NCES Winter Forum and 24th Annual Management Information Systems (MIS) Conference “Deep in the Heart of Data” February 21-25, 2011 Austin, TX ELLs, NCLB, and AMAOs: The WIDA Consortium’s approach to interpreting federal policy and providing guidance Rahul Joshi, M.S. Kristopher Stewart, M.A., M.P.A WIDA Consortium, University of Wisconsin-Madison H. Gary Cook, Ph.D., Research Director

No Child Left Behind Under Title III, states must define annual measurable achievement objectives for LEP students served that relate to their development and attainment of English language proficiency while meeting challenging State academic content and achievement standards as required under Title I, section 3122 of NCLB. Three specific AMAOs have been established under NCLB: AMAO 1: ELL students progressing in English language acquisition AMAO 2: ELL students exiting or reaching English language proficiency AMAO 3: ELL Adequate Yearly Progress (AYP) WIDA Consortium

Key Research Linquanti, R. & George, C. (2007). Establishing and utilizing an NCLB Title III accountability system: California’s approach and findings to date. Cook, H. G., Boals, T., Wilmes, C. & Santos, M. (2008). Issues in the development of annual measurable achievement objectives for WIDA consortium states. Established Key Criteria for Setting AMAOs WIDA Consortium

What is the WIDA Consortium? 24 states and D.C. 1.4 million students WIDA Consortium

The WIDA ELP Standards Standard 1 – Social & Instructional Language (SIL) English language learners communicate for social and instructional purposes in the school setting. Standard 2 – Language of Language Arts (LoLA) English language learners communicate information, ideas and concepts necessary for academic success in the content area of Language Arts. Standard 3 – Language of Mathematics (LoMA) English language learners communicate information, ideas and concepts necessary for academic success in the content area of Math. Standard 4 – Language of Science (LoSC) English language learners communicate information, ideas and concepts necessary for academic success in the content area of Science. Standard 5 – Language of Social Studies (LoSS) English language learners communicate information, ideas and concepts necessary for academic success in the content area of Social Studies. Within each standard, there are Model Performance Indicators (MPIs) for Listening, Speaking, Reading, and Writing for each grade-level cluster (PreK-K, 1-2, 3-5, 6-8, 9-12). WIDA Consortium 5

Four Language Domains Listening ─ process, understand, interpret, and evaluate spoken language in a variety of situations   Speaking ─ engage in oral communication in a variety of situations for a variety of purposes and audiences Reading ─ process, interpret, and evaluate written language, symbols, and text with understanding and fluency Writing ─ engage in written communication in a variety of forms for a variety of purposes and audiences Listening/Speaking/Reading/Writing are the four main language skills. Listening and Reading are comprehension skills that involve receptive language; Speaking and Writing are communication skills that require productive or expressive language. In order to give educators a comprehensive picture of ELL students’ English language proficiency – and to comply with federal law - all four domains are tested on the ACCESS for ELLs test and the W-APT screener. For more information on the specifics of how each domain is tested, please refer to the other presentations in the toolkit related to the ACCESS for ELLs test and the W-APT screener. WIDA Consortium 6

Organization of ELP Standards Snapshot of a page of ELP Standards. >four domains on the left side of the page >Example Topics >Proficiency Levels. WIDA Consortium 7

WIDA Research Areas WIDA Consortium WIDA Research Team Technical Assistance Database Management Applied Research Basic Research WIDA Consortium

Technical Assistance Projects and Policy Guidance – AMAO 1 Provide policy guidance on AMAO 1: Determine the scoring metric Determine the annual growth target Set the starting point Set the ending point Determine the annual growth rate Meet with the State stakeholders to discuss findings State stakeholders make recommendations to SEA/LEA WIDA Consortium

Technical Assistance Projects and Policy Guidance – AMAO 2 Provide policy guidance on AMAO 2: Define ELP Determine the cohort Set the starting point Set the ending point Determine annual growth rate Meet with the State stakeholders to discuss findings State stakeholders make recommendations to SEA/LEA WIDA Consortium

Sample AMAO 1 & 2 Analysis This slide is a graphical example related to AMAOs showing growth which accompanies a technical assistance report to a Consortium member. This chart represents four important aspects: grade level cluster, percentile rank, proficiency level, and scale score gain. Cluster being examined includes the third, fourth, and fifth grade. A student’s starting composite proficiency level (<2.0 - <5.0) accompanied by the percentile rank on the x axis shows the ELL student’s growth. For example, when you subtract last year’s scale score from the present year, that will give a student growth, let’s say for a scale score gain of 32. Then, based on the fourth grade student’s composite proficiency level of 3.3, you can see that student’s is at the 60th percentile based on a scale score gain of 32 and starting proficiency level of 3.3 within the 3-5 grade cluster. WIDA Consortium

Pits and Falls (and insights) Lack of capacity at SEAs in data analysis ELL policy: Need for communication between entities Influence federal ELL policy Share outcomes, successes, results Lots of data, but SEAs and LEAs know the kids best WIDA Consortium

Data Requirements for AMAO Analyses Longitudinal ACCESS test data  Students matched across years Data reasonably cleaned and/or validated Robust methodology for student matching Reliable procedure to handle cases with missing/invalid student identification fields On-demand, secure data delivery with remote availability Provide both atomic and aggregate information WIDA Consortium

Data Delivery - WIDA Data Warehouse High-performance, scalable SQL Server database Over a million individual test takers from 22 states across US (and growing…) ACCESS Test Information (test scores, restricted student identification data and demographics), Connected to selected NCES Research Data Collections Core database for WIDA Projects, Research Initiatives and Reporting Framework WIDA Consortium

WIDA Data Warehouse - Datasets This slide indicates the main data components of the WIDA Data Warehouse. The primary data source is the ACCESS for ELLs student performance data (indicated as 1) with restricted student identification information, for all the WIDA Consortium member states. Every student in this dataset has a unique student identifier and is longitudinally tracked across successive testing cycles. The data warehouse also houses selected data elements from some of the publicly available research datasets collected by NCES namely: Common Core Datasets (indicated as 2), School and Staffing Survey Datasets (indicated as 3) and National Assessment of Educational Progress Datasets (indicated as 4). The Common Core Datasets are connected with the ACCESS datasets at both the school and district level, using unique State School ID and State District ID's. In this case, nationwide uniquely assigned School CCD Identifier is stored and also used for connecting these datasets. Since School and Staffing Survey and National Assessment of Educational Progress Datasets are only available at the state level and hence connected accordingly. State level Common Core Data aggregates also form a part of the Common Core datasets. WIDA Consortium

Building Longitudinal System for AMAO It’s comprehensive and challenging, and why? Across successive test administrations: Old/Current Students Transfer to different state/ exit the ELL program New students enter the state and/or the ELL program Missing pieces in student identifier fields Matching can be only correct up to a certain confidence level Good quality data  Better understanding of student growth (and a key to happy life!) WIDA Consortium

Building Longitudinal Student Record System WIDA Consortium

Reporting Framework for WIDA Members Statewide and WIDA-wide performance dashboard reports More insightful reports soon based on ACCESS and NCES datasets State/WIDA Dashboard Key areas State Performance and State Growth in ACCESS Domains and Grade Clusters Native Language Distribution for States Comparison with NAEP Average Composite Scale Scores Largest 10 Districts based on student enrollment WIDA Consortium

State Performance by ACCESS Cluster WIDA Consortium

State Growth by ACCESS Cluster WIDA Consortium

Native Language Distribution in a State WIDA Consortium

Pits and Falls (and insights) Always question the data you have (for correctness and completeness) Good quality source data High Reliability on student matching Don’t assume quality of key student identifiers while longitudinally matching students Make the framework inherently longitudinal Data is a “double-edged sword:” Good data MAY lead to good decision making and policies Bad data CERTAINLY could lead to bad decision making on policies WIDA Consortium

Questions? WIDA Consortium www.wida.us Contact Information: Rahul: rgjoshi@wisc.edu Kris: stewart3@wisc.edu Gary: hcook@wisc.edu WIDA Consortium