Smarter Balanced Assessments A COMPLETE CHANGE IN THE PARADIGM OF STUDENT ASSESSMENT Smarter Balanced stats 29 states 15 million students 40,000 test items 12 week testing window for the mandatory summative test Not a timed test, likely shorter than current practice Fast turn around time on results Why is technology needed? Computer adaptive questions (more on this later) Some CAT are interactive in nature (ex: draw something) Performance tasks require Stimulus Information Processing tools and tasks Video clips Photos 2-D and 3-D modeling databases Data analysis software Simulations Spreadsheets Equation editor tools Chat rooms Products or performances Important terms to understand Computer adaptive testing Performance assessment Formative/interim assessment Summative assessment Computer adaptive testing (CAT) Based on student responses, the computer program adjusts the difficulty of questions throughout the assessment. For example, a student who answers a question correctly will receive a more challenging item, while an incorrect answer generates an easier question. By adapting to the student as the assessment is taking place, these assessments present an individually tailored set of questions to each student and can quickly identify which skills students have mastered. Weight-lifting analogy A “fixed” test is like lifting a 50 lb dumbbell Either you can or you can’t—doesn’t tell if you could lift one 5 lbs more or less If everyone can lift a 5 lb dumbbell, doesn’t tell you how much more you are able to lift CAT brings up the difficulty in small increments so that it is clear precisely “how much weight” you can handle Validity Program is being designed using Artificial Intelligence tools currently used for assessments such as GRE, GMAT, SAT About 10% of the constructed and extended response (includes written response of any length from short answer to sentences to essays) will be checked by a human scorer Testing versus assessment T: Results in a quantifiable score that demonstrates what students know or can do at a point in time A: Multiple measures over time that reflect mastery, growth, depth of knowledge, gaps in learning T: Right or wrong answers A: Multiple correct answers are acceptable; Answers illustrate student thinking and understandings/misunderstandings of content Testing versus assessment T: Occurs at the end of the learning process, designed to measure/rank students A: Ongoing and actionable during the learning process, designed to improve teaching/learning T: Evaluative—marks an end A: Diagnostic-opportunities to improve Performance assessment Performance tasks challenge students to apply their knowledge and skills to respond to complex real-world problems. They can best be described as collections of questions and activities that are coherently connected to a single theme or scenario. These activities are meant to measure capacities such as depth of understanding, writing and research skills and complex analysis, which cannot be adequately assessed with traditional assessment questions. No single correct answer Interim/formative assessments Optional comprehensive and content-cluster assessment to help identify specific needs of each student Can be administered throughout the year Provides clear examples of expected performance on Common Core standards Includes a variety of question types: selected response, short constructed response, extended constructed response, technology enhanced, and performance tasks Aligned to and reported on the same scale as the summative assessments Fully accessible for instruction and professional development Summative assessments Assesses the full range of Common Core in English language arts and mathematics for students in grades 3–8 and 11 (interim assessments can be used in grades 9 and 10) Measures current student achievement and growth across time, showing progress toward college and career readiness Can be given once or twice a year (mandatory testing window within the last 12 weeks of the instructional year) Includes a variety of question types: selected response, short constructed response, extended constructed response, technology enhanced, and performance tasks Item design All of the Common Core Standards will be clustered into testable “claims” Students will start with an entry level item aligned to grade level standards Items will get easier or harder depending on answer Every student will walk away saying “this was a hard test” Every student will persist with testing until they get half the items wrong Clear, precise picture of student ability Timeline School year 2011-12: test specifications, item development, technology readiness tools March 2013: CMT/CAPT tests given for the last time Spring 2014: CSDE applying for waiver from accountability testing in order to pilot versions of the Smarter Balanced Assessments School year 2014-15: districts will be able to use Smarter Balanced tools for formative and summative assessments Technology issues Testing will require access to Internet Decisions still pending on the variety of devices that can be used (security, protocol issues, authentications) There will be minimum hardware and software requirements needed: attempt to make accessible to older OS and less powerful machines Technology readiness tool will collect this information Technology Readiness Tool Degree to which hardware and software (devices) at the school level meet the minimum requirements set by each of the consortia. Adequate ratio of test taking devices to test takers within a prescribed time table. Sufficient bandwidth to handle the expected volume of traffic and content. Adequate number of trained personnel to support the process. System Architecture Resources and sample items will be open source Most of testing activity will occur in a cloud environment Security will mimic protocols used by banking industry Intent to deliver testing on tablets (either iPad or Android), minimum 9.5” screen since many districts are purchasing these-may need keyboards Tablets will be necessary for 25% of the math assessment planned for 2016-17 New purchase specifications Hardware 1 GHz or faster processor 1 GB RAM or greater memory 9.5 inch (10 inch class) or larger screen size 1024 x 768 or better screen resolution Operating System Windows 7 Mac 10.7 Linux (Ubuntu 11.10, Fedora 16) Chrome OS Apple iOS Android 4.0 Networking Wired or wireless Internet connection Device Type Desktops, laptops, netbooks, thin client, and tablets that meet the hardware, operating system, and networking specifications Security issues Every student’s test is likely to be somewhat different Test administration will likely take less time than a prescribed window Grade 3 Math 45-90 minute range for performance tasks and 45-60 minute range for extended response Grade 4-11 Math 60-120 minute range for performance tasks and 4560 minute range for extended response Tasks may cross days Whole group testing is unlikely; students are more likely to be assessed in smaller groups over the 12 week window for summative accountability testing Allowable teacher and peer interactions, group work Required tools and resources will be embedded within online system; not free access to Internet for searches Increasing rigor This is NOT CMT/CAPT given through a computer Because CAT changes the difficulty of questions, no one will “get them all right” Assessments will yield a precise measurement of what content students have mastered Scoring will be done with Artificial Intelligence programs currently used for GRE, GMAT, and other standardized assessments (only 10% human check) Typical CMT test item http://www.csde.state.ct.us/public/cedar/assessmen t/cmt/math_handbook.htm http://www.csde.state.ct.us/public/cedar/assessmen t/cmt/l_a_handbook.htm Typical Smarter Balanced test item http://www.smarterbalanced.org/wordpress/wp- content/uploads/2012/02/DRAFT%20Math%20Ite m%20Specifications.pdf http://www.smarterbalanced.org/wordpress/wpcontent/uploads/2012/02/DRAFT%20ELA%20Item %20Specifications.pdf http://www.smarterbalanced.org/wordpress/wpcontent/uploads/2012/02/DRAFT%20Performance %20Task%20Specifications.pdf Sample question Old question: Randa ate 3/8 of a pizza, and Marvin ate 1/8 of the same pizza. What fraction of the pizza did Randa and Marvin eat? a. 5/8 b. 3/8 c. 1/4 d. 1/2 (Answer: d) New question: Tito and Luis are stuffed with pizza! Tito ate one-fourth of a cheese pizza. Tito ate three-eighths of a pepperoni pizza. Tito ate one-half of a mushroom pizza. Luis ate five-eights of a cheese pizza. Luis ate the other half of the mushroom pizza. All the pizzas were the same size. Tito says he ate more pizza than Luis because Luis did not eat any pepperoni pizza. Luis says they each ate the same amount of pizza. Who is correct? Show all your mathematical thinking. (Answer: Luis is right — both ate 1 1/8 of a pizza).