Rahn, Naomi L., PhD; Coogle, Christan Grygas, PhD; Ottley, Jennifer R., PhD

SPECIAL EDUCATION law mandates that children with disabilities receive their education in the least restrictive environment (^{Individuals With Disabilities Education Improvement Act, 2004}). Because of this and our shifting social values regarding inclusive practices, children with and without disabilities are more frequently learning together (^{Division for Early Childhood/National Association for the Education of Young Children, 2009}). With this inclusive shift, early childhood special education (ECSE) teachers are required to address a variety of learning outcomes for children in early childhood classrooms. In classrooms serving young children with special needs, the ECSE teachers are encouraged to use naturalistic instruction (NI), which provides opportunities for children to practice developmentally and socially meaningful skills within the context of everyday activities (^{Division for Early Childhood [DEC], 2014}). Common features of NI include (a) implementation during typical routines and activities, (b) intentionally designed learning opportunities consisting of an antecedent and a consequence, (c) focusing on useful and important skills to ensure the child's participation in activities, and (d) implementation by familiar adults (^{Snyder, Rakap, et al., 2015}).

Naturalistic instruction has been identified as effective in promoting the learning of young children with special needs (^{Hart & Risley, 1975 ; Rule, Losardo, Dinnebeil, Kaiser, & Rowland, 1998}) and has been used to teach a variety of skills (^{Horn, Lieber, Li, Sandall, & Schwartz, 2000 ; VanDerHeyden, Snyder, Smith, Sevin, & Longwell, 2005}). In a recent review of NI, ^{Snyder, Rakap, et al. (2015)} found that almost all young children with special needs enrolled in the studies acquired target skills. ^{Rakap and Parlak-Rakap (2011)} conducted a review of 16 studies that provided insight into the usability of NI in a variety of settings (e.g., Head Start, community preschools, ECSE classrooms) and for targeting diverse child outcomes (e.g., cognitive, communication, motor). Collectively, these studies demonstrate the extensive research that has been conducted over the past 20 years on NI and the potential of this method in enhancing outcomes for young children with special needs. Even so, Rakap and Parlak-Rakap's review suggests that more research is necessary to better understand teachers' use of NI and the types of activities during which it may be ideal for teachers to implement the practice.

EMBEDDED LEARNING OPPORTUNITIES

A core component of NI is embedded learning opportunities (ELOs). Embedded learning opportunities are child-focused, instructional episodes implemented during typically occurring routines (^{Horn & Banerjee, 2009 ; Johnson, Rahn, & Bricker, 2015 ; Sandall & Schwartz, 2008}). When implementing ELOs, teachers use logically occurring antecedents (i.e., events that initiate the learning opportunity and trigger the desired behavior) and consequences (i.e., events that occur immediately after the desired behavior to reinforce the child) to develop functional and generalizable skills (^{Johnson et al., 2015}). Embedded learning opportunities are central to NI as they are child-specific and provide structure to promote children's development (^{Sandall & Schwartz, 2008}). Given the individualization of ELOs for each child, ELOs create opportunities to target goals within a child's individualized education plan (IEP). The IEP provides a road map for intervention through the selection of skills that “help a child become adaptive, competent, socially connected, and engaged and that promote learning in natural and inclusive environments” (^{DEC, 2014}, p. 11). Embedded learning opportunities support the implementation of the IEP by providing an intentional, structured, and planful way in which ECSE teachers can provide multiple opportunities for children to practice IEP skills within typical classroom routines and activities. As such, ELOs support children's acquisition of new skills. This is important, as the field has identified that children need multiple practice opportunities to attain skills (^{Johnson et al., 2015}).

Research suggests that children with special needs can achieve learning objectives when teachers use ELOs (e.g., ^{Fox & Hanline, 1993 ; Grisham-Brown, Hemmeter, Schuster, Collins, 2000 ; Horn et al., 2000}). However, research shows that teachers in early childhood classrooms use ELOs infrequently (^{Pretti-Frontczak & Bricker, 2001}). ^{Pretti-Frontczak and Bricker (2001)} coded observations of teachers with a target child for whom they had written two IEP objectives. Teachers used ELOs in 4.3%–18.1% (mean = 9.7%) of the possible 3,588 intervals. Embedded learning opportunities were most likely to consist of the teacher asking questions or providing a verbal model. Teachers were most likely to use ELOs when working one-on-one with children and least likely to use ELOs during large-group activities.

In a second study of ELOs, ^{Noh, Allen, and Squires (2009)} observed and recorded the use of ELOs by master's level preservice teachers in an inclusive early childhood program. Graduate research assistants (RAs) observed classroom activities for a total of 760 min. On average, the preservice teachers provided 7.87 ELOs per 20-min interval (range = 1.57–10.88 per min); however, Noh et al. did not assess whether this reflected significant differences across participants. Preservice teachers used ELOs most often during transitions and toileting and least often during arrival, departure, free play, and snack. Collectively, preservice and in-service teachers may not be using ELOs to address the IEP goals of children with special needs and when they do use ELOs, they may be selecting only a few antecedents or types of activities. It is unclear the extent to which differences exist in teachers' use of ELOs.

PURPOSE OF STUDY

Therefore, the purpose of this study was to better understand how ECSE teachers use ELOs to support children with special needs. First, we sought to describe how often and in what contexts ECSE teachers use ELOs to address children's IEP goals. Second, we were interested in understanding teachers' perceptions of and experiences using ELOs. Specific research questions included the following:

1. What is the proportion of 15-s intervals within 10-min videos during which ECSE teachers use ELOs?
2. During which type of activity (adult-directed, child-directed, routine) are ECSE teachers most likely to use ELOs?
3. Do ECSE teachers differ in the frequency with which they use ELOs?
4. Which types of antecedents do ECSE teachers use most frequently?
5. Which types of IEP goals are ECSE teachers most likely to address using ELOs?
6. What is the rationale for when and how ECSE teachers implement ELOs?
7. What barriers or challenges do ECSE teachers perceive in their use of ELOs?
8. What supports do ECSE teachers need to implement ELOs more frequently?

METHODS

Research design

We used quantitative and qualitative methods to examine ECSE teachers' use of ELOs in early childhood classrooms. To answer research questions 1 through 5, we used a descriptive design and incorporated numerous strategies to ensure the internal and external validity, reliability, and objectivity of our quantitative findings (see Table 1). For example, to strengthen the internal validity of the study, our ECSE teachers were blind to the objectives of the study and provided several video recordings so that we could observe their practices on multiple occasions. In addition, our data coders were naïve to the study objectives (objectivity) and double-coded 20% of sessions to ensure interobserver agreement (reliability).

Table 1

To answer research questions 6 through 8, we used qualitative methods. To increase rigor, we used multiple methodological strategies to ensure that the data were credible, transferable, dependable, and confirmable (^{Brantlinger, Jimenez, Klingner, Pugach, & Richardson, 2005 ; McWilliam, 2000 ; Patton, 2002}). These methods included utilizing a code–recode strategy (dependability) and having ECSE teachers review findings to ensure that our transcript was accurate and that it captured their perspectives of ELOs (credibility). We provide a thorough description of the methodological strategies used to increase rigor in Table 1 and within the “Interviews, Data Coding” section.

Setting and participants

Participants included eight White ECSE teachers working in three counties in one Mid-Atlantic state. Seven teachers were female and one was male. All were certified to teach ECSE. Their highest level of education completed was a bachelor's (n = 1), master's (n = 6), or doctoral (n = 1) degree. Teachers' mean years of experience working with children with special needs were 6.88 (range = 1-13) years. All taught in public preschool classrooms housed within elementary school buildings. Participants 1 through 5 and Participant 8 taught in publicly funded inclusive preschool classrooms designed to provide universal access to preschool to prepare all preschool children, including typically developing children and children with special needs, for kindergarten. Each inclusive classroom included up to 15 typically developing 4- and 5-year-old children who were enrolled in the universal preschool program and up to five children between 3 and 5 years of age with special needs. Participants 6 and 7 taught in self-contained ECSE classrooms. Self-contained classrooms did not include typically developing children; each classroom included up to five children with special needs between 3 and 5 years of age. Each classroom included one teacher and one paraprofessional. All programs were full day and operated 4 days per week.

Twenty-one children with IEPs participated in the study. Each classroom included between one and five participating children who ranged in age from 3 to 5 years at the beginning of the study. All children were identified as having a developmental delay; children had a wide range of needs from mild language and social delays to significant cognitive impairments with delays across all areas of development. Children had between two and 17 total IEP goals, up to three of which were coded for this study (see Table 2). Goals that were not observable or that included multiple skills were not coded to ensure reliable measurement. For children with more than three appropriate goals, the first author randomly selected three goals for coding to increase feasibility for coders who were watching videos for multiple children (e.g., four per classroom) who each had multiple goals (e.g., 12 total goals across the four children).

Table 2

Observational component

Procedures

Before beginning the study, the authors obtained institutional review board approval. After teachers provided informed consent and received consent from families of all children in their classrooms, the first author provided them with an iPad and a Swivl, a base for holding the iPad that moves with the teacher during video recording. The iPad included the Swivl application for video recording. Teachers video recorded themselves teaching during three activities (i.e., circle time [adult-directed], free play [child-directed], and mealtime [routine]) for at least 10 min on 3 separate days within a 1-month period. The researchers selected the three types of activities during which teachers would collect data based on an existing framework of early childhood classroom activities (^{Johnson et al., 2015}). Teachers started the video recording at the beginning of each activity. Teachers were asked to teach as usual and did not receive any training or directions to use ELOs to address IEP goals, as the goal of the study was to capture teachers' typical practice. We collected three videos of each activity for each teacher (i.e., nine total videos per teacher), with the exception of two teachers. Participant 8 provided only four videos because of personal constraints. Participant 3 provided nine total videos but unintentionally recorded an additional circle time session and one less free play activity. When all video recording was completed, the first author collected the equipment and a list of each child's IEP goals from the teacher.

Measures and materials

All three authors collaborated to develop a modified version of a 15-s partial interval sampling observation tool (i.e., PROBE; ^{Pretti-Frontczak, Capt, Leve, & Waddell, 1995}) to code teachers' use of ELOs. We modified the PROBE by eliminating a portion of the tool specific to identifying whether a skill was measurable or not measurable; this portion of the tool was unnecessary because we included only IEP goals that were observable in this study. We defined ELOs as teaching behaviors that (a) were directed toward an individual child with an IEP, (b) allowed the child to practice one of his or her IEP goals, (c) were part of an ongoing activity in the classroom, and (d) allowed the child to continue to participate in the activity in which he or she was already engaged. See Table 3 for a description of how coders determined whether or not a teaching behavior was an ELO. To supplement our coding of ELOs, we used definitions from an ecobehavioral observation tool (SCOPE; ^{Pretti-Frontczak & Capt, 1997}) used in previous research (^{Pretti-Frontczak & Bricker, 2001}) to record (a) type of activity, (b) the teacher's instructional antecedent, and (c) type of IEP goal. See Table 4 for definitions of antecedents. To determine the type of IEP goal, we used the Assessment, Evaluation, and Programming System (AEPS; ^{Bricker, 2002}), which allowed us to have a consistent method for identifying the developmental area (e.g., gross motor, cognitive) that was addressed in each IEP goal. The AEPS was selected because it is a widely used curriculum-based assessment that is part of a larger programming system focusing on ELOs within everyday routines and activities as a primary intervention strategy (^{Bricker, 2002}).

Table 3

Table 4

Coder training and reliability assessment

Four undergraduate RAs were trained to code teachers' ELOs based on the PROBE and SCOPE procedures using video clips that were not used in study results (i.e., sections of videos beyond the first 10 min) until at least 80% agreement was reached on whether or not an ELO was observed in an interval and if so, the antecedent type. When there were disagreements during training, coding procedures were refined as needed to reach consensus on codes. When 80% agreement was reached, a total of 670 min of video footage (the first 10 min of each activity) was coded using partial interval recording. Research assistants watched 15 s of a video, paused the video, and answered a series of questions leading to a decision about whether or not a teaching behavior was an ELO and if so, what type of antecedent was delivered. Research assistants defined the beginning of an ELO as the moment the teacher delivered an antecedent. For each 15-s interval, RAs recorded only the first observed ELO in the interval. Research assistants double-coded 25% of the videos to ensure interrater reliability. We calculated reliability using the total agreement method (^{Kennedy, 2005}). Reliability was 96.7% (range = 83%–100%) for occurrence of ELOs and 95.8% (range = 80%–100%) for antecedent type.

Data analysis

The first and third authors analyzed data from the video recordings using descriptive and inferential statistics to answer quantitative research questions 1–5. More specifically, we used proportions, frequency counts, means, and ranges to describe teachers' use of ELOs. To analyze data for Questions 2 and 3, we also used the Levene Statistic (testing for homogeneity of variance), χ² (Question 2) and η (Question 3; testing for statistical independence), 1-way analysis of variance (ANOVA), and the Welch Statistic (to correct for violations of homogeneity of variance).

Interviews

Procedures

After the teacher completed the video recordings, the second author invited the teacher to participate in a semistructured individual phone interview of 12 questions regarding each teacher's understanding, training, and use of ELOs to gain in-depth information related to participants' perspectives about ELOs (^{Patton, 2002}). Our semistructured interviews lasted approximately 45 min. The second author asked a scripted question and utilized probes when necessary to gain more information or clarification, and she typed the participant's responses. She then used member checking by e-mailing the typed record to the teacher and asking the teacher to verify that she had captured their responses correctly (^{Patton, 2002}). All participants engaged in this process except Participant 8, who did not complete the interview because of personal constraints. An external auditor (trained graduate assistant not involved in the study beyond reliability) then reviewed 25% of the data across themes and subthemes to determine agreement regarding analysis. Reliability was 100% for coding of participant responses to themes and subthemes.

Data coding

The second author copied all typed participant responses into an Excel spreadsheet to answer research Questions 6 through 8. The second author analyzed the data using starter codes that aligned with the interview questions to form themes (^{Miles & Huberman, 1994}). Our subthemes emerged from units of meaning that we identified from the participant responses, and we chunked these units of meaning together utilizing pattern coding (^{Miles & Huberman, 1994}).

We used multiple strategies to ensure the rigor and quality of the qualitative data analysis (^{Brantlinger et al., 2005 ; McWilliam, 2000}). To ensure the data's credibility, we used (a) member checking by sending participants the typed records of the interviews to confirm that our transcripts were both accurate and captured their perspectives regarding ELOs and (b) an external review of our research findings by an individual who was not engaged in this research (external auditor) to ensure interpretive quality. The external auditor also confirmed the dependability of the data, as did the code–recode strategy in which the authors reexamined previously coded content when new meaning units emerged and throughout the analysis process to ensure accuracy. To ensure transferability, we were particular about the selection and description of participants (ECSE teachers within inclusive classrooms) and engaged the external auditor to ensure that we adequately described the participants. Finally, we ensured confirmability by reflecting on our bias during analysis and utilizing an unbiased external auditor (see Table 1; ^{Brantlinger et al., 2005 ; Miles & Huberman, 1994 ; Patton, 2002}).

RESULTS

Teacher use of ELOs

Our first research question addressed the proportion of intervals during which teachers used ELOs that addressed IEP goals. On average, teachers used ELOs to address the targeted IEP goals in 9.0% of 15-s intervals (range = 0%–75%). Median use across teachers was 0%, with more than half (55.2%) of the observations containing no ELOs that addressed the targeted IEP goals (see Table 5).

Table 5

Our second question addressed the types of activities during which teachers were most likely to use ELOs. Teachers submitted 24 circle time videos, 22 free play videos, and 21 mealtime videos. Teachers used ELOs in 3.6% of intervals during circle time activities, 12.7% of intervals during free play activities, and 11.2% of intervals during mealtime activities (see Table 5). We checked for statistical independence using the χ² test of independence. We found no association between teachers and the type of activity; χ² (14) = 2.51, p = 1.0. We conducted a Levene test because the variances between activity types did not appear equal. The Levene Statistic was significant (p = .002), indicating that the variances were not equal between activity types. We then conducted an ANOVA and used the Welch statistic, which corrects for heterogeneity of variance among the groups. There was no significant difference (p = .065) in teachers' use of ELOs based on activity type.

Our third research question addressed whether teachers differed in the frequency with which they used ELOs for the targeted IEP goals. Teachers' mean use of ELOs ranged from 0.3% to 30% of intervals (see Table 5). We checked for statistical independence using the η value. We found an association between teachers and their frequency of ELO use (η = .698). We again conducted a Levene Statistic, which indicated that these differences were not equal between teachers (p = .000). An ANOVA with the Welch statistic indicated that teachers significantly differed in the frequency with which they used ELOs for the targeted IEP goals (p = .003).

Our fourth research question examined antecedents ECSE teachers used most frequently to address IEP goals. Teachers used directives, questions, prompts, and verbal models most frequently (in 49%–69% of intervals containing ELOs; see Table 5). Teachers used physical assistance and the provision of materials in 24%–27% of intervals. Teachers used clues, physical models, information, singing/reading/signing, and directs others in 0%–3% of intervals.

Our fifth research question examined which types of IEP goals teachers were most likely to address using ELOs. Whereas 33.9% of children's total IEP goals coded in this study addressed communication, 78.9% of these goals were addressed at least once in an ELO. Cognitive goals accounted for 37.5% of children's total IEP goals, but only 28.6% of these goals were targeted at least once during ELOs. Children's social and fine-motor goals accounted for 14.3% and 12.5% of the total IEP goals, respectively, with 37.5% and 42.9% of these goals being the focus of teachers' ELOs at least once. Only one IEP goal targeted adaptive skills and the teacher addressed it with an ELO.

Teacher perceptions and experiences using ELOs

Reported ELO use

Theme 1, reported ELO use, aligned with research Question 6, which addressed the rationale for when and how teachers implement ELOs. Theme 1 was defined as details surrounding ELOs, such as what skills were being taught, group size, how teachers used ELOs, specific teaching strategies, the setting, and teachers' description of their own use of ELOs. Reported ELO use made up 61% of the total data and seven participants had data within this theme (see Table 6 for percentage of total data and the number of participants providing data for each theme). The following subthemes made up Theme 1: types of activities, specific teaching strategies, child goals, group size, and the teacher's rationale for using ELOs.

Table 6

Types of activities

Seven participants indicated that they use ELOs during structured (e.g., circle time, art activity, game) activities or a combination of structured and unstructured (e.g., block play, play in housekeeping area, outdoor play) activities. For example, Participant 6 stated, “The best times are circle time and small group because you can really focus on that learning activity with that child and it is easier to manage that way.”

Specific strategies

Seven participants identified using ELO strategies such as in sight out of reach, prompting, reinforcement, and modeling. For example, Participant 3 said, “We frequently put preferred items out of reach but visible to increase requesting. With students who have some speech, we use modeling and prompting to elicit more than one word....” She identified specific strategies including, “Verbal and physical prompts, peer and adult modeling, verbal praise.”

Goals

Seven participants identified teaching goals related to the state's early learning standards as well as individual child needs. For example, Participant 2 said, “... When you look at the cognitive, social emotional skills those are all things that we focus on from the IEP and the early learning standards framework so those are things that we are required to work on.”

Group size

Six participants provided data related to working with children in a variety of group size settings. For example, when we asked Participant 4 about ELOs, she stated, “I would use them through all [types of routines] because I use them through all activities through the day. I do not think I do more of one than the other.”

Rationale

Four participants discussed functional learning opportunities as their rationale for using ELOs. For example, Participant 6 stated, “I really think especially in this kind of classroom every kind of activity is an opportunity for learning opportunities ... we embed it in every activity all day long.”

Challenges

Theme 2, challenges, aligned with research Question 7, which was related to difficulties experienced by teachers. Theme 2 was defined as specific barriers teachers identify that make it difficult for them to use ELOs. Seven participants had data within this theme. The following subthemes made up Theme 2: logistics and resources, and child skills.

Logistics and resources

Seven participants identified large class sizes, individualizing instruction, and access to resources as challenges they face. For example, Participant 1 stated, “I think in my classroom I have 20 kids this year and nine of them are on IEPs so finding time to work on ELOs for each kid can be stressful....” Participant 5 identified access to resources such as materials and time as challenges she experiences.

Child skills

Seven participants identified communication and social skills as difficult to address. Participants also acknowledged specific activities as challenging contexts for providing ELOs related to communication and social skills. Participant 3 said, “It is difficult to plan for communication skills ... of our four students with communication issues, two are willing and able to imitate speech acts when prompted. Two of them will not, so addressing communication skills is a challenge.” Participant 2 said, “I think the hardest ones to plan for are social skills ... most of the children with IEPs are [identified with] autism so they don't want to necessarily interact with one another ....” Participant 5 stated, “I think writing can be a challenging activity for some of the kids ... also cognitive skills such as trying to build things....”

Supports

Theme 3 aligned with research Question 8, which was related to supports needed by teachers to implement ELOs. Theme 3 was defined as resources teachers have received or would like to receive to enhance their use of ELOs. Seven participants provided data within this theme, and the two subthemes were desired supports and training.

Desired supports

Seven participants discussed a decrease in class size, more staff, training, money, and a decrease in nonteaching tasks as needed supports. Participant 6 said,

This year it is a matter of we are just feeling [more] outnumbered than we have in years past ... more adults would always help but it is unrealistic so [we use] visuals and schedules but not all children want or use them....

Participant 7 indicated that trainings specific to special education would be helpful. Participant 5 said, “I would like more money ... so if I think I need something like materials I get it.” Participant 4 stated, “The biggest help would be freeing up teachers' time....

Training

Seven participants identified both their higher education experiences and professional development as sources of training related to ELOs. For example, Participant 3 stated, “During my master's work in ECSE, almost all courses addressed ELO[s].” Participant 2 said, “... We have trainings now with PBS and how we can do those pro-social skills and integrate those things into the day....”

DISCUSSION

Our study expands on the existing literature by providing quantitative data that describe when and how teachers use ELOs and qualitative data that provide insights into teacher perceptions of ELOs, barriers to implementation, and supports needed to increase teacher use of this important mechanism for addressing the needs of young children with special needs.

Teachers' infrequent use of ELOs

The teachers in our sample used ELOs with about the same frequency as teachers in ^{Pretti-Frontczak and Bricker's (2001)} study. Teachers in the study by ^{Noh et al. (2009)} implemented ELOs more frequently than teachers in our study. Although the participants in the study by ^{Noh et al. (2009)} were preservice teachers, they had received very specific and intensive training in using ELOs and were working in a model ECSE classroom. Participants in the ^{Pretti-Frontczak and Bricker's (2001)} study had received training on writing IEP goals, wrote two IEP goals for each child, and were observed for their use of ELOs related specifically to those goals. In contrast, our data were collected with practicing teachers in school district classrooms who received no direct training or suggestions to use ELOs to address IEP goals. Thus, our study may provide a more representative and authentic sample of the frequency with which teachers use ELOs to address IEP goals within inclusive classroom settings.

Teachers' descriptions indicate that they perceived their use of ELOs to be more frequent than our quantitative data suggest. For example, Participant 3 reported “frequently [putting] preferred items out of reach but visible to increase requesting,” yet this participant used ELOs in fewer than 2% of the measured intervals. This discrepancy may be due to challenges with implementation (i.e., teachers know what ELOs are but have difficulty putting them into practice). This may also indicate that teachers need more structured ways to use the IEP as a road map for translating IEP goals into practice opportunities in the classroom. Another possible explanation is a difference in definitions. Teachers in our study appeared to define ELOs to include a broad range of skills. For example, one participant described using ELOs to address readiness and self-help skills that were in the state's early learning standards but not included in children's IEPs. Another possibility is that teachers may have addressed IEP goals that we did not code in our analysis; although this could have led to an underestimation of teachers' use of ELOs, our randomization of goals to include in the research would have made this difference purely due to chance. Finally, it may also be that teachers did not match their teaching to the child's interest or focus of attention. During coding, when teachers interrupted the child's activity, this was not coded as an ELO. In future training efforts, it may be critical to stress the key ingredients of ELOs including that the antecedent needs to focus on the IEP goal and allow the child to continue his or her current activity.

Teachers use ELOs across a variety of activities

Our findings were consistent with previous research (^{Pretti-Frontczak & Bricker, 2001}), in that teachers in our study used ELOs less often in adult-directed activities than in child-directed and routine activities; although this value was not statistically significant (p = .065), it did have practical importance in that teachers were three times more likely to use ELOs during child-directed and routine activities than adult-directed activities.

Interestingly, there were some discrepancies between what we observed and what teachers reported. Participants 2 and 6 indicated using ELOs during structured activities. Participant 1 indicated a preference toward structured activities but did suggest that she takes advantage of child-led activities as well. Participants 5 and 7 indicated using both structured and unstructured activities and wavered between which they perceived using more frequently. Participant 4 said that she uses ELOs during both structured and unstructured activities. In addition, all teachers except Participant 5 indicated using ELOs during one-on-one, small groups, and/or large group activities. Participant 3 indicated that she uses them in small groups, whereas Participants 1, 2, 6, and 7 indicated using ELOs across all types of activities; however, these participants also indicated preferences toward small group interactions due to aspects such as child focus and the type of IEP goal. Participants 4 and 5 did not identify a preference but rather indicated that they used ELOs across all types of activities.

Teachers use verbal antecedents to address communication goals

Similar to ^{Pretti-Frontczak and Bricker's (2001)} findings in which teachers used questions and verbal models most frequently, ECSE teachers in our study were most likely to use the following antecedents: directives, verbal prompts, questions, and verbal models. In addition, all teachers in our study who completed the interview reported using various antecedents such as prompting, reinforcement, modeling, and specific environmental arrangement strategies to promote communication. Both our quantitative and qualitative data show that ECSE teachers most often provide verbal antecedents, whether through directives, questions, or verbal models. Physical assistance and physical models were used less frequently and the remaining antecedents coded were used infrequently.

This study provides new information about teachers' use of ELOs and IEP goals. Over half of the ELOs teachers delivered addressed communication goals. Although communication goals were addressed with the greatest frequency in our quantitative analysis, when interviewed, some teachers discussed providing ELOs for communication goals as challenging. Teachers also indicated that social goals were challenging to address (e.g., for children with autism spectrum disorder); teachers addressed social goals much less frequently than communication goals in our sample. Despite our focus on IEP goals, teachers in our sample described providing ELOs targeting general skills (e.g., social skills, communication) for all children and addressing early learning standards in addition to individualized goals for children with IEPs. Importantly, this information indicates that ECSE teachers may be using ELOs to support children's learning more often than our data suggest, but that they are targeting early learning objectives more often than children's IEP goals. This finding may suggest that teachers are experiencing challenges associated with providing ELOs that align with children's IEP goals. In addition, although it is imperative to provide learning opportunities related to children's individualized goals, it is equally important for children to have access to the general education curriculum and to present these learning opportunities in an age- and developmentally appropriate manner, such as through ELOs for preschool children. This information presents a new direction for future research to describe the content targeted through ELOs and to what extent it provides access to the general education curriculum.

Supporting teachers to overcome barriers to implementing ELOs

Although all teachers used a range of 0–16 ELOs throughout all videos, during the interview, teachers in our study reported several barriers to using ELOs similar to those described in the study by ^{Horn et al. (2000)}, such as feeling overwhelmed by implementing ELOs within the context of balancing their other various responsibilities in the classroom. Teachers in our study reported aspects of program infrastructure that they believed would be helpful in increasing their use of ELOs (e.g., smaller class sizes, access to additional materials). Interestingly, these characteristics have been found to be unrelated to children's outcomes; rather, it is the quality of teacher–child interactions that have been positively associated with preschool children's school readiness (^{Mashburn et al., 2008}); however, providing teachers with resources that strengthen their use of ELOs and their interactions with children may be beneficial.

Implications for teacher training and classroom practice

Our results from the teacher interview suggest several implications for preservice teacher preparation and in-service professional development. Although some teachers in our study had many years of experience and reported knowing what ELOs were and how to use them, they still used them rarely in the classroom. To increase teachers' use of ELOs focused on children's IEP goals, teachers may need additional training to help them transfer skills learned in preservice programs to their work in the field. Additional research is needed to determine what types of support teachers may need to implement ELOs beyond their preservice preparation program. For example, there is a growing body of literature suggesting that coaching is a promising method for bringing evidence-based practices into the early childhood classroom (^{Coogle, Ottley, Rahn, & Storie, 2018 ; Coogle, Ottley, Storie, Rahn, & Burt, 2018 ; Snyder et al., 2018}). Providing coaching to in-service teachers on ELOs while they are in the act of teaching in their classrooms may help increase transfer of this skill from higher education classrooms to teaching practice.

Limitations and future research directions

There are several limitations to consider when interpreting the results of this study. First, our sample included only eight teachers from one geographic region of the United States, making it difficult to generalize findings. It is unknown whether our results would be replicated in a study with more teachers and with teachers from different parts of the United States or other countries. In addition to aiding with generalizing findings, coding data for more teachers might have allowed us to perform more nuanced analyses (e.g., looking at how ELO use varies by teacher and child characteristics). Second, we coded data for only three child IEP goals. Thus, it is possible that teachers provided ELOs on skills we did not measure or on the general education curriculum, which would have increased teachers' mean use of ELOs. In addition, the types of antecedents teachers used may have been impacted by children's IEP goals (e.g., more communication goals may have led teachers to use more verbal antecedents). Third, although consequences are an important part of the ELO to ensure learning, we did not code the teachers' use of consequences, as this was not part of the original PROBE measure and we were most interested in focusing on how teachers created opportunities for children to practice skills. Fourth, we coded only the first 10 min of each video. Teachers may have demonstrated increased use of ELOs in the middle or end of the routine. Fifth, we did not collect data throughout all parts of the classroom day, precluding conclusions about teacher use of ELOs during routines outside of meals, circle time, and free play. Sixth, only seven of the eight teachers participated in the interview, limiting our qualitative data to seven participating teachers.

In future studies, researchers might consider gathering data from a larger sample of teachers in several geographic regions of the United States to increase the generalizability of findings. In addition, researchers should code all observable IEP goals and consider coding early learning standards rather than limiting coding to three IEP goals. As consequences are a critical portion of the ELO for children's learning, researchers should consider including an analysis of consequences in addition to antecedents. Other valuable information might be related to the number of opportunities that are missed in providing ELOs. Researchers might consider exploring this topic. Researchers should also collect data during all routines, including transitions, and on more days of instruction to strengthen the conclusions that can be drawn about teachers' use of ELOs within each type of activity. Finally, although our study focused on teacher use of ELOs, future research should also examine child outcomes in relation to teacher use of ELOs (e.g., variations in child outcomes during particular activities or in relation to the types of antecedents implemented; ^{Snyder, Rakap, et al., 2015}).

CONCLUSION

Naturalistic instruction is a well-accepted and evidence-based method for ECSE teachers to provide opportunities for young children with special needs to practice critical skills to enhance their participation and independence in preschool classroom routines and activities (^{Snyder, Rakap, et al., 2015}). Within this framework, ELOs are the primary vehicle through which teachers plan and deliver intentional learning opportunities while allowing children to continue engaging in meaningful, ongoing activities (^{Horn & Banerjee, 2009}). For many young children, particularly those with significant disabilities, numerous opportunities to respond will be necessary to assist them in acquiring new skills (^{Johnson et al., 2015}). Ensuring that young children with special needs receive an adequate number of opportunities to practice targeted IEP goals will require more carefully planned preservice teacher training and ongoing professional development with particular attention to supports that will help teachers use ELOs with adequate frequency to meet the needs of young children with special needs.

REFERENCES