How New Schools, Tools and Policies Will Bring an End to the Big Test

President Obama joined the too-much-testing bandwagon with a late and vapid announcement. He can read opinion polls and probably sees the end of the current era of standards-based reform but he–and other people that care about equity–may be wondering what’s next?

The backlash against testing is driven by a combination of over testing, low quality assessments, and an inability to use the results of assessments to improve learning.

In Florida, for example, an average of 1% or less of the school year’s 900 hours of instruction time is taken up by statewide standardized assessments for all students.

However, local school districts also administer assessments. In 2013, ExcelinEd reviewed the 44 (out of 67 districts) assessment calendars that were available online. The review found an average of 98 local tests required on top of the state requirements. The numbers varied widely, with some districts requiring as few as eight additional tests to as many as 198 additional tests.

All too frequently, teachers and parents are unable to use the results of these assessments to improve student learning – either because the results are low quality or irrelevant or because the results are received too late to be helpful.

It’s clear that measurement has improved student learning and educational options for low-income families. It’s also clear that American schools spend a lot of time measuring.

Let’s recall how we got here.

• Federal commitment to equity. The 2002 reauthorization of federal education policy often referred to as No Child Left Behind incorporated a school accountability framework and requiring testing in grades 3-8 and once in high school.  To predict and improve performance on these tests, districts and schools added their own assessments.  
• Cheap and reliable. Since the introduction of norm referenced intelligence tests, the US has held an idiosyncratic fixation on affordability and reliability (rather than validity) when it comes to testing. That means long tests, multiple choice questions, testing windows, and security provisions.
• Development of aligned systems. After standards were introduced in every state, most urban districts tried to get kids out of lower academic quartiles by creating managed instruction systems including lesson plans, pacing guides, benchmark assessments, and professional development. Start the year with a district diagnostic test and end with a state summative test, add some school adopted tests in between and that’s a lot of testing.

Why test? Assessment plays four important roles in school systems:

1. Inform learning: continuous data feed that informs students, teachers, and parents about the learning process.
2. Manage matriculation: certify that students have learned enough to move on and ultimately graduate.
3. Evaluate educators: inform the practice and development of educators.
4. Check quality: measure program and school quality particularly what students know and can do and how fast they are progressing.

For the last 20 years, heavyweight end of year state tests were used for all four purposes–they were cheap and comparable but poorly suited for all four jobs.  But this doesn’t diminish the importance of these four roles.

The new PARCC and Smarter Balanced tests that are much better assessments of writing and thinking skills but were designed for the old system and remain focused on determining grade level proficiency. Because they include a lot of writing and problem solving, they take a lot longer than most old tests–and few of us predicted the backlash to harder longer tests.  

Even as new systems built on choice and personalization continue to develop, demand will increase for transparency and quicker test result turnaround. Educators are moving to more real-time decision making and families need to make choices. State assessments and accountability systems can certainly evolve but the four roles discussed above remain important.

What’s next? There have been eight important developments since NCLB was enacted in 2002:

1. Student internet access has improved sufficiently to support an expectation of frequent online learning and assessment.
2. Performance assessment tools make it easier to construct, manage, and assess projects and standards-aligned prompts (see features on LDC CoreTools, and Buck Institute).
3. Embedded assessments are incorporated into many forms of digital content.
4. Formative assessment systems have improved dramatically. Platforms like MasteryConnect, Acuity, Edmodo, OpenEd, and Schoology make it easy to build, administer, and share standards-aligned assessments.
5. Adaptive assessment, such as MAPS from NWEA, is widely used. Adaptive learning, which combines adaptive assessment and targeted tutoring, is gaining widespread use in blended learning models. Providers include DreamBox (K-8 math) i-Ready from Curriculum Associates (k-8 math and reading), ALEKS from McGraw Hill (mostly secondary use).
6. Broader aims of student success, including self management and relational skills, are widely recognized as important and are being incorporated into state and district goals. The hard to measure skills and dispositions require broader feedback systems than traditional standardized testing.
7. Competency-based school models, perhaps the most important sign of things to come, features mastery based progress. A series of small assessment gateways are used to manage student progress (see The Shift from Cohorts to Competency).
8. Micro-credentialing for educators (see Preparing Leaders for Deeper Learning) is another future trend. As educator preparation and development moves to competency-based systems, it will result in less reliance on summative assessments.

All of these developments lay the groundwork for the paradox of fewer tests but much better information. The most important sign of things to come is the trend towards personalized, blended, and competency-based education systems. This transition as well as the diverse education providers also entering the scene are what should be driving new assessment design. In other words, before a new assessment system can be contemplated, there must be agreement on what the instructional landscape will look like and what we want to measure.  

There are five steps states can take to lay the groundwork before states shrink or give up on end of year summative tests:

1. States should authorize the creation of innovation districts or schools to pilot a personalized, competency-based system and identify a pathway for statewide policy adoption.  Schools that commit to the transition can request flexibility from the rules or regulations that hinder innovation.  Participating districts and schools should be subject to high performance expectations in exchange for the flexibility.States should adopt competency-based graduation requirements and supporting policies.
2. States should encourage districts and networks to develop common local, formative assessment systems (see Assessment for Learning).
3. Districts and networks (and the EdTech sector) need to figure out how to combine all of the formative information (discussed here) that comes with the shift to digital learning. In particular we need to be able to compare proficiency levels and growth rates of students in different assessment rich environments.
4. States should adopt a micro-credentialing system for educator preparation and development.
5. States should adopt data backpacks–expanded electronic transcripts–that move with students

None of those are simple, but the direction of more and better data embedded into learning experiences is inevitable.  As new assessment systems are piloted, they should be tested for validity, reliability, and comparability.  Once these goals are accomplished, states must ensure that individual student growth calculations can be made, fair and equitable subgroup analysis remains a priority, and there is an ability to validate local decisions.

Good schools already know how every student is doing in every subject every day. The notion of taking a week off near the end of the year to determine proficiency is increasingly archaic. However, it will take a few years of R&D to build the tools and schools we need before we can ditch the big test.

For more see:

• Managed Instruction or Digital Learning?
• Helping Chief Academic Officers Make The Shift From Managed Instruction To Personalized Learning

This blog first appeared on redefinED. Feature image via redefinedonline.org.

Stay in-the-know with all things EdTech and innovations in learning by signing up to receive the weekly Smart Update. This post includes mentions of a Getting Smart partner. For a full list of partners, affiliate organizations and all other disclosures please see our Partner page.