Title page
Contents
Why this study? 2
Research questions 4
Findings 8
Limitations 15
Implications 15
References 17
Appendix A. About the study 19
Appendix B. Methods 22
Appendix C. Supporting analysis 58
Table 1. Correlations between predictors and students' outcomes measured two years out 12
Figure 1. Student self-reports of SEL competencies and school experiences peaked in elementary school, declined in middle school and early high school, and increased at the end of... 9
Figure 2. School-level teacher and student reports of students' SEL competencies and school experiences declined from elementary school to high school 9
Figure 3. Differences across schools in year-to-year changes in SEL competencies and school experiences were similar to those for academic measures 10
Figure 4. Adding SEL competency and school experience predictors improved classification accuracy by at most 5.2 percentage points relative to models that included demographic and... 13
Figure 5. Of the pairs of respondent reports, teacher and parent reports were the least correlated across schools 14
Figure 6. Average reports on perseverance and rigorous expectations differed across respondent types mainly because teachers responded differently than students and parents 15
Boxes
Box 1. Key terms 3
Box 2. Data sources, sample, and methods 6
Box 3. Interpreting classification accuracy 13
Table A.1. How this study addresses each Strategic Plan goal 20
Table B.1. Number of items in DCPS's customized Panorama Education survey 22
Table B.2. Sample sizes by respondent type and school year 23
Table B.3. Student characteristics and outcomes used in the study 24
Table B.4. Sample sizes by research question and analysis sample for the main analyses 26
Table B.5. Response rates (student survey) 29
Table B.6. Response rates (teacher survey) 29
Table B.7. Exploration of nonresponse bias for perseverance (student survey) 30
Table B.8. Exploration of nonresponse bias for self-management scale (student survey) 31
Table B.9. Exploration of nonresponse bias for self-efficacy (student survey) 32
Table B.10. Exploration of nonresponse bias for social awareness (student survey) 33
Table B.11. Exploration of nonresponse bias for rigorous expectations (student survey) 34
Table B.12. Exploration of nonresponse bias for student satisfaction (student survey) 35
Table B.13. Exploration of nonresponse bias for sense of belonging (student survey) 36
Table B.14. Exploration of nonresponse bias for changes in perseverance (student survey) 37
Table B.15. Exploration of nonresponse bias for changes in self-management (student survey) 38
Table B.16. Exploration of nonresponse bias for changes in self-efficacy (student survey) 39
Table B.17. Exploration of nonresponse bias for changes in social awareness (student survey) 40
Table B.18. Exploration of nonresponse bias for changes in rigorous expectations (student survey) 41
Table B.19. Exploration of nonresponse bias for changes in student satisfaction (student survey) 42
Table B.20. Exploration of nonresponse bias for changes in sense of belonging (student survey) 43
Table B.21. Exploration of nonresponse bias for perseverance (teacher survey) 44
Table B.22. Exploration of nonresponse bias for rigorous expectations (teacher survey) 46
Table B.23. Difference between survey sample and original sample for the composite of covariates with and without nonresponse weights 48
Table B.24. Correlations between cognitive tests and other future academic outcomes in existing literature 55
Table C.1. Average level of perseverance by subgroup and grade 59
Table C.2. Average level of rigorous expectations by subgroup and grade 62
Table C.3. Average level of self-efficacy by subgroup and grade 65
Table C.4. Average level of self-management by subgroup and grade 68
Table C.5. Average level of sense of belonging by subgroup and grade 71
Table C.6. Average level of social awareness by subgroup and grade 74
Table C.7. Average level of student satisfaction by subgroup and grade 77
Table C.8. Year-to-year correlations of SEL competencies and school experiences 81
Table C.9. Average year-to-year changes in measures for average and high positive-change schools 83
Table C.10. Pairwise correlations between grade-3 SEL competencies and school experiences and grade-3 ELA achievement 85
Table C.11. Pairwise correlations between grade-8 SEL competencies and school experiences in 2017/18 and grade-9 outcomes in 2018/19 85
Table C.12. Results from analyses of the accuracy of classifying high school graduation on credits behind in grade 9 86
Table C.13. Correlations between predictors and students' outcomes measured one year out 87
Table C.14. Correlations between school-level parent and teacher reports of perseverance and rigorous expectations and school-level outcomes measured one year out 88
Table C.15. Classification accuracy of predictors 89
Table C.16. Difference between respondent types overall and by school characteristics for perseverance 92
Table C.17. Difference between respondent types overall and by school characteristics for rigorous expectations 93
Table C.18. Correlation between school characteristics and average differences in responses for perseverance 94
Table C.19. Correlation between school characteristics and average differences in responses for rigorous expectations 95
Figure A.1. DCPS's Loved, Challenged, and Prepared Index reporting 20
Figure C.1. Students self-reported SEL competencies and school experiences exhibited similar patterns by age as by grade 58
Figure C.2. Students' SEL competencies and school experiences were correlated between years but less so than most academic measures 80
Figure C.3. The ICCs for year-to-year changes in measures suggest similar conclusions as the percentile differences presented in the main text 84
Figure C.4. The ICCs for most academic measures exceeded those of SEL competencies and school experiences 84
Figure C.5. A ROC curve illustrates how a model's classification threshold governs the tradeoff between accurately classifying students with positive outcomes and misclassifying those... 91
Figure C.6. Accounting for student and teacher nonresponse made little difference in the estimated correlations between respondent types 95
Figure C.7. Accounting for student and teacher nonresponse made little difference in the estimated average differences between respondent types 96