Exploring Teachers’ Intentions to Use Behavioral Classroom Interventions

Abstract

Multi-tiered systems of behavioral supports offer teachers tools to implement positive, antecedent- or consequence-based interventions for all students (i.e., Tier 1), and for those who need additional support (i.e., Tier 2), such as students with ADHD. Because these interventions may be challenging to use, targeted, theory-driven implementation strategies may assist teachers in implementing them with fidelity. This exploratory study examined teachers’ intended and self-reported use of specific Tier 1 and Tier 2 behavioral classroom interventions. Sixty-five K-8 teachers from five urban public schools completed an online survey about their intentions to use and self-reported use of four Tier 1 and Tier 2 behavioral classroom interventions. Teachers’ intentions varied by intervention, with the weakest intentions for using a daily behavior report (Tier 2), and weaker intentions for using high rates of specific praise than for other Tier 1 interventions. Teacher’s self-reported use was significantly lower than intended use for Tier 1 interventions, but not Tier 2 interventions. Results were generally similar whether the referent group was students with ADHD symptoms or the entire class. These results suggest specific factors to target to support teachers’ use of behavioral classroom interventions.

Schools are a common setting for the implementation of prevention practices and early interventions for children, including those with attention difficulties or disruptive behavior (Duong et al., 2021). K-12 teachers can deliver evidence-based prevention and early intervention in classrooms where children spend much of their time (Atkins et al., 2017). A growing number of schools have implemented Multi-Tiered Systems of Supports (MTSS) and Positive Behavior Interventions and Supports (PBIS; Kittelman et al., 2019). When using these frameworks, educators can apply practices with all students (i.e., universal or Tier 1 practices), as well as with students who need additional support (i.e., targeted or Tier 2 interventions). These evidence-based practices are grounded in behavioral theory and emphasize both positive, antecedent- and consequence-based approaches to manage student behavior in the classroom (Simonson & Sugai, 2019). These practices are of particular importance for students with or at risk for attention-deficit/hyperactivity disorder (ADHD) or other disruptive behavior disorders, as they are designed to maximize their academic and behavioral success in general education.

Despite the proliferation of MTSS and PBIS in schools, teachers use Tier 1 and Tier 2 behavioral classroom interventions inconsistently or not as designed (Owens et al., 2020). The burgeoning field of implementation science provides methods for developing and testing implementation strategies – “methods or techniques used to enhance the adoption, implementation or sustainability of a clinical program or practice” (Proctor et al., 2013). Teacher coaching and consultation are commonly used as implementation strategies to support teachers’ use of interventions to improve children’s social and behavioral outcomes, including Tier 1 and Tier 2 behavioral classroom interventions, and have generally been found to be effective (see Stormont et al., 2015 for a review). Teacher coaching and consultation models usually include performance feedback (e.g., Stormont et al., 2015), but vary considerably in the elements they include (e.g., modeling, goal setting, role-play). Because teacher coaching and consultation models require considerable resources (e.g., time from an expert coach; Stormont et al., 2015), there is a need for other forms of implementation strategies to support teacher use of Tier 1 and Tier 2 behavioral classroom interventions. Implementation strategies are most likely to be effective when they are theory driven and target appropriate mechanisms, meaning malleable factors that influence provider behavior (Lewis et al., 2018). Identifying malleable factors that influence teachers’ use of behavioral classroom intervention is a first step towards developing strategies to increase their use.

The Theory of Planned Behavior (TPB; Ajzen, 1991), a psychological theory of behavior prediction, is useful for understanding teachers’ use of evidence-based practices and informing mechanism-based strategies to improve implementation. The TBP posits that the strength of individuals’ intentions to perform a behavior (e.g., use a behavioral intervention) are a strong predictor of their behavior, when they can act on their intentions. The TPB is increasingly applied to understand providers’ implementation of evidence-based practices in many contexts (e.g., Fishman et al., 2018; Wolk et al., 2019), but little work to date applies the TPB to teachers’ use of behavioral interventions in the classroom (although see Sanetti et al., 2014). Doing so may have value for informing implementation strategy development for several reasons.

To apply TPB requires understanding the strength of teachers’ intentions to use behavioral interventions, and the extent to which they vary by specific intervention. Behavioral classroom interventions encompass a wide range of antecedent and contingency-based management practices. These discrete practices vary along several dimensions, including whether they are designed to be applied to the entire class (i.e., Tier 1), or to targeted students only (i.e., Tier 2), and the time they require to implement. Recent research regarding other psychosocial evidence-based practices (e.g., cognitive behavioral therapy for anxiety or depression, evidence-based classroom intervention for children with autism) suggests that provider intentions also vary meaningfully by practice component (Fishman et al., 2018; Wolk et al., 2019). Examining the extent to which teacher intentions vary by specific intervention may inform the development of implementation strategies tailored to specific teacher practices.

Additionally, it is important to understand the extent to which there is a “gap” between teacher intentions to use and self-reported use of behavioral classroom interventions. If this is the case (i.e., when teacher intentions are strong but teachers do not use the practice as intended), it suggests that implementation barriers interfere with ‘in the moment’ execution. When a discrepancy between intentions and behavior exists, rather than using implementation strategies focused on teacher motivation, it may be more effective to use strategies that support “in the moment” execution (e.g., reminders or building habits; Fishman et al., 2018). Because teachers’ self-report may overestimate their use of an intervention; an observed gap between intended and self-reported use of an intervention would be compelling evidence that something is interfering with “in the moment” execution.

Teachers’ intentions and self-reported use of behavioral classroom interventions also may vary for different groups of students. In particular, teachers’ use of Tier 1 and Tier 2 behavioral interventions in the classroom are especially important for students with or at risk for ADHD or other disruptive behavior disorders (e.g., Downs et al., 2019) but evidence suggests that teachers may use behavioral interventions such as praise less often with these students than with the class as a whole (McClowry et al., 2013). It is therefore important to understand how teachers’ intentions and self-reported use of behavioral classroom interventions vary by referent group.

Teachers’ perceptions of classroom behavioral interventions and their ability to use interventions as intended may vary based on characteristics of their school as well. Frameworks such as the one proposed by Domitrovich et al (2008) highlight factors at the macro level (e.g., policies and financing), and school level (e.g., resources, school climate) that may influence implementation of school-based interventions. The current study intentionally examines teacher perceptions in a large, urban, under-resourced school district serving a predominantly minoritized student population. Identifying implementation strategies to support behavioral classroom intervention in this context is particularly important, given that teacher-delivered interventions in the classroom offer promise for mitigating disparities in educational practice and outcomes for marginalized children (Atkins et al., 2017).

We conducted this exploratory study as a first step to applying the TPB to develop more precisely targeted implementation strategies. We gathered data on the intentions of K-8 teachers to use four key behavioral classroom interventions, each of which has an extensive research base supporting its effectiveness with children in the general population (Collier-Meek et al., 2019) as well as those with elevated ADHD symptoms (DuPaul & Stoner, 2015). The interventions included: specific praise (i.e., providing frequent verbal acknowledgment by specifically labeling praise-worthy behavior; e.g., Cook et al., 2017); precorrections (i.e., reminding students about behavioral norms prior to a time when behaviors of concern might be likely; e.g., De Pry & Sugai, 2002); brief and specific behavior corrections (i.e., consistently correcting behavior in a clear, concise and calm way; e.g., Owens et al., 2020); and use of daily behavior reports to provide feedback on specific behavioral goals (e.g., Holdaway et al., 2020). Note that daily behavior reports are typically applied at the Tier 2 level (i.e., for students with elevated levels of need, including ADHD symptoms), whereas the other three interventions are implemented class-wide (i.e., Tier 1) or specifically with students who have elevated behavioral need.

For each intervention, we gathered data about teachers’ intentions to use them for all students, as well as for students with symptoms of ADHD, because teachers may find it more challenging to use these practices with students with ADHD. We also collected teachers’ self-report of their use of each intervention, using parallel questions, to assess the gap between their intended use and self-reported use.

We examined two research questions: (1) To what extent do teachers’ intentions to use behavioral classroom interventions vary by specific intervention?; (2) How do intentions compare to self-reported use of the practices (i.e., to what extent is there a gap between intentions and self-reported behavior). For both questions, we examined whether results varied when thinking about students who have ADHD symptoms compared with the class as a whole.

Method

Setting

The study took place in five schools from a large urban school district in the Northeast United States. The district’s student body is racially and ethnically diverse (about 50% Black/African American, 21% Hispanic/Latino, 14% Non-Hispanic White, 7% Asian, and 5% Multiracial or Other races). Approximately 80% of these students live in households that are income-eligible for free or reduced-price meals.

Participants

A total of 128 teachers from five schools were invited to participate in the survey. The analytic sample consists of 65 teachers (51% response rate; 61 teachers completed all questions and 4 completed at least 50% the survey). An additional 12 teachers initiated the survey but did not provide enough data to be analyzed; data from these teachers were discarded. The sex, race/ethnicity, and years teaching experience of teachers who initiated but completed less than 50% of the survey did not differ significantly from the analytic sample.

Table 1 provides demographic characteristics of the sample. Teachers in the sample were similar to teachers in the school district as a whole in terms of race/ethnicity, although the sample had a greater percentage of females (86%) than the school district as a whole (74%).

Table 1.

Participant Demographics of Analytic Sample

Variables	Teachers (N=65), M (SD)/%
Years teaching experience (M, SD)	12.29 (8.53)
Years at current school (M, SD)	4.85 (5.32)
Sex:
Female	86%
Male	14%
Prefer Not to Say	--
Race:
White	70.8%
African-American or Black	26.2%
Asian	4.6%
American Indian/Alaska Native	1.5%
Native Hawaiian or Pacific Islander	1.5%
Prefer Not to Say	1.5 %
Ethnicity: Latinx/Hispanic/Spanish	3.1%
Grades Taught:
K-5	75.4%
6–8	24.6%
Education:
Doctorate	1.6%
Master’s	70.8%
Bachelor’s	27.2%
PhD	1.5%

Procedures

All procedures were approved by the school district research board and the hospital research board. K-8 teachers from five schools in a large urban school district in the Northeast were invited to participate in this study. Schools were recruited based on principal’s interest in the study, with an effort to select schools with student body demographics that are representative of the district. According to principal report, all five participating schools were using school-wide Positive Behavior Interventions and Supports (PBIS) programming at the time of the study.

After obtaining principal permission to conduct the research, we invited K-8 teachers in the five schools to participate in the confidential online survey via email. The survey included questions about teachers’ demographic and professional background characteristics and questions about their intended use and self-reported use of three Tier 1 one Tier 2 intervention (see Measures). To ensure that participants had a shared understanding of the four interventions, recorded presentations (each approximately 1–2 minutes long) describing each intervention were embedded within the survey; participants viewed each module prior to answering questions about the intervention. Because all participating teachers were in schools implementing school-wide PBIS at the time of the study, they may have received other school-wide trainings on the local PBIS framework regarding these practices, or similar practices; the goal of the present study was to understand teachers’ perceptions of the interventions within this context. Participants who completed the survey received a $15 electronic gift card to compensate them for their time.

At the time of the survey administration (i.e., between October, 2020 and January, 2021) all schools in the district were operating in a fully virtual model due to the COVID-19 pandemic. Because teachers’ intentions to use and experience with behavioral classroom interventions may differ between virtual and face-to-face settings, and because the focus of the broader study was to understand teacher behavior in more typical settings, throughout the survey teachers were instructed to think about their experiences teaching during the previous school year, prior to the COVID-19 pandemic.

Measures

Teacher demographic information

Teachers provided socio-demographic information including sex, race, ethnicity, grade levels taught, and years teaching experience.

Intentions and Use

Teachers were asked to rate their intended use of each of the three Tier 1 practices and Tier 2 intervention (i.e., “provide specific praise for appropriate behavior at least 4 times as often as I corrected behavior”; “use precorrections”; “use brief and specific error corrections when students broke classroom rules”; “use a written card, such as a daily report card, Check In Check Out card, or daily behavior report, to provide students feedback on specific behavioral goals”) for children exhibiting symptoms of ADHD (i.e., “think about the two students in your class who showed the most inattentive, impulsive, or hyperactive behavior”), as well as for students from the class as a whole (i.e., “for all students in my classroom”). The items used validated stems (Fishbein & Ajzen, 2010) to probe teachers’ intentions to use each intervention component (e.g., “I intended to provide specific praise for appropriate behavior at least 4 times as often as I corrected behavior”). Each item was rated on a 7-point scale (1 = Strongly disagree and 7 = Strongly agree), with higher numbers representing stronger intentions. Similar question stems were used to probe teachers’ self-reported use of the intervention (e.g., “I actually provided specific praise for appropriate behavior at least 4 times as often as I corrected behavior”) on the same scale.

Analytic Approach

To compare teacher’s intentions by intervention component, we compared intentions scores for the three Tier 1 practices and one Tier 2 intervention and for each of the two referent groups (i.e., students with ADHD symptoms, class as a whole) using a repeated measures analysis of variance (ANOVA). We used post-hoc pairwise comparisons to probe significant differences. To compare teachers’ intentions and self-reported behavior, we used paired sample t-tests to compare intended use and self-reported use of each intervention component. We examined these comparisons for both sets of referent student groups, and also used paired sample t-tests to compare across the referent groups. Given the exploratory nature of the study, we report p-values that are uncorrected for multiple comparisons. Effect sizes were computed using Cohen’s d.

Results

Intentions by Practice

Means and standard deviations of teachers’ intended use of the four interventions for whole class and for students with ADHD symptoms are displayed in Table 2. For the whole class, teachers’ average intended use of the three Tier 1 practices fell between 5 and 6.5 (on a 7-point scale). There was a significant difference in the strength of intention across the four interventions, F(3, 183) = 38.74, p < .001, partial eta squared = .39, with the strongest intentions for brief and specific behavior corrections and precorrections, and weakest intentions for using daily behavior reports. No other comparisons were statistically significant.

Table 2.

Teachers’ Intended and Actual Use of Behavioral Classroom Management Intervention Components (N = 65)

	Intended use M (SD)		Actual use M (SD)		p for comparison between intended and actual
	All students	Students with ADHD symptoms	All students	Students with ADHD symptoms	All students	Students with ADHD symptoms
Tier 1 Practices
Specific praise	6.00 (1.09)	5.57 (1.58)	5.55 (1.02)	4.98 (1.63)	p = .001	p < .001
Precorrections	6.19 (.88)	6.03 (1.12)	6.00 (1.02)	5.77 (1.10)	p = .03	p = .02
Brief and specific behavior corrections	6.24 (.78)	6.14 (1.03)	5.79 (.89)	5.72 (1.13)	p < .001	p < .001
Tier 2 Intervention
Daily behavior reports	4.31 (2.09)	4.61 (1.84)	4.15 (2.15)	4.50 (2.00)	p = .28	p = .32

Notes. Intended and actual use variables were measured on a scale from 1 (low) to 7 (high).

^aSelf-reported actual use of specific praise for the two referent groups differed significantly from each other (p = .007)

For students with ADHD symptoms, the mean score for teaches intentions to use a daily behavior report was 4.6, which falls in between “neither agree nor disagree” and “somewhat agree.” The mean score for using specific praise at least four times as often as behavior correction was 5.6, falling in between “somewhat agree” and “agree.” The mean scores for precorrections and brief and specific behavior corrections were 6.03 and 6.1, respectively, just above “agree.”

Results of the repeated-measures ANOVA showed a statistically significant difference in the strength of intentions across the four interventions, F(3, 189) = 20.65, p < .001, partial eta squared = .25 for students with ADHD symptoms. Post-hoc pairwise comparisons revealed that intentions to use daily behavior reports were significantly weaker than intentions to use the other three practices (all p’s < .01, all Cohen’s d’s > .56). Intentions to use high rates of specific praise were weaker than for precorrections (p = .018, Cohen’s d = .34), and brief and specific behavior corrections (p = .003, Cohen’s d = .43), which did not differ significantly from each other.

Intended Use Compared to Self-Reported Use

Teachers’ self-reported use of each intervention was compared to their intentions to use the same intervention (Table 2). Teacher’s self-reported use was significantly lower than intended use when the referent group was students with ADHD symptoms for specific praise (p < .001, Cohen’s d = .37), and brief and specific behavior corrections (p < .001, Cohen’s d = .39), and precorrections (p = .02, Cohen’s d = .23). Self-reported use and intended use of daily behavior reports did not differ significantly from each other. Results were similar when the referent group was all students in the class.

Teachers’ self-reported reported use of praise was significantly weaker when the referent was students with ADHD symptoms than for all students (t = −2.78, p = .007, Cohen’s d = .42). There were no other significant differences when the referent was all students versus those with ADHD symptoms.

Discussion

The goal of this study was to explore teachers’ intentions to implement three Tier 1 and one Tier 2 behavioral classroom interventions. We estimated the extent to which teachers’ intentions varied by specific intervention and differed from their self-reported use; we also examined how these patterns varied when the referent group was students with ADHD symptoms compared with all students in the classroom. Although exploratory, this is an important first step to developing theory-driven, tailored implementation strategies to improve the use of these interventions.

The results suggest that teachers’ intentions to use different behavioral classroom interventions vary depending on the specific intervention. Regarding both referent groups of students, teachers reported the strongest intentions to use precorrections and brief and specific behavior corrections. Their intentions to use high rates of specific praise were somewhat weaker, and intentions to use a daily behavior report were weakest. These results are consistent with other studies showing variation in provider intentions by specific intervention; the range of intentions reported were also similar to ranges reported using the same intentions scale regarding other evidence-based psychosocial intervention components (e.g., Fishman et al., 2018, Wolk et al., 2019). The fact that intentions were weakest for using a daily behavior report, the only Tier 2 intervention examined, is consistent with the hypothesis that Tier 2 interventions are more difficult and time intensive than Tier 1 interventions. Teachers showed a similar pattern of intentions across interventions when considering using the interventions for the whole class. This suggests that the interventions that teachers are most motivated to use with students with ADHD are the same as those they intend to use with their entire class.

We also compared teachers’ reports of their intended use of each of the behavioral classroom intervention components to their self-reported use. These results show that, even by their own report, which may overestimate use, teachers use praise, precorrections and brief and specific behavior corrections less than they intended, suggesting the importance of implementation strategies to support teachers in acting on strong intentions. This gap did not exist for using daily behavior reports, for which overall intentions were weaker, suggesting that implementation strategies aiming to promote intentions (rather than facilitating acting on intentions) may be more important for this practice.

In addition to reporting about their use of the behavioral classroom management interventions for students with ADHD symptoms, teachers also reported about their use of the interventions for all students in their classrooms. On average, teachers reported more use of specific praise for their class as a whole than for students with ADHD symptoms. These results stand in contrast to recommendations that students with ADHD symptoms or other forms of disruptive behavior should actually receive more frequent praise than their peers in order for them to receive a high ratio of praise to correction statements (Sutherland et al., 2000). As such, teacher coaching and other implementation strategies should provide specific support for teachers in understanding the importance of praising students with ADHD symptoms and finding opportunities to do so.

These results have several implications for research developing targeted implementation strategies, as well as for practitioners who support teachers in urban schools. Because teachers had relatively weak intentions to use daily behavior reports and did not report a gap between intentions and behavior for this practice, it may be most effective for teacher coaching or other implementation strategies to target teacher buy-in for this practice, by addressing attitudes, norms and self-efficacy, which the TPB posits are potential determinants of intentions (Ajzen, 1991). For example, implementation strategies may target teachers’ self-efficacy (i.e., confidence in their ability to use daily behavior reports) through goal setting, scaffolding, or skills practice. They also might target attitudes by messaging about the effectiveness of daily behavior reports, or norms by testimonials from other teachers about their use of this intervention. Although the clinical significance of changes in teachers’ reported intentions to use behavioral interventions is not known, the current results and the Theory of Planned Behavior nevertheless suggest potential targets for implementation strategies.

Teachers also reported weaker intentions to use specific praise compared to other Tier 1 practices. At the same time, teachers’ self-reported use of praise was lower than their intentions would suggest. This suggests that implementation strategies to support teachers in using high rates of specific praise, a core component of many behavioral intervention programs (Royer et al., 2019) should be designed to both strengthen teachers’ intentions to use praise (e.g., by promotion more positive beliefs about praise) and to support teachers in acting on strong intentions (e.g., through reminders).

Finally, teachers’ intentions to use precorrections and behavior corrections were relatively strong, but teachers reported less use of these practices than their intentions would suggest. This suggests that teacher coaching for these intervention components should emphasize supporting teachers in acting on their strong intentions; for example, strategies such as reminders may be particularly effective. It also may be helpful for coaches or consultants to help teachers identify what gets in the way of their acting on those intentions.

These results should be considered in the context of several limitations. First, data were collected during the COVID-19 pandemic (i.e., between October, 2020 and January, 2021), when teachers were engaged in fully remote instruction; teachers therefore were instructed to report retrospectively on their experiences during their most recent face-to-face instruction. Teachers may have varied in the accuracy of their recalls, and it is possible that recall was poorer for teachers who participated later in the school year, although it is likely not challenging for teachers to imagine and report about face-to-face instruction, given that their prior teaching experience had been in that format. Furthermore, there is no reason to believe this would have a differential impact on their report regarding the different intervention components. Second, all data included in the current study were from teacher self-report; there was no observational component. Self-report data are subject to a number of biases, and teacher self-reported use of the intervention component may not reflect what an observer in the classroom would have reported. Importantly, though, the goal of including a measure of self-reported use was not to describe the level of intervention use that occurred, but rather to compare whether teachers’ self-reported intentions and use were consistent with each other. Additionally, the sample size was limited to teachers from five schools, and was also limited to those who chose to respond to a voluntary survey. Results may not generalize to the whole district, or other urban school districts. It would be valuable for future work to measure these constructs among larger and more representative samples. The grade level taught by participating teachers varied considerably (i.e., K-8), which may have important implications for teachers’ experiences with behavioral classroom interventions (e.g., perceived developmental appropriateness, number of students taught throughout the school day); it will be important for future work to examine these constructs in more narrow grade ranges.

Despite these limitations, these results provide clear evidence that teachers’ intentions to use behavioral classroom interventions vary by specific intervention, with teachers showing overall stronger intentions to use Tier 1 interventions, as compared to the Tier 2 intervention. Results also showed a “gap” between teachers’ intended use and self-reported use of the three Tier 1 interventions, but not the Tier 2 intervention. This has direct implications for the types of implementation strategies that may be most effective for each type of intervention: specifically, these results suggest that implementation strategies should primarily aim to strengthen teacher intentions for the Tier 2 intervention (e.g., through building self-efficacy), whereas they should support teachers in acting on their intentions to use Tier 1 interventions (e.g., through reminders). This work provides a first step toward the goal of developing implementation strategies targeted more precisely to specific components of behavioral classroom interventions.