Teacher Perceptions of Skills, Knowledge, and Resources Needed to Promote Social and Emotional Learning in Rural Classrooms

The incorporation of social and emotional learning (SEL) in schools has been shown to improve academic and psychological health of students. Research has been limited regarding implementation of SEL programs in rural communities, where student needs are heightened. The current study examined factors that could impact teachers’ intentions to be early adopters of a SEL curriculum in a rural community. Seventy-six teachers provided self-report data regarding perceptions of professional strengths, school climate, school resources for student support, ability to educate diverse students, ability to teach specific SEL domains, and intentions to be an early adopter of a SEL program. Present results indicated positive perceptions of school climate, one’s ability to teach diverse students, and one’s ability to teach self-management skills positively predicted intentions to be an early adopter of a SEL curriculum. Implications for rural schools are explored and recommendations for adoption of SEL curricula in rural schools are provided.

Involvement in SEL curricula is associated with increased academic outcomes, prosocial behaviors, and psychological well-being as well as decreased behavioral issues, mental health problems, and dropout rates (CASEL, 2019;Durlak et al., 2011;Taylor et al. 2017). Even though SEL enhances student success and classroom behaviors while reducing conduct problems, it is not systematically offered in schools. Moreover, according to Durlak and colleagues (2011), research in rural schools regarding SEL is limited.

Rural Schools and SEL
The National Center for Education Statistics reported approximately one third of public school districts in the United States are in rural areas and about one fifth of United States' students attend rural schools (Provasnik et al., 2007). Rural school districts are situated in communities with low population density and tend to be geographically isolated with limited community resources. Youth in these rural schools face unique challenges. The United States' National Advisory Committee on Rural Health and Human Services (2018) reported that youth in rural areas can be disproportionately impacted by adverse childhood experiences (i.e., family violence, abuse, neglect, family members with substance abuse, and living with a family member with mental illness). Furthermore, students in rural schools experience higher rates of poverty compared with students in urban and suburban schools (Schaefer et al., 2016). Compounding all of these issues is the lack of access to quality mental health care that youth in rural areas face (Radunovich & Wines, 2012). Rural teachers are in a difficult position when helping these youth to be successful in the classroom, due to the many family and environmental stressors they face outside of school. As such, it is not surprising that students in rural schools, particularly those in low-income households, struggle more academically when compared to their suburban peers (Graham & Provost, 2012).
SEL has demonstrated effectiveness in addressing students' mental health and academic achievement, however, research surrounding SEL effectiveness in rural schools is limited (Center for Public Education, 2018;Reynolds, 2017). This is unfortunate because results from SEL research conducted in non-rural schools are promising. For example, SEL curricula enriches students' positive perceptions of self and others, prosocial behavior, academic performance, and sense of connectedness to school while also reducing conduct issues and stress (Payton et al., 2008). Furthermore, improved social and emotional competence among students can increase the probability of high school graduation, readiness for postsecondary education, career success, positive family relationships, and mental health as well as reduce criminal behavior (Hawkins et al., 2008;Jones et al., 2015). Students from lowincome families in rural schools often enter school struggling with social and emotional skills (Meyers et al., 2015). As such, SEL curricula could be especially beneficial for these students in rural education environments.

Teachers, Rural Schools, and SEL
Training surrounding SEL curricula is necessary for teachers to effectively promote SEL programs in rural classrooms. Unfortunately, teacher preparation programs often fail to train teachers on SEL curricula and teaching strategies (Schonert-Reichl et al., 2017). Therefore, it is more likely that when rural schools adopt SEL programs, teachers will have to learn how to teach the curricula on the job. This prospect is uniquely challenging for rural teachers who are often new to the teaching profession, frequently earn low salaries, struggle with access to quality professional development opportunities, and are less likely to hold an advanced degree (i.e., master's degree; Johnson & Howley, 2015;Player, 2015).
Considering characteristics that could influence rural teachers' likelihood of implementing SEL curricula in the classroom would be a helpful step in understanding how to encourage rural schools' successful adoption of SEL programs. Researchers have broadly explored characteristics which influence teachers' implementation of SEL. For example, Ransford and colleagues (2009) suggested that experiences of burnout and lack of support negatively impacted teachers SEL implementation; whereas teachers' efficacy and positive perceptions of administrative support and training positively impacted SEL implementation. Such research however, has not been conducted in rural school environments. Exploring factors that encourage rural teachers' SEL implementation, and specifically factors which may encourage teachers to be early adopters of SEL curricula is needed.

The Present Research
The purpose of the present study was to investigate the extent to which teachers in rural schools believed that they had the relevant skills, knowledge, and resources to provide SEL to students, and how this might predict their intentions to be an early adopter of a SEL curriculum. We hypothesized that teachers who reported having more skills, knowledge, and resources would also report that they intended to be early adopters of a SEL curriculum.

Method
The current study was conducted by an interdisciplinary group of faculty and students from a regional university in the southern United States in the areas of education, psychology, and counseling. Participants were teachers from rural public schools in the region near the university. The teachers attended an informational meeting at the end of the school year in order to prepare to implement the Jesse Lewis Choose Love SEL curriculum (JLCL, 2019), the following academic school year. This research was approved by the university's Institutional Review Board and teachers completed informed consent documents prior to participating in an online survey through Qualtrics.
This study was part of a larger research project in which a SEL curriculum was introduced into several rural schools in East Texas. This program, the Jesse Lewis Choose Love program, was designed by educators as a preventative and restorative way to meet the mental health needs of children and youth (JLCL, 2019). The curriculum focuses on four core values: courage, gratitude, forgiveness, and compassion in action and is aligned with CASEL standards in order to foster the development of SEL. Lessons in this program are specifically created for each grade level covering Pre-Kindergarten through 12 th grade and includes between four and six lessons per unit. Lessons are designed to be taught one time per week and take between 15 -20 minutes, depending on the lesson. For grades 6 -12, lessons are meant to be shorter and taught throughout the week. More information about this program can be found at https://chooselovemovement.org/chooselove-enrichment-program-at-a-glance/.

Participant Demographics
A total of 76 teachers completed the survey. Participants were predominantly female n = 59 (77.6%) and their race/ethnicity were distributed as follows: Black or African American n = 5 (6.6%), American Indian or Alaska Native n = 1 (1.3%), Hispanic or Latino n = 1 (1.3%), White n = 68 (89.5 %), more than two race/ethnicities n = 1 (1.3%). Additionally, participants identified a range of years of teaching experience, though most teachers reported more than twenty years of experience n = 22 (28.9 %). For further descriptive information, see Table 1.

Measures
The Panorama Teacher and Staff survey (Panorama Education, 2015) was used to measure teachers' self-perception of relevant skills, knowledge, and resources to implement SEL with students. Four scales were used. The first was a Teacher Perception of Professional Strengths scale, which contained eight items rated on a 5-point Likert scale (1 = Not at all; 5 = Extremely). This scale included questions such as, "How confident are you that you can engage students who typically are not motivated?" The second scale was a Teacher Perception of School Climate scale, which contained nine items rated on a 5-point Likert scale (1 = Not at all; 5 = Extremely). This scale included questions such as, "How supportive are students in their interactions with each other?" The third scale was a School Resources for Student Support scale, which contained three items rated on a 4-point Likert scale (1 = Not at all; 4 = Extremely). This scale included questions such as, "When students need help from an adult, how often do they have to wait to get that help?" The final scale was an Educating All Students or a Diversity scale, which contained nine items rated on a 5-point Likert scale (1 = Not at all; 5 = Extremely). This scale included questions such as, "How easy do you find interacting with students at your school who are from a different cultural background than your own?" In order to assess teachers' self-reported confidence in teaching SEL skills to students, one of the study authors (ES) wrote a Teacher Reflection Scale to assess each of the five SEL objectives that mirrored the CASEL criteria. Questionnaire items were rated on 5-point Likert scales (1 = Not at all; 5 = Extremely). This scale contained five items to assess Self-Awareness (e.g., "I am confident that I can teach my students to recognize strengths"), six items to assess Self-Management (e.g., "I am confident that I can teach my students impulse control"), four items to assess Social Awareness (e.g., "I am confident that I can teach my students perspective taking"), four items to assess Relationship Skills (e.g., "I am confident that I can teach my students communication skills"), and six items to assess Responsible Decision Making (e.g., I am confident that I can teach my students how to identify problems"). A copy of this scale is available on the Open Science Framework website (https://osf.io).
Finally, in order to assess whether teachers thought they would be an early adopter of the JLCL we asked, "I think I will be one of the first/fastest to adopt the Choose Love model out of all the teachers in my school." This item was rated on a 5-point Likert scale (1 = Not at all; 5 = Very much). Due to the complexity of using an ordinal variable as a predictor, a new variable was created. Early Adopter was dichotomized by splitting the responses by the median to represent 1 = not likely to become an early adopter of the JLCL curriculum, and 2 = likely to become an early adopter of the curriculum.

Reliability
Internal consistency reliability of the scores was assessed using Cronbach's . The reliability for the SEL subcales ranged from .87-.95: Social Awareness ( = .87), Self-Awareness ( = .94), Self-Management ( = .93), Relationship Skills ( = .91), and Decision Making ( = .95). Additionally, the reliability for the PANORAMA Teacher and Staff had a wider range of reliabilities in its scores. School Climate ( = .83), Diversity ( = .83), Professional Growth ( = .88) had moderately high reliablities; however, the reliability for the Resouces for Student Support subscale can be consired low ( = .60), but it is important to note this subscale only contains three items. The small number of items can explain the low reliability.

Results
All analyses were computed in the statistical software jamovi (v. 1.1.9). Descriptive statistics for the independent variables included in the analyses are provided in Table 2. On average, participants had high ratings in the Teacher Self-Reflection index variables; however, there was more variability in the Panorama Teacher and Staff Survey. The School Resources for Student Support index was notably low, although items on this scale were rated from 1 to 4 rather than 1 to 5 for the rest of the scales (M = 1.92, SD = 0.52).

Factors that Predict Reported Intentions of Early Adoption of JLCL
Before conducting the analyses, the data were screened for missing data, linearity, and collinearity. An exploration of scatterplots suggested the linearity assumption was satisfied. Collinearity diagnostics such as Variance Inflation Factor (VIF) and Tolerance showed no signs of multicollinearity (VIF < 10, Tolerance > 0). Binary logistic regression is an efficient and effective method to analyze the effects and unique contribution of a group of independent variables when the dependent variable is a binary outcome (Stoltzfus, 2011). Thus, a binary logistic regression was performed utilizing Intentions to be an Early Adopter as the dependent variable, and the four Panorama subscales (i.e., School Climate, Resources for Student Support, Diversity, and Professional Strength) as independent variables. The overall model test with all four predictors against a constant-only model was statistically reliable, 2 (4) = 23.20, p < .001, indicating that the predictors reliably distinguished between participants who intended to be an Early Adopter or not. Only School Climate and Diversity were significant predictors (β = -1.84, p = .008 and β = -1.37, p = .03 respectively). The variance accounted for in intentions to be an Early Adopter was moderate, with R 2 = .27 (Cox and Snell) and R 2 = .38 (Nagelkerke). School Climate and Diversity were significant predictors of intentions to be an Early Adopter, such that with each decrease in School Climate and Diversity, the odds of becoming an Early Adopter decreased by a factor of .16 and .25 respectively. Table 3 shows the correlation coefficients between the independent variables and Table 4 shows the results of this analysis. These results suggest that teachers who perceived themselves as working in a positive school climate and perceived themselves as capable of successfully engaging a diverse group of students in the classroom were more likely to intend to be early adopters of SEL curricula.
To further explore the data, multiple independent samples t-tests were conducted. The assumption of  normality was met using Shapiro-Wilk for all variables (p > .05) and the assumption of equality of variance was met as well using Levene's test (p > .05). Due to the number of t-tests computed, alpha was adjusted via a Bonferroni correction which yielded α = .05/4 = .01 and was the significance level used for all t-tests. Diversity and School Climate were significant (t = -3.39, p < .001 and t = -3.65, p < .001 respectively). Thus, participants' ratings of early adoption of the JLCL curriculum were higher overall across all the Panorama scales (see Table 5), except for School Resources for Student Support. These results suggest that when teachers consider adopting SEL curricula they will perceive themselves as capable of successfully engaging a diverse group of students in the classroom and working in a positive school climate as the higher ratings in these indices indicate.

Confidence in Teaching SEL and Intentions to be an Early Adopter of JLCL
A second logistic regression was performed using the Teacher Self-Reflection questions regarding confidence in teaching SEL, using intentions to be an Early Adopter of the JLCL as a dependent variable and the Social Awareness, Self-Awareness, Self-Management, Decision Making and Relationship Skills indices as independent variables. Correlations between the indices can be found in Table 3, and most of the correlations were r > .50. Notably, there was a high correlation between Self-Awareness and Self-Management (r = .91). Table 6 summarizes the results of this logistic regression. Due to multicollinearity, the variable Social Awareness was removed from the model. Subsequently, the overall  model test with four predictors was statistically significant, 2 (4) = 33.40, p < .001. However, only the Self-Management index was a significant predictor on its own (β = -3.46, p = .03) though Relationship Skills was marginally significant (β = -2.80, p = .052). For every unit decrease in Self-Management the odds of becoming an early adopter of the program decreased by a factor of 0.03. The variance explained by the model in intentions to be an Early Adopter variable was moderate, with R 2 = .36 (Cox and Snell) and R 2 = .50 (Nagelkerke). Indicating that the variance explained by the Teacher Self-Reflection indices in indicating early adoption of JLCL was 36% based on Cox and Snell's R 2 and 50% based on Nagelkerke's R 2 . These results suggest that teachers who perceived themselves as confident in their abilities to teach students self-management skills were more likely to identify themselves as intending to adopt SEL curricula early. Additionally, teachers who self-identified as confident in their abilities to teach students relationship skills were somewhat more likely to intend to be an early SEL curricula adopter.
Next, a series of independent samples t-tests were conducted. Due to the assumption of normality being violated, a Mann-Whitney U test, a nonparametric t-test, was conducted to compare the mean of the Self-Awareness, Self-Management, Social Awareness, and Relationship Skills scales for participants who intended to become early adopters of the JLCL curriculum and those who did not intend to become early adopters. Once again, the alpha was adjusted, and the significance level used was α = .01. The Mann-Whitney test indicated that ratings in each of the subscales of the Teacher Self-Reflection scale were significantly greater for all indices from those participants who favored the early adoption of JLCL. For example, participants who favored the early adoption of JLCL rated higher in the Self-Awareness index (M = 4.31, SD = .44) compared to those that did not favor the adoption (M = 3.79, SD = .58), Self-Management (M = 4.30, SD = .41) for those that favored the curriculum adoption compared to those that did not (M = 3.69, SD = .60), Social Awareness (M = 4.31, SD = .46 vs M = 3.85, SD = .52) compared to those that did not favor early adoption. Similarly, for Relationship Skills, the ratings were higher for those favoring the adoption of the curriculum (M = 4.42, SD = .48) compared to those that did not (M = 3.81, SD = .49). Table 7 shows the results of the Mann-Whitney tests. These results suggest that that teachers who rated themselves more confident in teaching SEL, and thus rated higher in all the components of the Teacher Self-Reflection scale (Self-Awareness, Self-Management, Social Awareness and Relationship skills), were more inclined to adopt the SEL curricula.

Discussion and Implications
Results from the current study shed light on some important factors regarding the likelihood of SEL curricula adoption in rural schools. For example, present findings suggest that teachers' perceptions < .001* Note: Items marked with an asterisk are significant at p = .05 or less, two tailed. of their own capabilities were influential regarding their attitudes about SEL and willingness to adopt the JLCL curriculum in the classroom. Teachers who perceived themselves as more capable of teaching diverse groups of students, who perceived the school as having a more positive climate, and who perceived themselves as capable of teaching student selfmanagement skills were more likely to report that they intended to be early adopters of JLCL. These results suggest that supporting knowledge and confidence in teaching self-management skills and fostering positive school climates may be helpful in rural teachers' intentions to adopt SEL programming. For example, it might be beneficial for rural school administrators to consider intentional professional development initiatives designed to bolster rural teachers' confidence in SEL-related skills. Professional development focused on helping teachers learn how to teach PreK-12 students appropriate impulse control, organizational skills, and behavior self-management could be impactful.
Additionally, findings from the current study suggest it is important to foster rural teachers' confidence related to teaching diverse groups of students. Professional development surrounding the intersection of race, ethnicity, socioeconomic status, language, and other multicultural factors with student learning experiences would likely be beneficial for teachers. Because rural teachers struggle with access to quality professional development (Johnson & Howley, 2015;Player, 2015), it may be advantageous for teachers and administrators to be creative when planning professional development in these areas. For example, book study groups which include readings and discussions surrounding privilege, power, and poverty culture could be integrated into professional development plans. Additionally, interactive professional development experiences that include activities such as service learning and role plays could help teachers gain experience with diverse groups of individuals and practice SEL skills. Beyond professional development, teacher preparation programs can play an important role in fostering these skills as well (Schonert-Reichl et al., 2015).
Another important finding from the current study was the influential role of perceived school climate on teachers' likelihood to be an early adopter of SEL curricula. This finding is not surprising; it is reasonable that teachers who perceive themselves as working in a positive environment would be motivated to include new curricula, such as SEL, in the classroom. Researchers have suggested that teachers' perceptions of a positive school climate are related to higher levels of teacher motivation and self-efficacy (Reaves & Cozzens, 2018). While it may be a somewhat obvious notion to encourage the development of positive school climates, particularly in rural schools, it can be challenging for faculty, staff, and administrators to find the time and resources in these rural areas to dedicate to positive school climate initiatives. Rural teachers, staff, and administrators are frequently burdened with stress related to meeting state standards, state testing, a lack of financial resources, and experiences of burnout (Garwood et al., 2018). Furthermore, recent events including the COVID-19 pandemic have created new stressors for rural schools. During these times, it is important for rural schools to remember the influential role that a positive school climate can play on teachers' and students' experiences at school and teachers' motivation to adopt new curricula such as SEL.

Limitations and Future Directions
The present study highlighted important factors that predict early adoption of a SEL program. However, several methodological and logistical limitations with our study should be noted. First, our approach relied upon self-reports. It is certainly possible that our participants engaged in selfpresentation in order to appear more interested in, or capable of implementing, the SEL program (Leary, 1996). This may be particularly true of our primary outcome measureintention to be an early adopter. Second, and of particular relevance to our main outcome measure, we did not measure early adoption behavior. Instead, we measured intention to become an early adopter. It is possible that our participants are mistaken about their abilities or motivation to actually adopt the SEL program. Third, our sample size was smaller than anticipated. It is possible that the sample of participants who completed the surveys differs from those who chose not to complete the survey, which could influence the generalizability of the results. In a similar vein, our sample was recruited exclusively from rural schools in East Texas which could influence the generalizability of the results. Further research with a larger sample size, from several rural regions, and using behavioral measures alongside self-report data would be a useful contribution to the literature on this topic.

Conclusion
Rural schools have a lot to gain from adopting SEL curricula (Oberle et al., 2015). Identifying teachers who are likely to be early adopters of such curricula is an important step in considering successful initiatives for SEL program implementation in rural schools. Teachers who are early adopters of SEL curricula and supporters of SEL curricula implementation can aid administrators in ensuring the success of SEL curricula in schools and across school districts. Such support can be particularly important in rural districts where administrative personnel and financial resources are often lacking. Greater attention to the concerns of rural schools and unique issues related to implementing program interventions, such as SEL, is needed (Fishman, 2015; National Association of State School boards of Education, 2014; Wang, 2014). SEL programs have the potential to positively impact the mental health of youth and academic achievement in rural communities where mental health disparities and academic achievement gaps often prevent students from reaching their full potential (Myers, 2019;Zhang, 2008).