We have spent 100's of hours talking to 100's of educators over the last 2 years about standardized testing, and the associated problems. We're under no illusions that the current testing crisis caused by reformers is complex, layered at the local, state and federal levels, and has stakeholders who hold divergent views.
In a perfect world, testing designed to evaluate an education system ought to be grade-span, and only require a sampling of students, akin to NAEP. It is over-kill to test every student in every subject, every year. However, we do not live in a perfect world, therefore Join the Future has identified the following problems, and provides the following recommendations as a way forward
The technology deployment for this years testing has been a predictable debacle. Perhaps half of Ohio's schools felt so lacking in technological capacity that they performed the tests on paper. For the remainder, most lacked sufficient technology to perform the testing in a short period of time, instead having to schedule weeks of testing that caused massive classroom disruption.
The testing software itself was a mess. Technology and media specialists in schools spent hundreds of hours, making thousands of tech support calls to testing companies trying to resolve technical problems with the software. Teachers were provided with inadequate training. Some students had to re-take the test up to 3 times because of technical difficulties. One can only imagine the stress a 3rd grader must have felt.
There was a lack of hardware compatibility between PARCC and AIR testing platforms. School IT professionals spent countless hours re-imaging computers to switch between testing use and general use. School libraries and computer labs in schools throughout the state have been unusable for most of the time tests have been taking place.
1. Schools must be provided with the flexibility to offer paper based testing for at least 3 years.
2. Any testing solution must be platform agnostic and work on desktops, laptops, Tablets and Chromebooks alike, without need for special software or re-imaging.
3. Educators must be provided with adequate time to train on the platform.
4. Students must be given adequate practice time with the testing platform. They don't need to be taking 2 tests in one - one on the intended content and the other on how to navigate a complex software product.
5. Schools must be provided with adequate resources to purchase compatible technology capable of testing the entire student population within one week.
6. Schools must be provided with adequate resources to ensure they have bandwidth to perform the testing seamlessly.
Problem: Testing Time
Testing cannot consume weeks and weeks of instruction time, and be repeated. Schools spent almost the entire month of February in a state of disruption, made worse by inclement weather and snow days, only to have to perform yet more tests in April. If schools had lost the entire month of February to 4 feet of snow it would have been viewed as a statewide crisis, yet that is the practical effect testing had on schools this year.
Testing Time Recommendations
1. Tests need to be shorter.
2. Tests must be align with a typical classroom schedule, e.g. if a typical classroom period is 30 minutes, the tests cannot be 40 minutes. That disrupts two periods and the subsequent schedule all day.
3. Schools need enough infrastructure to perform the tests in a single week, just once a year.
Problem: High Stakes
Attaching high stakes to a new and unstable testing regime is unfair and has caused lasting damage to morale and motivation. Stories of students getting physically and mentally sick with anxiety are common place. Educators have faced a high stakes evaluation system under constant change. Furthermore, it is becoming apparent to most students that the tests don't matter. Outside of 3rd grade and HS, the results of standardized testing has no impact on students, but deeply affects educators and their schools. This mis-alignment of stakes is a huge problem.
High Stakes Recommendations
1. Implement at least a 3 year moratorium on high stakes consequences until the system matures and proves itself.
2. Stakes must be aligned. It is simply not fair or appropriate for educators to face career consequences for tests students are taking, where the students themselves have no stake in the result. We recommend the elimination of student tests scores as a means to measure teacher quality at the individual level.
Problem: Inappropriate Test Content
Testing has been widely criticized for being age inappropriate. Many test questions were deployed at reading ages much higher than the target age range of the test takers. Often times, especially in Math, the root question itself might be appropriate, but couched in language far too complex for the student to comprehend. Not only does this need to cease, it needs to be investigated.
Reports that questions have spread across multiple pages or screens, forcing a student to remember large blocks of complex text, while navigating back and forth have been widespread. Students should be evaluated on content knowledge not their memories or ability to navigate complex software products
Test Content Recommendations
1. Tests must be age appropriate in content and reading level.
2. Questions must be easily navigable, or compact so that students can concentrate on the answer and not a page flipping ordeal.
3. The State should have a review panel to vet test questions in advance, with the ability to veto inappropriate questions.
4. An element of a testing companies contract should be tied to the appropriateness and accuracy of the tests they provide.
5. Test questions and their answers should be made public within 4 years.
Problem: Useful Results
Parents, students and schools should not have to wait 6 months to receive results, especially results performed electronically. By the time this years results are made available, many students will have moved classes, schools, districts and maybe even out of state.
Results need to provide more than a meaningless score. Educators would benefit from results that include identifying a students strength and weaknesses.
Furthermore, who is grading the tests needs to be closely examined. There has been widespread reports of craigslist advertisements for tests scorers, who require little training or knowledge.
Test Result Recommendations
1. Results should be available before the end of the school year, with a portion of a testing companies contract tied to timely delivery of results.
2. Results should include a description of where a student excelled and performed poorly.
3. All test answers should be publicly available within 4 years, with a portion of a testing companies contract tied to accuracy.
4. A system of qualifying test scorers, and evaluating their accuracy is needed.
Evidently the problems are significant and will require time to implement. Policy makers must resist the calls from corporate reformers for quick fixes. It has been their influence and misguided advice that has navigated us in to this crisis. If a test is being performed in a school the first questions ought to be how does this benefit the students education? Is the benefit of the test proportional to any of its impacts on student education?
Parents, students and educators alike would like to see a reduction in the total volume of testing, with the testing that remains primarily aimed at improving student learning.