2017 ASCA Grants Project: The School Counselor Ratio and Student Success

Added July 30, 2018

Executive Summary

With funding from ASCA, we sought to understand more about the relationship between school counselor ratios and student outcomes. We utilized data provided by the Department of Public Instruction in North Carolina and partnered with the Educational Policy Initiative at Carolina (EPIC - https://publicpolicy.unc.edu/epic-home/) at the University of North Carolina at Chapel Hill to conduct our analyses. Overall, we find that the quantity and quality of school counseling resources in elementary and
middle schools in North Carolina matters for student outcomes. The relationship between student outcomes and quantity of school counseling resources reflected by school counselor to student ratios is small in magnitude, but has a positive association for certain students and particular outcomes. Similarly, the relationship between student outcomes and the quality of school counseling resources reflected by RAMP designation also demonstrate some small and specific positive associations. 

Overall, the data on school counselor ratio suggest:

• School counselor ratio calculations need precise measurement. Data from the National Center of Educational Statistics rely on self-report data from states. It is unclear how
  states calculate ratios or if it the data is consistent from state to state. We present a more precise way to calculate ratio using both licensure and payroll data. These calculations account for part-time or shared school counselors, as well as licensed school counselors in varied roles or non-counselors acting in a school counselor category. The most recent data (2014-15) presented by ASCA notes a student to school counselor ratio of 1:378 for North Carolina. Our data for 2015-16 for North Carolina suggest 4.01 counselors per 1,000 students which equates to 1:249 ratio. We believe there is further need to examine the measurement and precision in school counselor ratio. Finally, we also believe there is value in presenting ratio data by school level noting how different the elementary school ratio (1:370) is from the high school (1:166) ratio in our data. The overall school counselor ratio masks the limited resources directed to prevention or proactive service, at the same time does not contextualize the extensive individual college and career readiness work done by high school counselors. 
• The most recent school counselor ratio data from ASCA includes trends over time (2004-2015). We presented our data (2008-2016) in that way in our interim report in October 2017. Again, these data differ based on calculations. For example, in contrast to the relatively flat pattern over time, our data demonstrate nuanced patters of increased school counselor ratio for elementary school counselors (more counselors per student). Generally, the ratio for elementary school counselor has steadily increased during the study period from 2.47 full time counselors per 1,000 students in 2008 to 2.70 full time counselors per 1,000 students in 2016. This indicates that North Carolina and its school districts are gradually investing more resources in elementary school counseling services. The ratio for middle school counselors show some declines and then a rise, but in all still lag behind pre-recession totals. High school patterns express up and down trends with again a reduction of overall school counselor resources. 
• An important question about school counselor ratio is the distribution of school counselor resources related to school and student characteristics. Philosophically, one could argue that resources that support student growth and learning should be allocated to those students who need it most. These data confirmed that North Carolina and its school districts concentrate financial resources for counselors into schools with more disadvantaged students (as measured by schools meeting the ASCA recommended body
  count ratio of 1:250). To more rigorously test the distribution of counseling services, we estimated linear regression models with our full time counseling ratio as the outcome variable and the percentage of minority and economically disadvantaged students as the focal predictor variables. Results confirm the focused distribution of counseling resources. For example, a 10 percentage point increase in the percentage of economically disadvantaged students in elementary schools predicts an additional 0.125 full time counselors per 1,000 students. In middle schools, a 10 percentage point increase in the percentage of minority students predicts an additional
  0.221 full time counselors per 1,000 students.
• The relationship of school counselor ratio to student outcomes show positive, but mixed results. We do find that increased quantity of school counselor resources (a lower school counselor to student ratio) has a statistically significant relationship for certain outcomes for specific school levels and/or populations. We examined these relationships in both school and student effects models. School effect models have been the prominent analytic strategy in previous research, but student effect models allow researchers to follow student outcomes over time accounting for variation across student mobility including the elementary and middle school transition (investigating the extent to which year-to-year changes in the school-level counseling resources to which students are exposed associate with year-to-year changes in student achievement levels). Significant relationships include:
  • End of Grade Achievement Test Scores in School Fixed Effect Models: School fixed effect models suggest a statistically significant positive relationship between end of grade achievement test scores in mathematics and school counselor ratios overall. Specifically, these relationships are positive for White students in elementary and middle school, Black students in middle school, and students who are not low SES (not on free or reduced lunch). Note, while these are statistically significant, they are miniscule and explain or are associated with less than 1% of the variance in student test scores. Student fixed effect models demonstrate a similar pattern of a positive relationship for student test scores in mathematics which includes Hispanic and low SES students. While we find a negative association between school counseling resources and mathematics achievement growth for elementary school students; we find that students who move to middle schools with relatively rich counseling resources experience a smaller achievement decline between 5th and 6th grade their peers who transition to middle schools with fewer counselor resources per student. While this effect is modest in size and does not appear to hold for reading achievement, it suggests that the resources that middle school counselors
    provide can help students make a more successful transition to challenging and typically highly tracked middle school mathematics curricula. In both models, the relationship between school counselor resources and end of grade test scores in reading are not significant (or demonstrate a slight negative relationship for White students).
  • Course Grades: Course grades are often considered a measure of school engagement rather than achievement. Only non, low-SES students in middle school
    demonstrated a positive significant relationship between school counselor resources and course grades in the school fixed effect model, whereas data
    demonstrate a significant and negative relationship between school counselor resources to course grades in the transition to middle school. Again, there's the exception with mathematics with positive, although slight in magnitude, relationships for Black and low SES students. It appears that counselor resources generally do not buffer
    the traditional grade (engagement) decline across school levels in course grades. Further inquiry is needed to understand this finding. 
  • School Absences: Overall, there is a negative relationship between school counselor resources and student absences (fewer student absences with increased
    school counselor resources) with few exceptions in school fixed effect models. Ratio explained about ¾ of a % of the variation in absences in elementary schools and nearly 1.5% of the variation in absences in middle schools. Student fixed effect models replicate similar results for elementary school, but instead demonstrate a positive relationship between absences and school counselor resources as students the transition to middle school. Similar to the engagement data with course grades, school counselor resources do not appear potent enough to buffer the typical decline in school engagement (in this case absences) in the transition to
    middle school. 

We were not able to examine behavioral outcomes with these data. Schools do not report behavioral data other than suspension, which does not have enough variation to conduct
reasonable or accurate analyses. Suspension data, and most other related behavioral data is not consistent (how behavior referrals are applied or recorded). In addition, the limited variance and highly skewed data on suspensions for elementary and middle school render analyses with little utility. We were also not able to examine the school counselor ratio relationship in high schools with these data. High schools report end of course achievement test data in North Carolina, as compared to end of grade data in elementary and middle schools. The variance in student course taking creates a high level of data alignment that was not possible with the grant resources allocated. However, we were able to extend our research with the dataset on the quality of school counseling resources by focusing on one large school district in North Carolina. This district has
a history of RAMP schools and allowed for some comparison within district. Overall the data on school counselor quality suggest:

• Demographic data highlight mixed patterns but demonstrate RAMP elementary schools are larger, have fewer minority and low SES students, more experienced and certified
  teachers, higher school counselor resources (better ratios) and higher achievement. RAMP middle schools are also larger with fewer minority and low SES students, but
  fewer school counselor resources and lower achievement then non-RAMP middle schools. 
• The relationships between RAMP and student outcomes demonstrate mixed results. While there was no statistically significant relationship between student
  achievement (as measured by end of grade achievement exams) and school counselor quality for elementary and middle school overall, a few positive relationships emerged for Hispanic students in middle school mathematics and White and non low-SES students in elementary reading (approached statistical significance). A statistically significant, negative relationship overall was found between RAMP and student absences in middle schools. Increased quality of school counselor resources was related to fewer student absences. Further, subgroup analyses demonstrated mixed but some statistically significant results (e.g., nearly 29% of the variance explained by RAMP for days absent for Hispanic students in middle school and 11% of the variance for low SES students in elementary school). 

In all, school counselor quantity (ratio) and quality (RAMP) matter, albeit the results are mixed and magnitude or practical effect is very small on broad measures of student outcomes. The quantitative analyses depend on the precision of measurement (e.g., school counselor ratio calculations) and chosen methodology (e.g., these are correlational relationships, not experimental/causal data). We believe these data provide an important start to deeper analyses into the impact of the quantity and quality of school counseling resources in elementary and middle schools (school levels with less research as compared to high school). In particular, some of these data reinforce the challenges of helping students when they make the transition from elementary to middle school apparent in most educational research.

More importantly, predictor variables such as ratio and RAMP are blunt data sources for quantity and quality of school counselor intervention. Similarly, test scores and absences may not be the best metrics for understanding the impact of school counseling services. These data do not account for a school counselor actually working individually or in group with a particular student, nor the duration or quality of that individual or group interaction. Unlike teacher or classroom research, direct counseling service remains highlight variable within a school (e.g. some of the students in the sample may have never or rarely interacted with a school counselor).

Valid conclusions about the actual impact of school counseling services would require randomly assigned experimental design that accounts for the actual quantity (how frequent and for how long a student sees a school counselor), modality (individual, group, classroom, programming) and quality (appropriate theory or technique) of the dosage of counseling per student. And even then, the appropriate outcome may not be a test score in mathematics, rather a reduction in anxiety, higher quality relationships or even higher subjective well-being. Certainly conclusions about the impact of school counseling practice require complex research design and careful interpretation.

 

Source:

Patrick Akos, Ph.D., Thurston Domina, Ph.D., Kevin Bastian, Ph.D., University of North Carolina at Chapel Hill