Building National Assessment Systems That Prioritize Learning and Adaptation

By Melissa Chiappetta and Benjamin Piper

Assessment often gets a bad rap. Education stakeholders worry that teachers are teaching to the test rather than teaching to ensure that students have the skills they need to be successful in the world. They also worry that assessments sometimes get used for the wrong purposes—to punish low-performing schools or teachers, for instance—or don’t get used at all, wasting scarce resources. However, when used right—to inform instruction—assessments are absolutely critical. In low-income countries, the lack of systematic large-scale assessments has often meant that governments and other stakeholders do not have a clear picture of learning gaps, which handicaps efforts to address those gaps.

A Global Look

According to the United Nations, 100 countries are failing to systematically assess student learning outcomes. Where there is evidence, it is sobering. The United Nations reports that 617 million primary and lower-secondary school children (or 55% of all school children) are not minimally proficient in reading or mathematics. This problem is particularly acute in sub-Saharan Africa, where more than 84% of students are not proficient in reading or math. Assessments, when used right, offer countries an opportunity to not only understand the scope of learning challenges but to explore why gaps in learning exist so that key stakeholders can target resources to address those gaps and improve learning.

However, many countries that do assess students often fail to use the results to target resources or adapt classroom instruction to address learning gaps. This hurts all students, but it is especially problematic for disadvantaged students, who frequently need more focused support. Measurement is worthwhile only if it is used. But knowledge of how to use assessment results is not widespread.

What the Evidence Says

Fortunately, the body of evidence on what effective assessment looks like is growing. Yet busy policy makers, donors, and partners often do not have time to wade through large volumes of text to identify the key actions they need to take to make large-scale assessments useful. For that reason, the Bill and Melinda Gates Foundation has partnered with RTI International to synthesize the evidence and provide a list of key actions that countries, donors, and partners seeking to design or refine national assessment systems should take. The result is a how-to guide titled Assessment-Informed Instruction: Systems Level (a second guide on Assessment-Informed Instruction: Classroom Level is also available)

The guide provides an overview of the various types of assessments in the education system and which ones work best for which purposes. It also offers practical instructions on how to examine an existing assessment system and sets out eight key recommendations for improving that system or building a new large-scale assessment system if one does not already exist. These eight key takeaways work together for one purpose—to ensure the use of assessment results to inform instruction and target resources—so that assessment becomes the invaluable tool it has the potential to be.  

The eight key recommendations focus on the following:

1. Embed assessment into the education system. Assessments need to be owned and run by the government. This means that the establishment, conduct, and financing of assessment programs and agencies, including information on which agency will conduct the assessments and which agencies will use the results, needs to be legislated. And donors and partners should work with countries’ existing assessment systems rather than creating parallel ones.
    
2. Be realistic about costs and sustainability. When deciding which assessments will be most useful within a system, countries and their partners need to weigh the costs and benefits. They need to weigh sample size, rigor, comprehensiveness of questions, the use of complimentary data collection tools, the use of dashboards, and who will conduct the assessment (e.g., government officials versus trained enumerators) against the purpose and budget of the assessment. For instance, if an assessment is going to be used to inform teacher training, it may be helpful to pair the assessment with classroom observations and to make the data available to teachers via an online dashboard.
    
3. Ensure alignment of assessment within the education system. Assessments need to be aligned across levels but also with the country’s performance standards, benchmarks, targets, curricula, teacher training, materials, and classroom instruction so that students are not being assessed on content to which they haven’t been exposed and are getting credit for knowledge they have gained in the classroom.
    
4. Clarify the purpose of assessments. Assessments have three main purposes: to set expectations for learning outcomes, to monitor and hold schools accountable for meeting those expectations, and to intervene to support struggling schools and students and ensure equity for all students. Understanding the purpose of the assessments is critical for making decisions on the cost trade-offs discussed above, as well as ensuring that results are adequately presented to the key users.
    
5. Focus assessments on what teachers and policy makers need to know. Assessments should differentiate between key skills, be inclusive, and contain demographic information to support targeting. Without this information, key assessment users will not be able to determine why there might be gaps in learning (so they can work to improve instruction), nor will they have the information needed to target resources and interventions to the groups most in need.
    
6. Decide when and how often to collect assessment data. Timing must be linked to purpose. For instance, school-level assessments should occur several times a year to ensure that teachers are able to adapt instruction to individual student needs and that governments are able to target teacher coaching resources to the schools and teachers most in need. Assessments focused on monitoring system changes, on the other hand, need occur only every two years, as large-scale changes take longer to show up in the education system.
    
7. Find a balance between accountability and learning. Assessment results should be disseminated only to those who have a clear need for the data, and these users should be provided with explicit instructions on how to use the data. For instance, teachers should get information on key gaps in student skills, with details on how they might address those gaps through example lessons, pedagogical techniques, and information on the students who need the most support. Also, incentives are more effective than punishments, as punishments make assessments high stakes and make it more likely that teachers and schools will teach to the assessment or look for other ways to “game” the system.
    
8. Ensure targeted dissemination. Dissemination needs to be tailored to each specific audience. For instance, policy makers and decentralized government officials need clear, concise information, with indications of the schools and teachers who need the most support and what type of support is needed, whereas teachers and administrators need more granular information on which items students answered correctly and incorrectly, as well as remediation methods they might use to help fill student learning gaps.
    

The Road Ahead

This guide is meant to help countries build more impactful assessment systems—systems that prioritize learning and adaptation in order to continually improve instruction and equity. In doing so, countries will further their progress toward Sustainable Development Goal 4.1, which aims to ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes.

That said, the guide is just a starting point. Inevitably, as more and more countries design and refine their learning assessment systems and implement the recommendations from the guide, more nuanced evidence will emerge on what works. For that reason, countries, donors, and partners should track the effects of their practices on awareness of learning gaps, targeting of resources, instruction, and learning outcomes.

Education is a fundamental human right, and it empowers people to advocate on their own behalf in the struggle against poverty and inequality. Yet globally, we are failing 617 million children and adolescents who are not meeting minimum proficiency standards and who will continue to miss the mark until we take assessment—and especially the use of the data it gathers—seriously. This guide is a resource for all of those who wish to have an impact on addressing our global learning crisis.