Experts in student mental health increasingly call for schoolwide approaches to service delivery, sometimes referred to as integrated models (Cook et al. 2015; Domitrovich et al. 2010) or multi-tiered systems of support (Strein et al. 2003). Such models reflect an ecological perspective, which recognizes the importance of the school context and its multiple actors (e.g., teachers, mental health professionals, students, parents) as influences on students’ mental health (Cappella et al. 2008). Typically, schoolwide approaches to mental health combine preventive services that promote social–emotional competence for all students along with intervention services for students exhibiting behavioral challenges.

The current study reports on a pilot of a novel school-based mental health program based on a university–school partnership. Like other schoolwide programs that deploy on-site consultants, the School-Based Mental Health Collaboration (SBMHC) integrates prevention and treatment services, but it is the first, to our knowledge, that is guided by psychodynamic theory and practice rather than a scripted curriculum or manualized program. SBMHC follows what Levitt, Neimeyer, and Williams (2005) call a principles rather than rules approach to clinical practice. A principles approach guides decision-making according to non-specific principles, whereas a rules approach relies on specific, formalized prescriptions. Although rules may result in greater certainty than principles, principles accommodate greater complexity. In this instance, the complexity derives from the program’s customization of goals by classroom according to teachers’ priorities and skills, as well their students’ mental health competencies and unmet service needs. During a 2-year pilot of SBMHC in three high-poverty urban schools, we evaluated two key indicators of program success: teachers’ satisfaction with the program and changes in their self-efficacy in promoting student mental health.

SBMHC’s Mission and Principles

SBMHC’s central mission is to meet a school’s most pressing mental health needs while applying key psychodynamic principles to work not only with students but also with all other stakeholders, including teachers, mental health staff (if present), school leaders, and parents (Eppler-Wolff et al. 2019). The first of these principles is that individuals within the school system, from students through principals, are more likely to reflect on their behaviors and attempt changes to promote their own or others’ mental health in the context of a trusting relationship, whether that relationship is with an SBMHC clinician or with another individual in the school system. Similar to Winnicott’s (1965) description of a “holding environment” for children in which parents attune to, and strive to meet, their child’s needs, SBMHC tries to provide a “holding environment” for staff and students alike (Hyman 2012) by showing empathy, validating needs, and marshaling resources to meet those needs. This, in turn, builds emotionally secure environments in which school leaders, teachers, students, and parents may risk reflection and change.

The second principle derived from psychoanalytic theory and practice is that mentalizing—that is, attempting to understand both one’s own and others’ mental states (Fonagy and Target 1997; Steele et al. 2015)—fosters trusting relationships and joint problem-solving among school stakeholders. SBMHC consultants adopt a mentalizing stance when dealing with all stakeholders while also encouraging those stakeholders to do the same with each other. Whether the task involves helping teachers deal with challenging students or helping leaders establish schoolwide codes of behavior, we assume a posture of gentle curiosity when trying to understand the origins and functions of each individual’s behaviors. This perspective recognizes, for example, that sad, anxious, off-task, or aggressive behaviors carry different meanings across students, and even within students across different situations. Similarly, teachers’ responses to unwanted child behaviors carry different meanings across teachers based on their own values, personal history, and teaching philosophy. SBMHC wonders not only about the minds of students and teachers but also about the school leaders and mental health providers who determine schoolwide mental health policies and procedures. In trying to promote schoolwide mental health, SBMHC posits itself as an information-seeking ally rather than as an expert with an “all-knowing stance” (Hagelquist 2018).

The successful schoolwide operationalization of these principles—first, that trusting relationships enable self-reflection and change, and second, that gentle curiosity encourages an understanding of others’ perspectives—does not lend itself to highly scripted protocols and procedures, but rather to a reliance on clinical acumen on a situational basis. As a result, classroom consultants, who are graduate students being trained as clinicians, are allowed discretion in determining their priorities with each teacher. However, there are standard functions performed by all consultants, which we now review.

Classroom Consultants’ Activities

At the start of the year, each teacher is assigned a classroom consultant (CC), who spends one half-day per week in her/his classroom. Particularly at the beginning of the year, the CC spends time simply observing the classroom, making note of the teacher’s classroom management style, his/her efforts to promote social–emotional competence, and his/her approaches to student behavioral challenges. CCs also spend their time meeting with the teacher, leading or co-leading social–emotional learning lessons, intervening with students, and making in- or out-of-school referrals for students with pronounced mental health needs.

Teacher consultations

CCs meet approximately weekly with teachers to discuss overall classroom management as well as particular students displaying internalizing or externalizing behaviors. CCs exercise discretion in setting priorities with each teacher, but, in general, strive to promote teachers’ competencies in three domains: supporting social–emotional wellbeing, responding to challenging behaviors, and building mentalization capacity. However, teachers’ needs for support vary across these competencies. CCs begin by observing how each teacher handles behavioral challenges in the classroom posed by that year’s student population. Rather than criticizing teachers, or simply recommending alternative teaching strategies, CCs validate the teacher’s frustration in dealing with difficult behaviors and position themselves as allies in the search for more effective solutions.

For example, after having observed a teacher’s maladaptive response to a student’s rule-breaking behavior, the CC might offer that she, too, felt angry and frustrated by the student. She may then go on to model a stance of gentle curiosity about the cause of that student’s behavior, thereby encouraging empathy in the teacher and guiding their joint problem-solving efforts towards root causes such as stressors in the home or untreated learning difficulties. The CC may suggest some techniques that the teacher might apply to calm herself or offer a strategy to encourage alternative behaviors in the child. After experimenting with an agreed-upon plan of action, the CC and the teacher evaluate its success and, if necessary, formulate alternative plans. The teacher is thus invited to explore strategies addressing the cause of an undesirable behavior, rather than simply punishing or otherwise trying to eliminate the behavior. Under clinical supervision, CCs formulate goals for each teacher based on her or his capacity and willingness for self-reflection and behavior change. The CC keeps the teacher’s mind in her or his mind while encouraging the teacher to do the same for her or his students.

SEL Curriculum

SBMHC requires that partnering schools implement an evidence-based social–emotional learning (SEL) curriculum that promotes student competences in five core competencies identified by the Collaborative for Academic, Social and Emotional Learning (www.casel.org): relationship skills, responsible decision-making, self-awareness, self-management, and social awareness To date, all three schools have chosen Second Step (Committee for Children 2003a, 2003b, 2003c). CCs co-teach the weekly lesson with teachers and spend part of their weekly visit time discussing curriculum implementation and enhancement. Grade-specific lessons fall within four major units: Skills For Learning, Empathy, Emotion Management, and Problem Solving. There are weekly take-home sheets for parents, as well as activities teachers can use during the week to integrate material from the lessons into their daily curriculum.

Direct Intervention with Students

CCs directly intervene with students in the classroom following Hagelquist’s (2018) Security/Trauma Focused/Obtaining Skills/Resource/

Mentalization (STORM) model, which takes a mentalization approach with children with a history of trauma. This model aims to provide children with a sense of safety and security, acknowledge and educate others about the far-reaching effects of trauma, help children obtain self-regulatory skills, identify and draw on the child’s strengths, and encourage mentalization by drawing attention to the child’s and others’ mental states. The operationalization of these goals varies by classroom depending on individual students’ needs. CCs might confer a sense of security and trust by providing a consistent and supportive presence to a child perceived by the teacher as a “problem.” They might promote self-regulatory skills by re-arranging a child’s physical environment to minimize distractions, or by sharing strategies on cooling down during upsetting situations. They might encourage mentalization by labeling and validating a child’s emotions, or by helping children in conflict to consider each other’s perspectives.

Referrals

When CCs and teachers determine that a particular child needs services outside the classroom, they and the SBMHC Clinical Supervisor meet with parents for consultation. For parents who seek services, CCs provide referrals to community-based providers. Importantly, they follow-up with parents to ask if the services were received, and if so, whether they were satisfactory (and if not, make other referrals as needed). When families engage in these services, CCs serve as a liaison among the provider, school, and parents to ensure that all parties are apprised of the progress made at home and school. Although not all classrooms require this type of intervention, in those that did, the CC filled a vital gap caused by a dearth of on-site mental health personnel. The option of working with SBMHC also met the needs of families who preferred not to have their children assessed through the City’s Department of Education.

Customization of Activities

There is no prescribed distribution of CCs’ time across these activities. In some classrooms, CCs may spend more time directly intervening with students to model behavior management strategies for the teacher, or to address children’s immediate needs for comfort or support. In classrooms with more students exhibiting pronounced mental health problems, CCs may spend more time meeting with parents and arranging referrals. In classrooms with few or no students exhibiting such problems, CCs may spend more time refining implementation of the Second Step curriculum by devising activities to reinforce lessons throughout the week. Furthermore, the proportion of time spent on each of these activities often shifts throughout the school year depending on the needs of the individual teacher, the children, and classroom.

Teacher Satisfaction with School-based Mental Health Programs

Teachers’ satisfaction with a mental health program, also referred to as program acceptability or perceived utility, generally predicts the extent to which they implement it as designed (Biggs et al. 2008; Domitrovich et al. 2019; Ransford et al. 2009). This satisfaction is the product of multiple factors. First, teachers are more apt to “buy into” programs whose underlying model is compatible with their own values, practices, and norms, as well as those of the school (Flaspohler et al. 2008; Greenhalgh et al. 2004). They are more motivated to implement programs that are perceived as filling a need and having a high likelihood of success (Han and Weiss 2005). Teachers may also view new programs as more credible when existing users reflect the culture of the school and the families served (Guerra and Knox 2008). Proposed programs may further gain credibility in urban schools if they address the burdens that economic deprivation and racism place on students (Barnes 2019).

Past literature also highlights the importance of the working relationship between teacher and consultant—that is, the extent to which they agree on goals and share an affective bond (Gessnitzer and Kauffeld 2015; Wehby et al. 2012). Some studies of school-based mental health programs demonstrate high teacher satisfaction with coaches or consultants placed in their classrooms (Ratkalkar et al. 2017; Reinke et al. 2013). Yet the applicability of these findings to SBMHC was unknown because of limited comparability between program models. Existing school-based mental health programs typically use consultants to coach teachers on the use of a specific curriculum that fosters student competencies, such as empathy or perspective-taking, or minimizes undesired behaviors such as aggression or off-task behavior (Domitrovich et al. 2019; Jones et al. 2011; Wehby et al. 2012). For example, Becker et al. (2013) provided teachers with weekly coaching on implementation of the PAX Good Behavior Game (Embry et al. 2003), which rewards teams of students with tokens for desired behaviors. Other consultation programs encourage teachers to adopt specified teaching behaviors to promote student mental health (Coles et al. 2015; McCormick et al. 2015; Sutherland et al. 2015). In the BRIDGE program (Cappella et al. 2012), for example, consultants score teachers using the CLASS Emotional Support and Classroom Management quality rating scales (Pianta et al. 2008). They then coach teachers on improving their scores by enhancing the behaviors measured by the scales (e.g., showing regard for students’ perspectives).

By contrast, SBMHC does not standardize its goals or activities across classrooms, but is rather guided by an overarching approach to promoting mental health by practicing and encouraging the use of mentalizing with all stakeholders. However, the practice of mentalization may be unfamiliar to many teachers serving high-poverty and largely non-White students, many of whom have urgent, unmet mental health needs. For example, when CCs and teachers first meet, many teachers express a need for help with disciplining particular students, whereas CCs’ first priority is to examine those students’ behaviors as clues towards understanding their origin and function. They model a stance of gentle curiosity about, and empathy towards, these behaviors in their meetings with teachers. This approach may seem overly theoretical or to take too long to implement in the “real world.” It may also be identified with psychodynamic psychotherapy, which is often perceived as the province of White elites (Sanders Thompson et al. 2004). As it happens, the CCs are drawn from the population of graduate students in psychology in New York City, and to date have been more likely to be White than not; by contrast, the teachers they work with, like the students themselves, tend to be Black or Hispanic. Indeed, in one school, many of the teachers are themselves former students who have never lived anywhere outside that neighborhood. It was possible, therefore, that teachers would view CCs as outsiders whose ivory tower experience bore little relationship to their own or those of their students.

A related complication of the SBMHC model is that it relies in part on teachers’ willingness and ability to reflect on their teaching. For example, a CC may ask a teacher whether his or her response to a child’s troubling behavior had the desired impact, and if not, to consider why. In a study of the Incredible Years SEL curriculum, coaches demonstrated lower fidelity when helping teachers reflect on their personal style than in their other activities with teachers (Reinke et al. 2013). A mental health consultation program for early childhood centers found that it wasn’t until the second year of implementation that teachers shifted from problematizing children’s behaviors to showing empathy and curiosity about the meaning of those behaviors (Alkon et al. 2003). In the BRIDGE mental health program, which promoted teachers’ emotionally supportive behaviors, the challenge of engaging in self-reflection was one of the themes that emerged in focus groups with teachers (Cappella et al. 2016). It was therefore unclear how teachers would respond to CCs’ use of modeling and encouragement of self-reflection.

It was also unknown how teachers would respond to the relatively unscripted and non-directive approach to improving classroom management taken by SBMHC. In a recent survey of elementary school teachers’ preferences for classroom-based mental health programs, teachers preferred interventions that both helped them understand their students’ problems, and provided step-by-step instructions on how to solve them, over interventions that helped with just one competency or the other (Egan et al. 2019). Although the Second Step curriculum provided a common language for CCs, teachers, and students, there was no single behavioral goal for students or teachers that served as an easily identifiable lodestar for the intervention. Rather, CCs tailored their priorities and activities with each teacher over the course of the school year. Other studies have shown that school-based mental health consultants naturally customize their work with teachers based on their individual strengths and weaknesses (Becker et al. 2013; Heller et al. 2011). CCs’ responsiveness to individual teachers’ needs was thus expected to foster teachers’ perceptions of program effectiveness. However, it was also possible that in foregoing scripted instructions or universal priorities, CCs’ guidance would be deemed ineffective or difficult to implement. It was therefore important to understand teachers’ perceived impacts of the support provided by their CCs, both across the various activities they engaged in and the multiple competencies they promoted.

School-based Mental Health Programs’ Effects on Teacher Self-Efficacy

Some literature suggests that teachers’ perceptions of their own ability, or self-efficacy, in promoting students’ social–emotional competence affects the fidelity with which they implement social–emotional learning programs. For example, a study of implementation of the PATHS social–emotional curriculum for students through the use of teacher training and coaching found that teachers’ self-efficacy predicted the number of supplemental activities they conducted each week, although it did not predict the number of lessons delivered (Ransford et al. 2009). A trial of the RULER SEL curriculum, which also relied on teacher training and coaching, found that teachers who scored lower on their quality of implementation also scored lower on self-efficacy with respect to their ability to change practices when needed (Reyes et al. 2012).

Little attention has been paid to the possible reverse directionality of the relationship between self-efficacy and implementation. That is, the implementation of an SEL or mental health intervention may itself increase teachers’ confidence in their ability to not only implement that specific intervention but also to affect their students’ mental health more broadly. To the extent that an intervention supplies teachers with new strategies for enhancing classroom social–emotional competence, it may boost teachers’ perceived ability in that domain. Similarly, to the extent that an intervention gives teachers new strategies for addressing disruptive behaviors, it may boost their perceived ability in that domain. An intervention that fosters teachers’ understanding of their students’ experiences, or increases referrals to community-based providers, may boost teachers’ perceived abilities in those domains. SBMHC helps teachers promote class-wide social–emotional wellbeing, address particular students’ disruptive behaviors, increase their own understanding of students’ experiences, and make referrals to community-based providers. Therefore, it may raise teachers’ perceived abilities across these varied mental health competencies.

It may, however, be more difficult for consultants to boost teachers’ self-efficacy in some capacities than others. For example, it may be relatively easy to increase teachers’ self-efficacy surrounding implementation of SBMHC’s SEL curriculum. The Second Step curriculum consists of weekly lesson plans and prescribed materials. Teachers may gain confidence in their ability to deliver the lessons by virtue of practice along with CC feedback over the course of the year. For example, a greater number of positive procedural interactions between teachers and students in the PAX Good Behavior Game was found to improve teacher self-efficacy (Huber et al. 2016). It may be far more challenging for CCs to raise teachers’ confidence in their ability to deal with disruptive students, a major source of distress for teachers (Emmer and Stough 2001). A recent study asked urban teachers to describe their thoughts and feelings following a challenging classroom situation (Camacho et al. 2018). Teachers most commonly felt angry and unable to resolve the situation effectively. Less common responses were specific plans to resolve the situation or an acknowledgment of the factors in students’ lives that affect their behavior in the classroom.

It is not surprising that teachers in urban schools serving economically disadvantaged students, who are already overburdened, may find aggressive or off-task behavior unmanageable (Shernoff et al. 2011; Walter et al. 2006). It can take teachers years to hone their classroom management skills and acquire confidence in that arena (Emmer and Stough 2001). Teachers who are chronically stressed may find it particularly difficult to maintain self-control in the face of student non-compliance. Given that SBMHC CCs not only aid teachers with challenging behaviors but also with the universal promotion of social–emotional competence, it was unclear which aspects of teachers’ self-efficacy surrounding mental health competencies would be impacted.

The Current Study

SBMHC diverges from past school-based mental health consultation programs in that it was designed to promote psychodynamic principles and practices rather than a specific set of teacher or student behaviors. Its consultants prioritize building trusting relationships with teachers and promoting mentalization skills, and are allowed discretion in determining their priorities and activities with each teacher. These features of the program model were designed to “meet teachers where they are” by validating their perspectives and supporting the needs they perceived as most acute. CCs strive to be non-judgmental about, and responsive to, the teacher’s agenda instead of positioning themselves as advocates for a particular set of teaching strategies or champions of particular student behaviors. The customization of priorities and activities across each CC–teacher dyad was expected to yield high teacher satisfaction, as well as increased self-efficacy in promoting student mental health. Yet, the focus on mentalization, while intended to elicit emotional growth and creative problem-solving among teachers, may also have seemed overly theoretical, culturally irrelevant, or difficult to implement. Furthermore, the diversity of teacher and student priorities and activities across classrooms raised questions about which CC activities teachers would find most helpful, and which aspects of their self-efficacy would be most improved.

To answer these questions, the current study draws on data from a 2-year pilot of SBMHC at three high-poverty public elementary schools in New York City. In evaluating teacher acceptance, we attended to three different dimensions of the program model. First, we measured teachers’ satisfaction with their CC according to activity: co-teaching the SEL curriculum, helping integrate the SEL curriculum into their daily curriculum, weekly coaching, direct intervention, referring parents to mental health services, and coordinating between the teacher and students’ mental health providers. Second, we measured teachers’ reports of the quality of the relationship with their CC. Third, we measured changes over the school year in teachers’ self-efficacy across three broad competencies: promoting social–emotional skills, addressing behavior challenges, and understanding students’ behaviors. To do so, we expanded the sample to include teachers in a comparison group as well as teachers who received the intervention. The results will speak to the potential value of classroom consultation programs with a psychodynamic orientation.

Methods

Procedures

In Year 1 of the pilot, SBMHC operated at two public elementary schools in New York City. At School 1, SBMHC had been introduced the previous year in its formative stage, whereas at School 2, SBMHC was new to the school. School 1 had only one classroom in grades K and 1, both of which were assigned to the intervention group, and two classrooms for grades 2–5, one of which was assigned to the intervention group and the other to the comparison group. School 2 had two classrooms for grades K-5, and assigned one classroom in each grade to the intervention group and one to the comparison group.

In Year 2 of the pilot, SBMHC left School 1 because of turnover in school leadership. School 2 continued SBMHC implementation for a second year, but as agreed with the principal as a condition for piloting, all K-5 classrooms were assigned to the intervention group. SBMHC was introduced to School 3, which had two classrooms per grade (K-5). One classroom in each grade was assigned to the intervention group and one was assigned to the comparison group.

Pooling across both years of the pilot, there was a total of 30 intervention classrooms and 16 comparison classrooms. Lead teachers in all classrooms (n = 54; some classrooms had co-lead teachers) were recruited for study participation. Of these, all but one consented. However, seven teachers were dropped from analyses because they did not complete data collection (n = 5) or they left school midyear (n = 2). The final analytic sample comprised 47 teachers (30 in the intervention group, representing 28 classrooms, and 17 in the comparison group, representing 15 classrooms), although there was some item missingness (n’s are denoted in tables accordingly). This represented a participation rate of 83% at School 1, of 77% in Year 1 and 100% in Year 2 at School 2, and of 88% at School 3. It should be noted that five teachers at School 2 participated in both study years.

All three schools were located in neighborhoods where roughly one-quarter (23–29%) of the population falls below the poverty line (U.S. Census Bureau n.d.) The percentage of Black students ranged from 20 to 60% across schools, while the percentage of Hispanic students ranged from 35 to 56% (NYC Department of Education n.d.). The percentage of students with special needs ranged from 14 to 36%, and the percentage who were chronically absent ranged from 16 to 47%. Finally, there were low rates of academic achievement at these schools. Only 8–29% of students passed the state English test, and only 9–28% of students passed the state math test. In sum, these schools served disadvantaged and high-needs students.

CC Recruitment and Training

There were four CCs in each of the study years. CCs were selected each year from graduate students in psychology at Teachers College and other accredited universities in New York City. The SBMHC Director provided one day of training on the theoretical underpinnings of the program (e.g., attachment and mentalization theory), and another day of training with an accompanying guidebook on consultants’ roles and responsibilities (Eppler-Wolff et al. 2020). She also led a 4-week training course about the application of attachment theory to work with children in schools. All CCs received 1 h of individual and 1 h of group clinical supervision each week with the Director and Clinical Coordinator, both of whom are licensed psychologists. At the end of the year, CCs were credited with a psychology externship that is necessary for graduation and eventual licensure (https://nynjadot.apa.org).

CC Activities

Each CC was assigned to three classrooms, and spent one half-day a week in each classroom. CCs recorded their activities and impressions after each classroom visit. Based on these records, it was calculated that CCs visited each classroom 27 times a year, on average. During these visits, on average, they co-taught the SEL curriculum 20 times; they observed the classroom 25 times; and they coached the teacher 23 times per year. Coaching sessions lasted 25 min, on average. CCs discussed referrals for mental health services for specific children with teachers during 11% of their coaching sessions. CCs and teachers met with the parents of 28 students over the 2-year pilot period.

Measures

Teachers in the intervention group alone were administered a satisfaction survey in the spring of each year. Teachers in both the intervention and comparison groups were administered surveys that measured their mental health promotion self-efficacy in the fall and spring of each year. Although it would have been preferable to use validated measures of teacher satisfaction and self-efficacy, it was necessary to create original measures that captured the specific dimensions of SBMHC’s program model.

Satisfaction

Teachers completed a 10-item survey assessing satisfaction with their CC. All items were endorsed on a 5-point scale (1 = strongly disagree to 5 = strongly agree). Three items measured the extent to which teachers believed their CC had improved their competencies, namely, by helping them promote social–emotional skills in the classroom, by increasing their awareness and understanding of their students’ experiences and feelings, and by teaching them useful strategies for dealing with students’ behavioral challenges. One item measured the quality of the CC-teacher relationship (“I have a good relationship with the Classroom Consultant”). Six items measured teachers’ satisfaction with CCs’ various roles and responsibilities. Teachers indicated the extent to which CCs supported them through co-teaching the Second Step curriculum, helping them integrate Second Step into their classroom culture, their weekly coaching sessions, direct intervention with the students, referring parents to mental health services in the community, and coordinating between them and their students’ community health providers.

Self-Efficacy

In the fall and spring, teachers in the intervention and comparison groups indicated their perceived ability across a range of social–emotional teaching competencies. They were asked to endorse nine items on a 5-point scale (1 = strongly disagree to 5 = strongly agree). Three items tapped their self-efficacy with respect to the social–emotional curriculum: “I feel equipped to promote the five pillars of emotional competency: self-awareness, self-management, social awareness, relationship skills, and responsible decision-making,” “ I am able to integrate social and emotional learning concepts into my daily curriculum,” and “It is my responsibility as a teacher to promote the five pillars of emotional competency.” One item tapped teachers’ perceived ability to recognize students’ feelings: “I am aware of and can understand my students’ experiences and feelings.” Three items assessed teachers’ perceived ability to deal with students’ externalizing and internalizing behavior: “I know of effective strategies I can use to deal with students’ behavioral challenges,” “I feel equipped to intervene with a child who is having self-regulatory issues (i.e., difficulties in self-control, tantrums, etc.),” and “I feel equipped to intervene with a child who is feeling sad and anxious.” Two items tapped teachers’ perceived ability to deal with children’s parents and mental health providers: “I feel equipped to communicate with parents about their child’s behavioral challenges” and “I feel equipped to communicate with any mental health providers my students may have.”

Controls

In the fall survey, all teachers reported their number of years of teaching experience (11, on average). The number of children in each classroom as of the fall was tabulated based on class rosters. Both teacher years of experience and classroom size were included as controls in models of SBMHC’s associations with change in teacher self-efficacy from fall to spring. We did not collect data on teachers’ race/ethnicity, age, or household structure to avoid the appearance of prying into or judging teachers’ personal lives.

Analytic Plan

To assess the satisfaction of teachers in the intervention group, scores for each item are presented descriptively, combining data across study years. Sensitivity analyses revealed that there were no differences in satisfaction across the three schools, so data were analyzed in aggregate.

To assess the effects of SBMHC on teacher self-efficacy, regression models drew on data from teachers in both the intervention and comparison groups. For each indicator of self-efficacy, the spring score on that item was regressed on intervention group status, controlling for the fall score on that item as well as teacher experience and classroom size. Thus, coefficients may be interpreted as indicating the change in score from fall to spring. Mixed models were run with random intercepts by classroom. In sensitivity analyses, separate models were run by school and year of the study. Results did not differ significantly, and thus, to ease interpretability and maximize sample size, classrooms were combined across schools and years.

Results

Teacher Satisfaction

As shown in Table 1, all mean satisfaction scores ranged between 4 (agree) and 5 (strongly agree). With respect to enhancing competencies, teachers agreed most strongly that CCs helped them promote social–emotional skills in the classroom (M = 4.27, SD = 0.58). They agreed slightly less strongly that CCs taught them useful strategies for dealing with students’ behavioral challenges (M = 4.21, SD = 0.63). They agreed yet less strongly that CCs increased their awareness and understanding of their students’ experiences and feelings (M = 4.05, SD = 0.62).

Table 1 Intervention group teachers’ satisfaction with their Classroom Consultant
Full size table

Teachers strongly agreed that they had a good relationship with their CC (M = 4.67, SD = 0.55). Among the services CCs provided, teachers agreed most strongly that they were supported by co-teaching the SEL curriculum (M = 4.50, SD = 0.63). The other services’ supportiveness to teachers were, in order of endorsement, directly intervening with students (M = 4.43, SD = 0.56); referring parents to mental health services in the community (M = 4.42, SD = 0.62); helping integrate the SEL curriculum into classroom culture (M = 4.37, SD = 0.49) and providing weekly coaching sessions (M = 4.37, SD = 0.67); and coordinating with students’ mental health providers (M = 4.07, SD = 0.65).

Teacher Self-Efficacy

Regression models indicated that teachers in the intervention group significantly increased their self-efficacy in four out of the nine competencies relative to the comparison group (Table 2). Specifically, intervention group teachers’ self-efficacy improved more than comparison group teachers with respect to their perceived ability to promote the five pillars of emotional competency (b = 0.69, SD = 0.18, p < .001, ES = .81), to integrate social–emotional learning into their curriculum (b = 0.47, SD = 0.23, p < .05, ES = .67), and to communicate with mental health providers (b = 0.48, SD = 0.18, p < .01, ES = .55). Intervention group teachers also improved more than comparison group teachers on their knowledge of effective strategies for dealing with students’ behavioral challenges (b = 0.37, SD = 0.13, p < .01, ES = .51). Intervention and comparison group teachers improved similarly on their belief that it is their responsibility to promote the five pillars of emotional competency, their awareness and understanding of students’ feelings, their perceived ability to intervene with children who have self-regulatory issues or are sad or anxious, and their perceived ability to communicate with parents.

Table 2 Results from models regressing teachers’ self-efficacy on intervention group status
Full size table

Discussion

SBMHC is a novel mental health classroom consultation program for elementary schools serving high-poverty populations. Like some other programs (Jones et al. 2011; Ransford et al. 2009; Ratkalkar et al. 2017; Reinke et al. 2013; Wehby et al. 2012), SBMHC includes an SEL curriculum and provides each classroom with a consultant. Uniquely, however, consultants are not charged primarily with helping teachers implement the SEL curriculum. SBMHC’s psychodynamic orientation dictates that consultants strive to create a safe environment for each teacher, in which she or he can reflect on her own teaching strategies and risk making changes. While keeping teachers’ minds in mind, CCs encourage those teachers to keep their students’ minds in mind. Rather than being viewed as critics or advocates for a particular agenda, CCs aim to be seen as non-judgmental allies who work toward meeting the teacher’s and students’ mental health needs in each classroom. Consequently, they have discretion in setting priorities with each teacher, and in determining how to allocate their time across activities such as co-teaching the SEL curriculum, directly intervening with students, or meeting with teachers.

We had several concerns about the acceptability of this program to teachers. First, it was possible that the flexibility and customization of the program across classrooms might be viewed as a sign of incoherence or ineffectiveness. It was also possible that the emphasis on reasons for, rather than consequences of, student behavior might be viewed as culturally irrelevant. Indeed, as the university in a university-school partnership, we ran the risk of being perceived as grafting theoretical principles from the “ivory tower” onto a high-poverty school with urgent unmet mental health needs. It was therefore critical that we examine how teachers viewed the program at large and, in particular, the perceived usefulness of their assigned consultant.

The results showed that teachers in the intervention group generally agreed that they had a good relationship with their CC. Among the competencies CCs impacted, teachers rated promoting social–emotional skills in the classroom the highest, followed by knowing useful strategies for dealing with students’ behavioral challenges, and being aware of and understanding students’ experiences and feelings, in that order. Mirroring these results, the CC activity teachers found most supportive was helping them co-teach the SEL curriculum. This activity was rated higher than CCs’ coaching sessions, direct intervention with students, and parent referrals to mental health providers. The activity rated as least supportive was coordinating between teachers and students’ mental health providers in the community. It is likely that this item scored lowest because CCs did not refer parents to community mental health providers in all classrooms, and when they did, not all parents enrolled in services.

These findings suggest that of the activities falling within CCs’ purview, helping teachers implement the SEL curriculum was the one most appreciated. The results of the self-efficacy analysis suggest that that it may also have been the most successful at raising teachers’ confidence. Of the four competencies for which CCs boosted teachers’ self-efficacy, two pertained to the SEL curriculum: integrating the SEL program into their daily curriculum and promoting the pillars of emotional competency. Effect sizes were moderate (.67) and large (.81), respectively, according to Cohen (1988). There were also moderate effect sizes for the increases in teachers’ knowledge of effective strategies for dealing with behavioral challenges and their ability to communicate with mental health providers.

Notably, teachers in the intervention group did not significantly increase their reported awareness and understanding of students’ feelings. This result is consistent with the finding that while teachers generally agreed that CCs increased their awareness and understanding of students’ experiences and feelings, CCs were more effective at improving teachers’ skills in promoting SEL skills and dealing with challenging behaviors. It may be more difficult for SBMHC CCs to alter teachers’ cognitions than it is to alter their behaviors. Other mental health consultation programs have identified the difficulty of eliciting self-reflection from teachers. One study of mental health consultation in early childhood centers found that it took 2 years of program participation for teachers to begin shifting their orientation surrounding challenging student behaviors away from problematization and towards empathy and curiosity (Alkon et al. 2003). In sensitivity analyses, we examined whether teachers in their second year of exposure to SBMHC improved more than others on self-efficacy, but results did not affirm this hypothesis.

Directions for Future Study

Several questions remain about potential modifications to the program if adopted at other high needs, low-resource elementary schools. Despite teachers’ overall acceptance of their CCs, it may be wise for future configurations of SBMHC to experiment with longer lasting teacher–CC pairings. Because this was a university-based intervention in which graduate students in psychology were performing a 1-year externship, there was necessarily turnover each year. The advantages of using graduate students as CCs—low-cost labor and built-in clinical supervision—make SBMHC possible to implement without extensive funding, a necessary and sufficient condition for many high needs public schools like those in this study. On the other hand, there may be drawbacks associated with assigning each teacher a new CC every school year. Teachers must form a working relationship and define priorities with a new CC each year rather than building on the previous year’s. In cases of less-than-optimal teacher-CC matching, however, this chance to refresh may actually be an asset. In the future, a university program adopting SBMHC might randomly assign their CCs to 1- versus 2-year terms. This would allow a test for differences in teacher acceptance and self-efficacy between classrooms whose CC–teacher partnerships were in their second versus first year. Perhaps with more time in a consistent partnership, CCs may better improve teachers’ understanding and interpretation of their students’ behavior, one of the competencies we found most difficult to affect.

Another possible direction for future research would be to increase the amount of time spent by CCs in classrooms. In response to an open-ended question asking teachers how to improve SBMHC, the most common answer was to receive more of the CC’s time. However, even then, it is not immediately evident just how a greater dosage of SBMHC would affect teachers’ self-efficacy and satisfaction. In an open-ended question asking teachers how their CC was most helpful to them, respondents cited a wide range of functions fulfilled by CCs. Some cited their CC’s intervention with particularly challenging students, others cited delivery of the SEL curriculum, and still others cited the availability of a sounding board or brainstorming partner. It is clear that in many classrooms, teachers were grateful for a second set of hands, with one teacher noting, for example, that their CC “always fills in where she sees there is a need.” In filling those needs, CCs clearly elicit teachers’ gratitude, but may not influence their knowledge base or behaviors in the long term. It would be ideal for a future trial of CCs’ effects on teachers to provide a co-teacher to classrooms in the comparison group for the same length of time CCs spend in intervention classrooms. This design would distinguish the added value of CCs in particular from that of a second adult in the classroom.

On a similar note, it would be helpful to establish SBMHC at a school with more mental health services available on-site than in the schools reported here. As mentioned earlier, not all parents who were referred to community-based providers took up those services (though not all those services were for mental health). Parents commonly reported long waitlists at centers that were free or low cost. To the extent that the school can itself provide needed services, it is possible that more families will take them up, and that coordination among CCs, providers, teachers, and families will improve. It is also possible that some of CCs’ time spent with select high needs students will be redistributed to other activities.

Finally, it is important to acknowledge that the SBMHC program has not been implemented in a location without the leadership of the SBMHC founder and director. Therefore, even without the potential modifications recommended for future study, a replication of the current model would be greatly useful for the field. Also, as the program grows, it will be essential to conduct evaluative research to better understand just how SBMHC affects specific teaching practices (e.g., implementing Second Step, managing challenging student behaviors, communicating with parents and mental health providers), as well as teachers’ own feelings of mental wellbeing.