Transcription

ACCESS for ELLsInterpretive Guidefor Score ReportsGrades K–12SPRING 2021UNDERSTANDING STUDENT SCORES

ContentsACCESS for ELLs . 1Understanding Scores . 1ACCESS for ELLs Score Reports . 2Individual Student Report . 2Student Roster Report . 2Frequency Report. 2Individual Student Scores . 3Domain Scores. 3Composite Scores . 6Kindergarten Scores . 7Interpreting Student Scores . 7Understanding Student Growth . 8Group Scores . 8Student Roster Report . 8Frequency Reports . 10Proficiency Level Descriptors (Grades 1–12) . 11Proficiency Level Descriptors (Kindergarten) . 15Reading the ACCESS for ELLs Individual Student Report . 17This document helps educators understand what students’ ACCESS for ELLs scores mean andwhat to do with that information. It also introduces some of the tools available to programcoordinators and district administrators interested in reviewing and taking action on groupperformance on ACCESS for ELLs.This document presents WIDA recommendations for interpreting and using test scores.State and district policies on test score use differ from each other and may vary from therecommendations presented in this document.The Every Student Succeeds Act of 2015 requires that all students identified as Englishlanguage learners (ELLs) be assessed annually for English language proficiency. ACCESS forELLs meets federal accountability requirements and provides educators with a measure of theEnglish language proficiency growth of ELLs. 2021 Board of Regents of the University of Wisconsin System, on behalf of WIDA.Comments on this document? Email [email protected] revised 3/1/21

ACCESS for ELLsACCESS for ELLs is a suite of large-scale English language proficiency tests for K–12 students. It is onecomponent of WIDA’s comprehensive, standards-driven system that supports the teaching andlearning of English language learners (ELLs). The purpose of ACCESS for ELLs is to monitor studentprogress in English language proficiency on a yearly basis and to serve as just one of the manycriteria that educators consider as they determine whether English learners have attained an Englishlanguage proficiency level that will allow them to meaningfully participate in English languageclassroom instruction. Visit wida.wisc.edu/assess/access for detail on ACCESS for ELLs.ACCESS for ELLs is a standards-referenced test, which means that student performance is comparedto English language development standards WIDA has defined. Any student can achieve any score,and students are not ranked against each other or against the expected performance of monolingualEnglish speakers. Visit wida.wisc.edu/teach/standards for detail on WIDA standards.The WIDA English Language Development Standards Framework, 2020 Edition will be the basisof future test development. However, all tests available in the 2020–2021 school year werebased on the 2012 standards.Understanding ScoresBefore diving into your students’ score reports, take some time to familiarize yourself with theresources on the Can Do Descriptors page of the WIDA website. The Can Do Descriptors and thecorresponding WIDA Performance Definitions for Speaking and Writing and Listening and Readingcan help you understand what test scores mean in practical terms. As you examine and discuss theEnglish language proficiency profile that each Individual Student Report shows, use WIDA resourcesto help you move from scores to concrete recommendations for the services, instructional support,and future assessment needs of each student.Consider holding an in-service session for your school or district so that educators can talk throughthe WIDA English Language Development Standards Framework, review sample score reports, anddiscuss how students’ scores might inform plans for classroom instruction and support.WIDA offers a variety of professional development resources that can help educators andadministrators fully understand and make the best use of WIDA assessments. Check out the currentprofessional learning offerings and the webinars available in the Secure Portal Download Library.Don’t keep ACCESS for ELLs information to yourself! Scores can help parents or guardiansand other educators better understand a student’s abilities. Find resources for sharing scores onthe Family Engagement page of the WIDA website.Use resources like the Model Performance Indicators, included in the 2012 Amplification of the EnglishLanguage Development Standards, to identify and describe the language abilities a student alreadyhas, the skills a student can work on, and the instructional supports that might be effective as astudent develops new language abilities. Share the profile and plans you develop with your students’content teachers. Translate your plans into the student’s home language and share them with thestudent’s family during conferences, family nights, or home visits so that home can be a place ofactive language learning.1

ACCESS for ELLs Score ReportsIndividual Student ReportAudience: Students, Parents and Guardians, Teachers, School TeamsDetailed report of a single student’s performance, including proficiency level andscale scores for each language domain and four composite areas. Share withstudents to set language goals. Share with parents and guardians as part ofdiscussions around student progress and achievement. Share with the student’steachers to inform individualized classroom instruction and assessment.Translations of the Individual Student Report are available in the following languages in WIDA AMS.Albanian, Amharic, Arabic (MSA), Bengali, Bosnian, Burmese, Chamorro, Chinese (Simplified), Chinese(Traditional), Chuukese, French (European), German, Gujarati, Haitian Creole, Hawaiian, Hindi, Hmong,Ilokano, Italian, Japanese, Karen, Khmer (Cambodian), Korean, Lao, Malayalam, Mandingo, Marshallese,Nepali, Polish, Portuguese (Brazilian), Punjabi, Romanian, Russian, Samoan, Serbian, Somali, Spanish(International), Swahili, Tagalog, Telugu, Tongan, Turkish, Ukrainian, Urdu, Vietnamese, WolofTranslated reports should always accompany—not replace!—official reports in English.Student Roster ReportAudience: Teachers, Program Coordinators and Directors, AdministratorsOverview report on the performances of a group of students, includingproficiency level and scale scores for each language domain andcomposite area by school, grade, student, tier, and grade-level cluster.Share with administrators and teachers to inform classroom instruction andassessment.Frequency ReportHigh-level report for a single grade within a school, district, or state on the number andpercentage of tested students that achieved each proficiency level for each languagedomain and composite area.State Frequency ReportSchool Frequency ReportDistrict Frequency ReportAudience: ProgramCoordinators and Directors,AdministratorsAudience: Program Coordinators Audience: State and DistrictProgram Staff, Policy Makers andand Directors, Administrators,LegislatorsBoards of EducationShare with school and districtstaff to inform school-levelprogrammatic decisions.Share with district staff to informdistrict-level programmaticdecisions.Use to prepare reports forpolicymakers and legislators andto inform state- and district-levelprogrammatic decisions.View sample reports at wida.wisc.edu/assess/access/scores-reports.2

Individual Student ScoresDomain ScoresThe Individual Student Report contains detailed information about a student’s performance on eachsection of ACCESS for ELLs. It is primarily for students, parents or guardians, and teachers. It providesa snapshot of how well the student understands and can produce the language needed to access theacademic content presented in an English language classroom. The Individual Student Report shows aproficiency level and a scale score for each of the four language domains.Proficiency levels are interpretive scores. In other words, they are based on, but separate from, scalescores. The proficiency level score describes the student’s performance in terms of the six WIDAEnglish Language Proficiency Levels:Level 1EnteringLevel 2EmergingLevel 3DevelopingLevel 4ExpandingLevel 5BridgingLevel 6ReachingThe proficiency level score is a whole number followed by a decimal. The whole number reflects thestudent’s proficiency level, and the number after the decimal reflects how far the student hasprogressed within that level. For example, a student with a score of 3.7 is at proficiency level 3 and isover halfway toward achieving proficiency level 4. At the bottom of the Individual Student Report,each proficiency level the student achieved is explained in terms of what the student can do usingEnglish. A complete list of the proficiency level descriptors is included in this document.Take care when comparing proficiency level scores across grades. A second grader with a 4.0 inListening and a 3.0 in Speaking is demonstrating more developed listening skills than speaking skills.However, proficiency levels are relevant to the context of a particular grade level. A second graderwith a 4.0 in Listening and an eighth grader with a 4.0 in Listening are exposed to very different,grade-level appropriate content as they test. While their score reports reflect the same proficiencylevel, the eighth grader is demonstrating more skill by responding to more challenging content.It’s also important to consider grade-appropriate expectations when students in different grades takethe same grade-level cluster test. For example, when a sixth grader and an eighth grader take thegrades 6–8 test and both earn proficiency level scores of 4.0, this is the result of the eighth graderearning a higher scale score. The eighth grader must perform better than the sixth grader to earn thesame proficiency level score because the proficiency level is grade specific.3

Use proficiency levels to make comparisons across domains but not across grades. with the WIDA Can Do Descriptors to develop a student-specific English language skillprofile. as one of multiple criteria to determine a student’s eligibility for English languagesupport services.Proficiency levels are grade specific.A fifth grader who earns a scale scoreof 355 is at proficiency level 4, whilethat same scale score for a thirdgrader might generate a proficiencylevel score of 5.2.Proficiency levels are domain specific.A sixth grader who earns a scale score of370 in Listening is at proficiency level 4.3.That same student who earns a scale scoreof 370 in Reading has a Readingproficiency level of only 3.8.Context matters! These three studentshave all earned a proficiency levelscore of 4 on grade-level appropriate8th gradetests. The eighth grader hasth6grade2nd gradedemonstrated the most languageskill by responding to the mostdemonstrated performance to earn proficiency level 4challenging content.Think of it like this: When students in different grades each receive an “A” on a math test, theequivalent grades do not reflect an equivalent knowledge of math. The student in the highergrade likely understands math concepts the student in the lower grade doesn’t. Similarly, thegrade 8 student in this example has shown she can understand and produce more languagethan the grade 2 student can, even though they are both at proficiency level 4.Use scale scores to make comparisons across grade levels but not across domains. A scale score of 355in Listening is not the same as a 355 in Speaking! to monitor student growth over time within a domain.Scale scores are a means of comparing equivalent knowledge across grades. However,increasing expectations at higher grades mean that scale scores do not translate toequivalent proficiency levels across grades. For example, consider how a scale score of355 in Listening translates to a proficiency level score:Grade 3Proficiency Level 5.2Grade 4Proficiency Level 4.64Grade 5Proficiency Level 4.0

Scale scores precisely track student growth over time and across grades. Because scale scores takeinto account differences in item difficulty, they place all students on a single continuum that stretchesfrom kindergarten through grade 12. In addition, scale scores allow you to compare studentperformance across grades, within each domain, with more granularity than you’ll see with proficiencylevels. For example, using scale scores, you can track how much a student’s listening ability increasesfrom sixth to seventh grade or you might compare the speaking skills of your school’s second gradersto that of the fifth graders when evaluating curricula.Scale scores are not raw scores. A raw score is simply a tally of correct responses. Raw scores arenot reported for ACCESS for ELLs because they do not provide a meaningful measure of studentperformance. For example, consider two students taking ACCESS for ELLs Online. As the studentsmove through the test, their performances determine which questions they see. The lower-proficiencystudent sees easier items, and the higher-proficiency student sees more difficult items. Scale scoresreflect the fact that a student who correctly answers 10 difficult questions demonstrates a higher levelof proficiency than a student who correctly answers 10 easy questions.A scale score is reported as a single pointwithin a confidence band that shows theStandard Error of Measurement (SEM).In other words, the box beneath the scale score shows the range of scores a student mightreceive if that student took the test again and again at a single point in time.Confidence bands are a reminder that scales scores represent just one point in a range ofpotential student performance outcomes. Consider, for example, these scenarios:1)The student is healthy and well rested. The testing session goes smoothly.2) The student isn’t feeling well. The testing session goes smoothly.3) The student is healthy and well rested. The testing session is repeatedly interruptedby loud noises in the room next door.Even though the student is the same, has the same proficiency level, and responds to thesame test questions in all three scenarios, she is most likely to achieve the highest score inthe first scenario. Because ACCESS for ELLs is a statistically reliable assessment, the scoresin each scenario would be similar—but probably not exactly the same. The confidenceband reflects the expected score variation.Different methods are used to score the different domain tests of ACCESS for ELLs. The multiple-choice items of the Listening and Reading tests are machine scored. The constructed response items of the Writing test are scored by trained raters who use theWIDA Writing Scoring Scale. The constructed response items of the Speaking domain are scored locally by certified testadministrators when students take ACCESS for ELLs Paper. Recorded responses are scoredcentrally by trained raters when students take ACCESS for ELLs Online. Both testadministrators and centralized raters use the WIDA Speaking Scoring Scale.5

More tailored to instructional planning than the scoring scales are the WIDA Speaking and WritingRubrics. These rubrics detail the types of spoken and written language expected of students at eachproficiency level. For example, one characteristic of students at Level 2—Emerging is “repetitivesentence and phrasal patterns and formulaic grammatical structures.” Students at this proficiencylevel might benefit from classroom activities that encourage them to practice new phrases andsentence structures. These documents, from which the scoring scales are derived, provide a practicallist of specific skills that educators can reference as they plan classroom supports.Find WIDA rubrics and scoring scales at wida.wisc.edu/resources?keys rubrics.Composite ScoresIn addition to proficiency level and scale scores for each language domain, students receive aproficiency level score and a scale score for different combinations of the language domains. Thesecomposite scores are Oral Language, Literacy, Comprehension, and Overall.Proficiency levels are always calculated from scale scores. A Literacyproficiency level, for example, is based on the Literacy scale score. TheLiteracy proficiency level doesn’t reflect the midpoint between thestudent’s proficiency level scores in the Reading and Writing domains.Composite scores demand careful consideration. Composite scores can helpfully summarizestudent skills. However, similar composite scores can detract from critical differences betweenstudents. For example, two students with identical Overall scores might have very differentprofiles in terms of their oral language and literacy development. One student might have verystrong speaking skills, while another might excel at reading. Because a high score in onelanguage domain can inflate a composite score, a student’s individual performance in eachdomain is more informative than a single composite score.Only students who complete all four domains receive all four composite scores. If a student does notcomplete a particular domain, scores for that domain and any associated composite scores will bemissing from the student’s score report.The letters NA appear on the Individual Student Report when information recorded on a test bookletor entered in WIDA AMS specifies that a particular domain test should not be scored. When NAappears for an individual language domain, NA also appears for the composite scores calculatedusing that domain, including the Overall score.6

Spaces are blank when a test booklet is returned or an online test submitted without any evidencethat the student engaged with the test content of an entire domain test. (Practice items are notscored, so completing the practice items does not indicate that the student attempted to completethe test.) Indications that a student engaged with the test content are: Listening and Reading: A response is captured or marked for at least one scored item.Speaking: A task was scored on a paper score sheet. OR The Record function in the onlinetest platform was activated for at least one scored item. A human voice need not