1. Greg Lobdell, President & Director Research greg@effectiveness.org In Collaboration with the State Board of Education: Ben Rarick Andrew Parr, and Linda Drake

2. Understanding Former-ELL Work began in fall, 2013 as AI development was underway Led to conversations with SBE and OSPI staff about what I was seeing in the data Current Project: Research report / article highlighting what we see and policy implications from deeply understanding the Former-ELL subgroup

3. Collaboration State Board of Education – Ben Rarick – Andrew Parr – Linda Drake CEE: Greg Lobdell

4. Objectives of the Work Journey is just beginning… What can we learn about the Former-ELL subgroup – Size and characteristics What is the performance based on the 3 indicators used in the Achievement Index (AI) What factors matter? Size, poverty, diversity, language, grade, grade-configuration, etc. Positive Outliers- What do we see in those schools that are significant positive outliers? – Do we have other “data” we can look at for these schools? – Upward feeder pattern analysis Policy implications

5. What can the AI data tell us? Building is the unit of analysis Continuously enrolled students only Performance Indicators: – Proficiency: reading, math, writing, science percent of students meeting standard – Median Student Growth Percentiles: reading and math – Graduation Rate: 5-year Adjusted Cohort method

6. Sizing the Groups 3-Year Average Size Grade 3Grade 4Grade 5Grade 6Grade 7Grade 8Grade 10 Former-ELL Subgroup 5,497 7,069 7,559 7,832 7,730 7,646 6,472 ELL Subgroup 8,578 6,947 6,030 5,026 3,978 3,477 2,875 ALL 61,116 61,481 62,170 62,642 62,654 62,298 59,693 Source: OSPI MSP and HSPE Reading raw data for Achievement Index. Continuously Enrolled Students only. Testing years 2011, 2012, and 2013.

7. Your Turn: What do you see? Snapshot of Performance Indicators – Grade by grade proficiency in all content areas for All-students, ELL, and Former-ELL – Student Growth: mean 3-year Median Growth Percentiles – Graduation Rate: 3 years of 5-year Adjusted Cohort graduation rates What Observations do you have? What other questions do you have that would help you understand this more deeply?

8. Are schools with reportable Former ELL populations different from schools without Former ELL populations with respect to student demographics? If so, is the pattern of differences consistent across grade levels? Group Differences – Q1

9. School Differences t-Test Results School Measure Former ELL Group Mean Standard Deviation Standard Error of the Mean t-Test Result PCT_MIGRANT 0.693.399.091 p < 0.001 1 3.457.175.256 PCT_ELL 0 3.686.736.1813 p < 0.001 1 16.7115.665.5591 PCT_SWD 0 15.8116.409.44170 p < 0.001 1 12.264.035.14400 PCT_FRL 0 44.4223.036.62011 p < 0.001 1 55.6525.718.91790 PCT_504 0 2.073.012.0813 p = 0.113 1 2.252.055.0733 PCT_FOSTER 0.21.715.019 p < 0.001 1.14.273.010 TOTAL_N 0 363.19313.8038.447 p < 0.001 1 683.30292.04013.993

10. Group Differences – Q2 Is the academic performance of the All Students group different at Former ELL schools as compared to Non-Former ELL schools? If so, is the pattern of differences consistent across grade levels?

11. School Academics t-Test Results School Measure Former ELL Group NMean Standard Deviation Standard Error of the Mean t-Test Result Reading_3-Yr_Pct_Met 0 104272.6313.611.422 p < 0.001 1 77768.6413.981.502 Math_3-Yr_Percent_Met 0 103463.1115.577.484 p = 0.001 1 77660.7215.456.559 RandM_3YR_AVG_PRO 0 103467.8914.257.443 p < 0.001 1 77664.7014.408.517 Reading_3Yr_MSGP 0 99749.557.402.234 p = 0.231 1 77349.966.997.252 Math_3Yr_MSGP 0 99550.169.370.297 p = 0.291 1 77250.649.433.340 RandM_3YR_AVG_MGP 0 99249.857.470.237 p = 0.214 1 77250.297.302.263

12. Group Differences – Q3 For schools with reportable Former ELL populations, how do the academic measures for the Former ELL students compare to the academic measures for the All Students group? How do the measures vary by content area and by school level?

13. Group Differences Descriptive Statistics School MeasureNMinimumMaximumMean Std. Deviation FELL_R_PRO_3YR_AVG 45536.298.074.511.70 FELL_M_PRO_3YR_AVG 45017.197.965.515.45 FELL_RandM_PRO_3YR_AVG 45027.497.069.913.11 FELL_R_MGP_3YR_AVG 50528.376.852.28.48 FELL_M_MGP_3YR_AVG 50426.782.053.910.62 FELL_RandM_GRO_3YR_AVG 50330.677.053.18.419 School MeasureNMinimumMaximumMeanStd. Deviation Reading_3Yr_Pct_Met 7778.099.468.613.98 Math_3Yr_Percent_Met 7765.299.760.715.46 RandM_3YR_AVG_PRO 7766.699.564.714.41 Reading_3Yr_MSGP 77329.270.750.07.00 Math_3Yr_MSGP 77219.577.750.69.43 RandM_3YR_AVG_MGP77227.871.850.37.30

14. Group Differences – Q4 For each of the academic performance indicators and school level, which schools have the greatest demonstrable success with their respective Former ELL students?

15. Highest Performing Schools School MeasureESMSHS N90 th 95 th N90 th 95 th N90 th 95 th FELL_R_PRO_3YR_AVG 23191.293.915979.484.05590.593.7 FELL_M_PRO_3YR_AVG 23188.193.315973.682.55081.788.5 FELL_RandM_PRO_3YR_AVG 23189.093.015976.383.65086.690.2 FELL_R_MGP_3YR_AVG 22964.165.920062.064.96459.360.9 FELL_M_MGP_3YR_AVG 23069.473.620067.870.86266.170.4 FELL_RandM_GRO_3YR_AVG 22965.067.620062.865.66260.663.6 FELL_GRAD_3YR_AVG 1591.1

16. Positive Outliers- Further Investigation is Necessary to Understand K-5 Elementary – 520 students – 88% Free-reduced – 40% ELL – 11% Asian/Pac. Islander – 5% Black/African American – 51% Hispanic – 25% White

17. Guided Conversation… Would it be more appropriate to compare Former-ELL performance to a “never-ELL group or the all-students group? What additional analyses would be useful? – How should this study be expanded or refined?

18. If you think of things later… greg@effectiveness.org