competitive employment in their communities after exiting the public school system. The majority of students were enrolled in the program for two years and had IEP goals related to job skills training, social skills, and community participation. All young adult participants met the following inclusion criteria: (a) diagnosis of IDD, including ASD and/or ID as confirmed by formal documentation such as a medical report or current IEP, (b) between the ages of 18–26, and (c) indicated a desire to eat lunch with peers and potentially improve their conversation skills. Individuals with co-occurring conditions and limited verbal language skills were eligible to participate. Young adults or their parents/guardians provided signed consent to indicate their willingness to participate in all procedures required for the study and particularly the audio recording of the conversation sessions. Participants whose parents served as their legal guardians participated in an assent process with the first author.

Research team members administered the Social Responsiveness Scale, 2nd edition (SRS-2; Constantino & Gruber, 2012) to describe the current social/communication skills of each participant (see Table 1). The SRS-2 measures social responsiveness and reports scores within normal, moderate, or high range of social deficiencies. Parents/guardians completed the Adult (Relative/Other Report) form which gives parents/guardians the opportunity to rate the participants’ social abilities across a variety of contexts. Participants completed the Adult (Self-Report) form with assistance from the first author and the 18–21 transition program teacher. Parents/guardians and participants may report different perceptions regarding the social functioning of the participants. All participants used verbal speech as their primary mode of communication and are described below using pseudonyms.

Table 1 Participant SRS-2 standardized assessment results
Full size table

Jackson

Jackson was a 21-year-old Black male with ASD and ID. His IEP indicated that he was independent and able to advocate for his preferences and ask questions when needed. Jackson reported small talk as a weakness of his and a desire to improve socialization skills. He was observed to have moderate vocal conversation skills during baseline observations. Jackson’s postsecondary goal was to obtain employment at Amazon after exiting the transition program.

Tony

Tony was a 21-year-old White male with ASD. He navigated his schedule independently and regularly walked the campus unsupervised with his peers. As reported on his IEP, Tony required support answering questions beyond a yes/no response. Tony exhibited echolalia and limited reciprocal conversation during baseline observations. His postsecondary goal was to continue employment at a bowling alley with a job coach after exiting the transition program.

Peer Coaches

We recruited undergraduate students to serve as peer coaches from the same university where the 18–21 transition program was housed. Research team members visited education undergraduate classes in person to distribute recruitment flyers and consent forms. Peer coaches met the following inclusion criteria: (a) ages 18 or older; (b) no documented disability that would impede their ability to interact with participants with IDD; and (c) indicated a desire to practice conversation skills with same-age peers IDD. Nine students provided signed consent and gave permission to audio record the conversation sessions. All peer coaches were female and between the ages of 18–22. Three studied special education, and six were in related fields such as elementary education and psychology. One peer coach was Hispanic, one was Black, and seven were White. In general, peer coaches reported having minimal to no experience interacting with same-age peers with disabilities prior to participation in the study. Both participants with IDD (i.e., target participants) and peer coaches received a $25 stipend for their time and participation.

Setting and Materials

All data collection sessions occurred during regular lunchtime according to the 18–21 transition program schedule. Target participants and peer coaches ate lunch together in the dining area of the university student union, which was a three-level common space (196,000 square feet) that included a movie theater, game hall, convenience store, and leisure areas. The dining area was the same location in which participants with IDD ate lunch together prior to the study. The student union is open to all individuals on the university campus and frequented by students, faculty, and community members. Within the student union, the dining hall is a large area that consists of several tables and chairs with a variety of food kiosks and restaurants organized around the tables. Materials used in this study included audio recording devices, peer coaching strategies worksheet (see Fig. 1), and goal setting and conversation rating scale worksheet (see online Supplementary Material). The audio recording devices were Sony ICD-PX370 Mono Digital Voice Recorders that were 1.5 × 4.5 inches. The devices included built-in USB ports to transfer the audio files to a computer for storage and data collection. The small size of the devices allowed research team members to position them on dining hall tables in an unobtrusive manner that did not interfere with conversations.

Fig. 1
figure 1

Peer coaching strategies worksheet. Source. Bambara et al. (2016) and adapted from Thiemann and Goldstein (2001, 2004)

Full size image

Dependent Variable

The dependent variable was conversational engagement as measured by partial interval recording procedures. Each conversational session was 10 min in duration and audio recorded for data collection purposes. Research team members listened to the audio recordings to record if a positive conversational act was delivered by the target participant within the 10-s interval. Data collectors marked a + on the data collection page if the participant exhibited a positive conversational act during the 10-s interval. We created operational definitions of conversational acts by adapting definitions from other interventions studies to teach conversation skills (Bambara et al., 2016, 2018a, 2021; Gregori et al., 2021). See Table 2 for the operational definitions and examples used in our study. Positive conversational acts included initiation statements, any type of question, and verbal comment or response. Likewise, data collectors marked a—on the data collection page if exhibited an absence of a conversational act or negative conversational act during the 10-s interval. These included no response, non-engagement, and off-topic responses. It is important to note if the target participant was quiet, this was recorded as absence of a conversational act. The 10-min conversation sessions divided into 10-s intervals generated 60 datum of conversational acts. The percentage of overall conversational engagement for each session was calculated by the following formula (Intervals with positive conversational acts / 60 = % of conversational engagement).

Table 2 Operational definitions and examples
Full size table

Experimental Design and Data Analysis

We, a priori, selected an alternating treatments design for Experiment 1 to compare the effects of peer coaching alone vs. peer coaching + goal setting. Baseline conditions are optional but recommended in alternating treatments designs (Cooper et al., 2020; Wolery et al., 2018). We planned for two adjacent baseline conditions. The first baseline condition (A1) consisted of target participants eating lunch with other 18–21 transition program classmates, as this was representative of routine practices prior to the study. The second baseline (B1) consisted of target participants eating lunch with the peer coaches but no explicit coaching strategies implemented (i.e., natural supports condition). For the alternating treatments condition (C1), we randomized the sequence of implementation of the peer coaching intervention alone vs. peer coaching + goal setting (Wolery et al., 2018). Each intervention was implemented no more than two consecutive times. Rather than a best-treatment condition, we implemented a student choice condition (D1), as the final condition in which participants selected if they wanted to participate in the peer coaching alone or peer coaching + goal setting interventions.

To analyze the graphed behavioral data in Experiment 1, we implemented a systematic process of visual analysis as described by Barton and colleagues (2018). We analyzed the level, trend, and variability/ stability in each condition to determine if there were clear data patterns within conditions. Experiment 1 Meets What Works Clearinghouse (WWC) Standards Without Reservations (Institute of Education Science, [IES], 2022) because there was a minimum of five data points for each intervention in the alternating treatments condition. We utilized summative visual analysis to determine if peer coaching alone or peer coaching + goal setting was most effective for Jackson and Tony. We also conducted visual analysis of the adjacent conditions of baseline and natural supports condition, as well as the natural supports condition compared to the alternating treatments condition.

Procedures

Peer Coach Training

All peer coaches participated in a one time, one hour training with research team members to learn the specific conversation coaching strategies adapted from the academic literature (Bambara et al., 2016; Thiemann & Goldstein, 2001, 2004). Researchers explained the purpose of the study and provided coaches with the peer coaching strategies worksheet and goal setting and conversation rating scale worksheet. The peer coaches learned two overarching strategies: strategies to support conversation and strategies to promote initiation and follow-up questions. To support conversation, peer coaches were trained to (a) Look, listen, and wait, (b) Gain your partner’s attention, (c) Keep the conversation going, (d) Help your partner respond, and (e) Move on. To support the promotion of initiation, peer coaches were trained to prompt their peers by saying, “You start,” “Ask me a question,” or “Tell me something.” To support the promotion of follow-up questions, peer coaches were trained to pose a question and make a statement, and if the peer did not ask, to then prompt them by saying, “Ask me back” or “Ask me about it.” Researchers modeled the coaching strategies and then peer coaches practiced by having a conversation with researchers who gave feedback. Researchers also explained the goal setting and conversation rating scale procedures to peer coaches. During the training, it was made clear the peers were not meant to rate the quality of the conversations themselves; rather, the purpose was to provide support to target participants as they evaluated their own conversation skills. Peer coaches could ask questions and receive clarification as needed any time during the training.

General Procedures

We rotated the use of peer coaches throughout all conditions of the study according to their schedule and availability. Participants with IDD routinely ate lunch at four or five tables in the same area of the student union. For data collection purposes, research team members asked target participants and peer coaches to sit next to one another. A minimum of two people sat at one table together. Research team members followed the lead of target participants regarding where they wanted to sit. Prior to starting the audio recording device, research team members asked all participant types if it was okay to record their conversational session. If participants said yes, research team members placed the audio recording device in front of participants and recorded for a duration of 10 min. Participants could decline to be audio recorded or make a request to eat lunch in a different location on any given day and still participate in the study.

Experiment 1 Baseline

Jackson and Tony ate lunch with their 18–21 transition program classmates during baseline, which represents their regular routine prior to participation in the study. Peer coaches were not present in the student union for the baseline condition. Neither peer coaching nor goal setting materials were present. Research team members audio recorded the conversations between Jackson and Tony with their peers with disabilities. No prompting or support from research team member related to conversation skills was provided.

Experiment 1 Natural Supports Condition

Jackson and Tony ate lunch with a peer coach during the natural supports condition. However, the peer coaches were told to not provide any specific peer coaching strategies or implement the goal setting intervention as learned during the peer coach training. Target participants and peer coaches could ask one another questions and speak on any topic they desired. Research team members operated the audio recording devices but not did provide any coaching or support related to conversation skills.

Experiment 1 Alternating Treatments Condition

We implemented the peer coaching alone intervention and the peer coaching + goal setting in a rapid manner for the alternating treatments condition for Jackson and Tony.

Peer Coaching Alone

The peer coaching alone sessions consisted of the peer coach delivering the specific strategies to support conversation as trained. Researchers gave peer coaches a hard copy of the peer coaching strategy worksheet (see Fig. 1) to refer to during the sessions if they desired. Peers implemented overall strategies to support conversation (e.g., gaining the partner’s attention), as well as a minimum of 1 strategy to promote initiations (e.g., “You start”) and a minimum of 1 strategy to promote follow-up questions (e.g., “Ask me about it”). Number of specific coaching prompts provided by the peers ranged according to needs of the target participants, but an approximate range was 2–5 for each conversational session.

Peer Coaching + Goal Setting

The peer coaching + goal setting sessions consisted of the same peer coaching strategies described above with the added component of goal setting. Researchers provided both the peer coaching strategy worksheet and goal setting worksheet to peer coaches during these sessions. Prior to the audio recording for data collection purposes, target participants selected a goal for themselves for that specific conversational session and wrote it on the worksheet (e.g., “I will ask my conversational partner at least 3 questions”). After the conversation ended, target participants rated the quality of the conversation using a 4-point scale for conversation initiations, follow-up questions, showing interest, appropriate statements, and overall quality of the conversation. On the rating scale, 1 indicated that the conversation was not so good with many conversational roadblocks, 2 indicated that the conversation was okay with some conversational roadblocks, 3 indicated that the conversation was a pretty good and engaging conversation, and 4 indicated that the conversation was really fun, good, and engaging. Target participants and peer coaches discussed the scores and quality of the conversation together with participants encouraged to be self-reflective rather than peer coaches being evaluative.

Experiment 1 Student Choice Condition

We implemented a student choice condition by asking Jackson and Tony which intervention they preferred: peer coaching alone vs. peer coaching + goal setting. The first author explained they could choose either intervention for the last 5 sessions. This condition was incorporated to enhance the social validity of the intervention and to promote decision making skills within the participants with IDD.

Interobserver Agreement and Procedural Fidelity

Research team members participated in training to ensure consistent coding of the operational definitions of conversational acts to report overall percentage of conversational engagement for each session. The training involved research team discussing all operational definitions in Table 2 using examples and non-examples and then practicing scoring with three sample audio recordings. The sample audio recordings were from a previous conversation study in which young adults with IDD spoke to typically developing peers. All research team members calculated interobserver agreement (IOA) using a point-by-point agreement formula (Cooper et al., 2020) with the first author serving as the primary coder. The IOA training occurred prior to the first baseline session for data collection purposes with a minimum of 80% agreement achieved across all team members. Two research team members conducted IOA of the primary dependent variable for a minimum of 30% of sessions across all conditions and participants of the study. Mean IOA for the entire study for Tony was 88% (range = 78–100%) and for Jackson was 92% (range = 87–98%).

We created procedural fidelity checklists for all conditions of the study for Experiment 1 (see Supplementary Online Material). The peer coaching alone checklist consisted of five steps, and the peer coaching + goal setting checklist consisted of nine steps. The mean procedural fidelity for Experiment 1 for Jackson was: Jackson baseline condition 100%; Jackson natural supports condition 100%; Jackson peer coaching alone condition 94% (range = 88%–100%); Jackson peer coaching + goal setting condition 98% (range = 88%–100%); and Jackson student choice condition 100%. The mean procedural fidelity of Experiment 1 for Tony was: Tony baseline condition 100%; Tony natural supports condition 100%; Tony peer coaching alone condition 97% (range = 88%–100%); Tony peer coaching + goal setting condition 99% (range = 88%–100%); and Tony student choice condition 100%.

Social Validity

We administered social validity questionnaires and conducted exit interviews to evaluate potential satisfaction with the interventions for all participant types. The questionnaire was Likert-style in which target participants and peer coaches rated their level of agreement using the anchors of strongly disagree, disagree, agree, and strongly agree. For participants in Experiment 1, the questions focused on experiences with the peer coaching and goal setting interventions. Each social validity questionnaire also included 4 open-ended questions in which research team members verbally asked all participant types and audio-recorded their responses. We transcribed the social validity exit interviews and conducted inductive qualitative coding procedures (Bhattacharya, 2017) to identify themes reported across all participant types. The open-ended questions focused on what component of the conversation sessions all participants liked the best and what aspects were challenging. We also asked all participant types for suggestions to expand or continue the research study at another university. The qualitative coding procedures consisted of the first author reading each social validity exit interview transcription in full and conducting inductive analysis to identify “nodes” or common themes reported across several participants (Bhattacharya, 2017). The first author collected example quotes from participants that supported each identified node and wrote a brief description of each node. Then, the first author shared nodes and descriptions with other research team members to participate in a debriefing session and reflect on the process. The second and fourth authors independently coded all transcriptions to ensure trustworthiness of the themes identified (Brantlinger et al., 2005).

Experiment 1 Results

Jackson

Figures 2, 3, 4 and 5 present graphed data for all participants. Jackson’s mean conversational engagement during the initial baseline condition was 8% (range = 2–17%). Jackson’s initial natural supports condition indicated a mean conversational engagement of 32% (range = 12–61%) with a stable data path. For the alternating treatments condition, Jackson’s mean conversational engagement was 56% (range = 23–70%) for the peer coaching alone intervention and 57% (range = 35–77%) for the peer coaching + goal setting intervention. There was not a clear separation between the data paths in the alternating treatments condition. Upon return to the natural supports condition, Jackson’s mean conversational engagement was 22% (range = 8–53%) with a stable data path. Jackson selected peer coaching + goal setting for the student choice condition, and his mean conversational engagement was 45% (range = 15–70%). Summative visual analysis for Jackson indicated no clear difference between the peer coaching alone intervention vs. the additive effects of goal setting. However, Jackson indicated a preference for peer coaching + goal setting.

Fig. 2
figure 2

Jackson’s percentage of conversational engagement (Experiment 1). Note. PC = Peer coaching. Open circles = Peer coaching alone. Open triangles = Peer coaching + goal setting

Full size image
Fig. 3
figure 3

Tony’s percentage of conversational engagement (Experiment 1). Note. PC = Peer coaching. Open circles = Peer coaching alone. Open triangles = Peer coaching + goal setting

Full size image
Fig. 4
figure 4

Lola’s percentage of conversational engagement (Experiment 2). Note. Closed circles = Access to peers with disabilities. Closed squares = Access to peers without disabilities

Full size image
Fig. 5
figure 5

Nathan’s percentage of conversational engagement (Experiment 2). Note. Closed circles = Access to peers with disabilities. Closed squares = Access to peers without disabilities

Full size image

Tony

Tony’s mean conversational engagement during the initial baseline condition was 6% (range = 0–17%). Tony’s initial natural supports condition mean was 19% (range = 10–32%) with a stable data path. For the alternating treatments condition, Tony’s mean conversational engagement was 51% (range = 25–82%) for the peer coaching alone intervention and 48% (range 33–78%) for the peer coaching + goal setting intervention. There was not a clear separation between data paths in the alternating treatments condition. Upon return to the natural supports condition, Tony’s mean conversational engagement was 12% (range = 5–20%) with a stable data path. Tony selected peer coaching alone for the student choice condition, and his mean conversational engagement was 46% (range = 40–48%), as well as an immediacy of change was observed. Summative visual analysis for Tony indicated no clear difference between the peer coaching alone intervention vs. the additive effects of goal setting. However, Tony reported a preference for peer coaching alone.

Social Validity

Jackson and Tony Social Validity Results

Young adult social validity questionnaire results can be found in Table 3. Jackson and Tony agreed they felt independent when they participated in conversation during lunch. They strongly agreed they would like to continue eating lunch with a peer after the conclusion of the study. Jackson and Tony agreed they liked setting goals for themselves and strongly agreed they liked receiving coaching from a peer to improve their conversational skills. Jackson reported the following related to his participation in the study: “I will be better communicating at work.” Tony selected peer coaching alone for the final student choice condition whereas Jackson selected peer coaching + goal setting.

Table 3 Young adult social validity questionnaire results
Full size table

Peer Coach Social Validity Results

Table 4 presents the peer coach social validity questionnaire results, and Fig. 6 includes sample quotes aligned to each identified node. All peer coaches strongly agreed or agreed they liked being a conversational partner, and they felt prepared when they had the conversations during lunch. In addition, peer coaches reported a desire to continue eating lunch with the target participants after the conclusion of the study and to continue being a coach in different settings. Results of the qualitative coding of the exit interview transcriptions yielded the following 4 nodes: challenges or difficulties presented during the conversations, positive outcomes reported (including peer friendships or skills gained), ideas for training and future studies, and prior experience of the peer coaches related to working with peers with disabilities. Some of the challenges reported by the peer coaches included the loud environment in the student union, generating topics, maintaining back-and-forth conversation, and mixture of different people at the table. For example, Lacey, a 19-year-old elementary education undergraduate student said, “Yeah, I didn’t want to keep asking the same questions but sometimes I didn’t know what else to ask because I didn’t know anything more about them.” Peer coaches reported the most positive aspect of participating in the study was the friendships they created. Vanessa, a 20-year-old special education undergraduate student noted that all the peer coaches were approximately the same age as the young adults with IDD and this assisted to keep the conversations natural. Vanessa stated, “…I think that we were working because we were the same age as all the other students, and we’re all college students. We have things to relate together and I think that helped a lot in the conversations. Conversations were natural. A natural conversation you would have with your friends in college.”

Table 4 Peer coach social validity questionnaire results
Full size table
Fig. 6
figure 6

Peer coach and young adult example quotes

Full size image

Overall, the peer coaches reported they found the training they received to be sufficient and had relevant ideas for future studies. Addison, an 18-year-old elementary education student, stated the following about the training: “I think it was very short, but like very simple, but in a good way if that makes sense.” Several peer coaches reported matching interests across young adults with IDD and peer coaches would be advantageous for future studies. For instance, when asked for ways to improve the study, one of the peer coaches (Lacey) stated: “…to pair people that have the same interests maybe because I feel like they are able to connect a bit more.” Finally, several peer coaches reported little to no prior experiences with same-age peers with disabilities and their participation in the study was positive. Addison stated, “I felt like it was a great way to diversify my experience in life basically because as a person without a disability, I’m not really exposed to people with disabilities. So, I felt like it was a great opportunity to diversify my experiences with people with disabilities and create like you know a friendship with them.”

Experiment 2

The purpose of Experiment 2 was to determine if natural supports were sufficient for some young adults with IDD rather than structured peer-mediated interventions. Experiment 2 was guided by the following research questions: (a) Do young adults with IDD demonstrate higher levels of conversational engagement when speaking with peers with or without disabilities? and (b) What is the social validity of natural supports to enhance conversation skills as reported by young adults with IDD and peer coaches who participated in the intervention?

Method

We recruited participants in Experiment 2 from the same 18–21 transition program on the university campus described above in Experiment 1. The peer coaches without disabilities who participated in Experiment 2 were the same peer coaches in Experiment 1. The student union setting and all materials used in Experiment 2 were the same as Experiment 1. The dependent variable of overall conversational engagement and general procedures for Experiment 2 (i.e., rotation of peers, use of audio recording devices) were also the same as Experiment 1.

Lola

Lola was a 22-year-old Black female with ID. As reported on her IEP, she regularly walked the university campus unsupervised and followed a schedule independently. She enjoyed spending time in a student lounge area on campus. During baseline observations, Lola exhibited high levels of vocal communication and reported a desire to improve her conversation skills to initiate conversation with others. Her teacher described Lola as friendly, and her postsecondary goal was to obtain competitive employment at a local restaurant after exiting the transition program.

Nathan

Nathan was a 21-year-old Black male with ID. He was described as friendly and independent on his IEP. Nathan’s teacher reported that he had taken a leadership role in his internship on campus to help other students complete tasks. Nathan exhibited moderate levels of vocal communication during baseline observations and reported a desire to improve his conversation skills. Nathan's postsecondary goal was to obtain competitive employment in a retail store after exiting the transition program.

Experimental Design and Data Analysis

Due to unexpected findings during the second baseline phase (B1) for Lola and Nathan (i.e., natural supports condition), we ad hoc decided to implement a withdrawal design instead of the planned alternating treatments design for these two participants. The withdrawal design consisted of the following two conditions: (A) Lola and Nathan ate lunch with their fellow 18–21 transition program classmates with disabilities or (B) Lola and Nathan ate lunch with peer coaches without disabilities but no explicit strategies were provided. Each condition was implemented for a total of 3 phases for an ABABAB design with 5 opportunities for demonstrations of effect.

We analyzed the graphed behavioral data in Experiment 2 using the same systematic process of visual analysis as Experiment 1 (Barton et al., 2018). We conducted within condition visual analyses of level, trend, and variability/stability in each condition. We changed conditions when data in the phase were stable and conducted more sessions until the data stabilized. Experiment 2 Meets WWC Standards With Reservations (IES, 2022) because the initial baseline had only three data points. We evaluated Experiment 2 data to determine if there was a functional relation between conversational engagement and type of peers serving as conversational partners (i.e., peers with or without disabilities).

Procedures

Experiment 2 Baseline

Lola and Nathan ate lunch with their 18–21 transition program classmates with disabilities during the baseline condition. Participants sat in close proximity to one another at two or three tables in the dining area of the student union. Peer coaches were not present in the student union for the baseline condition, and research team members gave no specific instruction or support related to conversation skills. Lola, Nathan, and their classmates with disabilities could ask questions to one another and speak on any topic they desired.

Experiment 2 Natural Supports Condition

Lola and Nathan ate lunch with peer coaches during the natural supports condition, but the peers did not implement any specific coaching strategies or the goal setting intervention. In particular, peers were told to not provide prompts to promote initiation (e.g., “You start”) or to promote follow-up questions (e.g., “Ask me back”) as these were coaching strategies learned during the training. Research team members told the peers to interact with Lola and Nathan in a natural manner. All participant types (Lola, Nathan, and peer coaches) could ask questions to one another and speak on any topic. No peer coaching or goal setting materials were present during the natural supports condition.

Experiment 2 Withdrawal Condition

Experiment 2 withdrawal condition was identical to the initial baseline condition. The peer coaches were not present and Lola and Nathan ate lunch with their classmates with disabilities. If a target participant inquired about a peer coach, research team members said peer coaches would be eating lunch together soon according to their schedule.

Interobserver Agreement and Procedural Fidelity

Research team members conducted IOA in Experiment 2 in the same manner as Experiment 1. Mean IOA across all conditions of Experiment 2 for Lola was 81% (range = 81–93%) and Nathan was 90% (range = 82–96%). Likewise, we created procedural fidelity checklists for both conditions of Experiment 2 (see Supplementary Online Material). Both the baseline and natural supports checklists consisted of five steps. The mean procedural fidelity for Experiment 2 for Lola was as follows: Lola baseline condition 100%; and Lola natural supports condition 100%. The mean procedural fidelity of Experiment 2 for Nathan was as follows: Nathan baseline condition 100%; and Nathan natural supports condition 100%.

Experiment 2 Results

Lola

Lola’s mean conversational engagement during baseline (A1) was 21% (range = 10–37%). During the initial natural supports condition (B1), her mean conversational engagement improved to 70% (range = 40–93%), and an immediacy of change was observed. Upon withdrawal of the typically developing peers (A2), Lola’s mean conversational engagement reduced to 32% (range = 0–82%) although no immediacy of change was observed in this condition. When the peer coaches returned for the B2 condition, a large immediacy of change was observed, and Lola’s mean conversational engagement increased to 74% (range = 63–97%). In the final withdrawal condition (A3), Lola’s mean conversational engagement was reduced to 26% (range = 13–43%). During the final natural supports condition (B3), the mean was 60% (range = 35–98%). There were clear changes in level for four out of the five phase changes. Summative visual analysis indicated a functional relation was present for Lola given four demonstrations of effect were observed. Specifically, Lola demonstrated higher conversational engagement when speaking with the peer coaches without disabilities who provided natural supports rather than her 18–21 transition program classmates with disabilities.

Nathan

Nathan’s mean conversational engagement during baseline (A1) was 15% (range = 0–37%). During the initial natural supports condition (B1), his mean conversational engagement improved to 74% (range = 41–93%), and an immediacy of change was observed. Upon withdrawal of the peer coaches (A2), Nathan’s mean conversational engagement decreased to 22% (range = 0–42%), and an immediacy of change was observed in this condition. When the peer coaches returned (B2), an immediacy of change was observed and Nathan’s mean conversational engagement increased to 71% (range = 53–98%). During the final withdrawal condition (A3), Nathan’s mean conversational engagement decreased to 20% (range = 8–45%). During the final natural supports (B3), Nathan’s mean was 78% (range = 50–87%), and an immediacy of change was observed. Summative visual analysis indicated a functional relation was present for Nathan given four demonstrations of effect were observed. Nathan demonstrated higher levels of conversation engagement when speaking with the peer coaches who provided natural supports rather than his 18–21 transition program classmates with disabilities.

Social Validity

Lola strongly agreed with all social validity questionnaire statements, and Nathan agreed with all statements. Both Lola and Nathan stated they liked having conversations with peer coaches, and they would like to have more opportunities for conversations in different settings. Both Lola and Nathan stated they felt independent when participating in conversation during lunch, and they would like to continue eating lunch with a peer coach after the conclusion of the study. When asked what they liked best about participating in the conversations, Nathan said during the exit interview, “Talking about stuff and having fun eating lunch, and talking to my peers.” Lola reported that, “I liked that you guys came to the student union and had all your students come and talk to us…lunch bunch helped.” Please see above for the peer coach social validity results reported in Experiment 1 which were the same for Experiment 2.

Overall Discussion

The peer-mediated intervention in its various forms enhanced the conversation skills of all participants with IDD in Experiments 1 and 2. First, results of Experiment 1 indicated no clear difference between the peer coaching alone intervention vs. peer coaching + goal setting. Jackson and Tony demonstrated lower conversational engagement during the natural supports condition, which indicated they were not likely to improve their conversation skills without access a structured peer-mediated intervention. Jackson selected the peer coaching + goal setting intervention during the student choice condition whereas Tony selected peer coaching alone. Findings from the alternating treatments condition indicated peer coaching and goal setting were more effective than natural supports for Jackson and Tony. Results of Experiment 2 indicated Lola and Nathan demonstrated greater conversational engagement when speaking with peer coaches without disabilities compared to 18–21 transition program classmates with disabilities. The unexpectedly high conversational engagement during the first natural supports condition for Lola and Nathan was the reason we implemented a withdrawal design in Experiment 2 rather than the planned alternating treatments design in Experiment 1. Findings from Experiment 2 indicate some young adults with IDD on university campuses may benefit from mere opportunities to interact with typically developing peers in which the peers provide natural supports rather than a structured intervention. In the natural supports condition in Experiment 2, the peer coaches were being kind and inclusive to Lola and Nathan, as well as showing interest in them. Prior to their participation in the study, Lola and Nathan only ate lunch with their 18–21 transition program classmates, despite the fact they had access to typically developing peers in the student union. Postsecondary education programs for students with IDD should provide opportunities for students with and without disabilities to interact with one another. The demand for social conversation increases as young adults interact by spending time with one another and talking (Bambara et al., 2018b). The different outcomes of Experiment 1 and 2 can be framed around intervention intensity, that is Jackson and Tony needed more intensive interventions than Lola and Nathan. Dimensions of intervention intensity described by Fuchs et al. (2017) relevant to this study are behavioral support and individualization. The peer coaches in both Experiments 1 and 2 were able to individualize the levels of behavioral support in the form of conversational coaching provided to all four participants with IDD.

Results from both experiments provide supporting evidence that peer-mediated interventions are likely to be effective for older learners with IDD in postschool environments in a similar manner to younger learners in K-12 school settings. Contextual fit and preferences of older learners must be considered when designing and implementing interventions in postschool environments such as university campuses, community-based locations, or employment settings. Results of Experiment 1 (the alternating treatments design) did not indicate a notable difference between peer coaching alone vs. peer coaching + goal setting. Peer coaching alone was equally as effective as the additive effects of goal setting. Accordingly, service providers of young adults with IDD may consider using a single intervention before selecting an intervention package (e.g., peer coaching + goal setting) to leverage time and resources as well as using learner preferences to guide intervention selection. Service providers can also ask young adults with IDD if they prefer to participate in interventions alone, with same-age peers with disabilities, same-age peers without disabilities, or combination approach. Given that preferences and interests of students must be incorporated in the transition planning process (Individuals with Disabilities Education Improvement Act [IDEIA], 2004), it is logical and important to consider students’ satisfaction with their post-school activities.

We evaluated the social validity of the peer-mediated interventions in Experiments 1 and 2 using Likert-style questionnaires and exit interviews. Both target participants and peer coaches reported overall high satisfaction with the peer-mediated intervention and their participation in the study. However, a traditional researcher-created and administered social validity questionnaire such as the one used in this study is a subjective measure in which participants are likely to report high satisfaction (Ledford et al., 2016). Results of the exit interviews yielded more nuanced social validity findings, such as challenges experienced during the conversations, friendships formed, desire for natural interactions, and ideas for future implementation. In particular, several peer coaches suggested matching interests would have been advantageous to improve interactions. This finding is consistent with other peer-mediated intervention research that describes the importance of matching interests (e.g., Schwartz et al., 2020). Matching can be conducted related to shared hobbies, interests, and/or experiences.

Experiments 1 and 2 have implications for service providers who seek to enhance the social experiences of young adults with IDD on university campuses. The 18–21 transition program in this study can be classified as college-based transition services which entail transition-age students with disabilities receiving their final years of transition services on a college or university campus (Grigal et al., 2018). These programs are funded by public schools as students are still receiving special education services as described by IDEIA (2004). Although college-based transition services can provide rich experiences to students with disabilities in the college environment (e.g., job internships, access to university facilities), students enrolled in such programs are vulnerable to social exclusion from students without disabilities who attend the same university. Specifically, Jackson, Tony, Lola, and Nathan did not interact with college students without disabilities on a regular basis. Rather, they participated in classroom activities with their 18–21 transition program classmates, attended job internships on the university campus at various placements (e.g., library, recreation center), and then went home. Moreover, the 18–21 transition program students followed the daily high school schedule rather than a traditional university schedule. As a result, they arrived on the campus much earlier and also departed earlier than university students without disabilities. Lunchtime was truly the only free time in the students’ regular school day. However, despite typically developing peers being in close proximity to them, all target participants ate lunch with their 18–21 transition program classmates or alone prior to their participation in this study. Results of both experiments indicated conversational engagement was higher when the target participants conversed with the peer coaches without disabilities, which highlights the importance of fostering interactions between students with and without disabilities on university campuses.

Findings from this study have implications for students enrolled in postsecondary programs other than college-based transition services, such as non-degree seeking certificate programs. Think College (2024) reported there are currently 339 postsecondary education programs across the United States for students with ID. These programs are affiliated with an accredited institution of higher education and can be housed at two-year colleges or four-year universities. Completing a non-degree certificate program can contribute to higher rates of employment (Burns et al., 2020) and provide valuable experiences such as living in a dorm, taking classes, and obtaining job skills. Regardless of the increased prevalence of young adults with IDD pursuing postsecondary education, their social experiences and inclusion in the university environment and community has been understudied in the academic literature (Prohn et al., 2018). Both experiments contribute to the peer-mediated intervention literature base in postsecondary settings. Findings were consistent with previous peer-mediated intervention research for secondary students with IDD (Bambara et al., 2016, 2018a, 2021), as well as other goal-setting studies (Burke et al., 2021; Carr et al., 2014). Developing the skills necessary for positive social relationships is one of the foundational elements that determines if college students with IDD are socially included or excluded (Prohn et al., 2019). In addition, developing self-determination skills is important to enhance the postschool outcomes of young adults with disabilities including IDD (Shogren & Wehmeyer, 2016). Although much of the self-determination literature focuses on long-term goals (e.g., attending college, obtaining a job), the short-term goals that young adults with IDD set for themselves in the study can potentially develop important self-determination and goal setting skills. Essentially, findings from this study indicate service providers in different types of programs on university campuses must plan for different types of peer support: (a) explicit instruction in the area of social/communication skills such as peer coaching or goal setting and (b) natural supports by providing opportunity and access to typically developing peers. In either type of peer support, service providers will need to train the peers without disabilities by modeling how best to interact with and support the young adults with IDD in an age-appropriate and non-stigmatizing manner.

Limitations and Suggestions for Future Research

Results of this study should be interpreted while also considering several limitations that can guide future research. First, collecting primary dependent variable data on target participants with IDD and not all study participants (i.e., peer coaches without disabilities) narrows the scope of evaluating conversational behavior and potential effects of different types of conversational partners. Accordingly, future researchers may consider also measuring the conversational engagement of all individuals participating in the conversation, not only the target participants with disabilities. Second, all data collectors were aware of the conditions of the study when coding the conversational sessions. Future researchers may consider using naïve data collectors who are unaware of the study conditions when coding. Third, in Experiment 1, generalization and maintenance data were not collected due to the semester ending and participants graduating from the 18–21 transition program. However, across both experiments young adults conversed with different peers across sessions promoting generalized responding across people. Future research is needed to evaluate if the conversational engagement skills maintain over time and generalize to other settings and people. Although the withdrawal design used in Experiment 2 demonstrated a strong functional relation, it did not allow for generalization or maintenance data to be collected on the young adults’ conversational engagement.

Fourth, Experiment 1 demonstrated that both the peer coaching alone and the peer coaching + goal setting condition were effective in increasing conversational engagement; however, goal setting alone was not evaluated. Future research should evaluate if goal setting alone can increase conversational engagement of young adults with IDD. Additionally, within the alternating treatments design it is possible that there were multiple treatment effects wherein participants carried over skills learned during one condition to another. Due to each condition containing peer coaching, it is more likely that goal setting carried over to the peer coaching alone condition if there were multiple treatment effects. Finally, additional research is needed to examine peer-mediated interventions in other contexts relevant to the lives of young adults with IDD such as employment settings (Bowman-Perrott et al., 2023). Implementing similar peer coaching strategies used in this study, peer networks could be established with co-workers in community-based employment settings to provide on-the-job supports to employees with IDD.

Conclusion

As increasing numbers of young adults with IDD are on university campuses for a variety of pursuits, it is essential that researchers and service providers identify ways to support their access to positive and inclusive experiences. Young adults with IDD can benefit from structured opportunities to interact with typically developing peers throughout their school day. Examples of such opportunities beyond eating lunch include attending a sporting event, participating in activities at a campus recreation center, studying together at the library, or walking to various points of interest on campus. Young adults with IDD will likely benefit from both opportunity and instruction in the form of individually designed, socially valid interventions. Results of this study indicated the young adults with IDD benefited from varying levels of support. That is, some target participants required only natural supports, and other participants needed additional supports through peer coaching or goal setting to improve their conversation skills. With wide access to same-age peers, differentiated supports are well-suited to meet the needs of young adults with IDD to be fully engaged and meaningfully included in everyday activities on a university campus.