Assessment has always been an important part of learning but lately, we’ve grown to rely on quizzes and tests that are artificial, expensive, single-purpose, uninformative, and inequitable. Our peculiar American affliction is our fixation with reliability (consistent and cheap) over validity (authentic and meaningful).
In the same way the business world has shifted to a focus on customer experience informed by big data, the future of education is learner experience informed by measurement. And by measurement, it’s not just to planned assessments but embedded checks, adaptive adjustments, gathered observations, reflections, and system diagnostics.
The 13 building blocks outlined below build on David Conley’s outline of next-generation assessment. They describe how measurement systems will incorporate lots of small measures embedded in and associated with learning experiences rather than big inauthentic end of year tests. The first half describes the role of measurement in a learning journey. The second half describes robust measurement systems.
Smart Measurement Rather than Dumb Assessment
1. Learner growth: Measurement is primarily focused on informing learner growth. Program administration and system accountability come second. As David Conley said, “Students are actors, not objects.
2. Continuum: Measurement views growth on a novice-expert continuum. It’s not fixated on artificial age cohort proficiency targets measured by year-end standardized tests.
3. Embedded: Most measurement is embedded in or a reflection on learning experiences. Feedback is often immediate, always informative. It’s sometimes quantitative, sometimes qualitative.
4. Application: Measurement is focused on the application of knowledge and skills. It’s not about regurgitating facts and formulas. Sometimes the stakes are high when a mastery judgment is involved (for a certification or move to the next level) but measurement remains authentic and fair.
5. Ownership: Measurement promotes student ownership of learning and helps them identify interests and develop self-knowledge. It is as much about building success skills as academic progress with feedback that helps learners identify the next steps.
6. Actionable: Measurement produces actionable information. It doesn’t just put learners in categories. It helps them understand their own learning and promotes goal setting and persistence.
7. Insight: Measurement provides insight into the application of knowledge in context. Because expertise is context and domain-dependent, measurement is not an isolated one-shot affair. Important skills are applied to different problems in different settings to demonstrate transferability. Measurement informs teachers of what was learned and about the nature of the task and context.
Measurement Systems Rather than Assessment Events
8. Adaptive: Measurement is incorporated into adaptive systems that identify the next best steps–both in level and type of challenge. Adaptive systems particularly good for intensive skill-building that enable equitable contributions to extended challenges.
9. Profiles: Measurement is used to create comprehensive learner records in ways that aid individual development and goal attainment, that identify the most productive environments and experiences for growth.
In a paper on the future of assessment, Australian nonprofit High Resolves described profiles as “Very large, dynamic, database of all archived cognitive, affective and behavioral indicators from multiple activity‑based assessments”. Comprehensive records inform not only individual next steps but (with full privacy) provides valuable insights into the performance of subgroups and the efficacy of learning experiences.
10. Proactive: Measurement spots evidence of desired competencies in powerful immersive experiences–both inside and outside of school. It doesn’t rely exclusively on elaborate and inauthentic deconstructed post hoc assessment tasks (those big end of year multiple-choice tests). Teachers, advisors, and algorithms can all be part of proactively spotting competencies.
You can plan for opportunities to provoke creative problem solving–but you can also spot it when and where it emerges given conditions that value curiosity and self-direction.
11. Cumulative Validity: Measurement takes advantage of cumulative validity. Combining hundreds of data points (e.g., multitrait feedback on 30 writing samples from several classes over two years–each with 2-3 revisions) can provide a much more accurate picture of writing competence than an end of year standardized test. Automated feedback systems can augment human judgment in assessing skills progressions.
Good schools know how every learner is doing on every important competency every day–they take advantage of the cumulative benefit of high-quality formative assessment and don’t need a day (or week) long test at the end of the year.
12. Equity: Good measurement systems address equity issues. They identify learners that need more time and support; they power early warning systems. Equitable measurement systems avoid tasks and tests that incorporate bias.
13. Credentials: Measurement systems help communicate milestones in capability development in the form of credentials and portfolios of artifacts that enable learners to tell their stories.
There are a handful of built-from-scratch school systems and postsecondary programs that do many of these things well. It’s harder for an existing system of schools to adopt these principles but a few dozen are on the path. CompetencyWorks notes progress in most states.
Conley sees an accelerating change in college admissions–an opening for multiple measures. And research suggests that a wider range of measures generally helps identify a more diverse candidate pool.
States can advance smart measurement with pilot programs that support schools moving in this direction. Future state accountability systems could authorize networks of schools that can present comprehensive datasets that consistently and accurately describe learner growth.
The future of learning is smart measurement–it’s a collection of authentic observations over time, it’s often adaptive and embedded, it’s proactive and results in useful profiles that help learners tell their story.
For more, see:
- A Proposal to End Standardized Testing (2019)
- Measuring What Matters (2018 podcast)
- David Conley on Next Generation Assessment (2018 podcast)
- Measuring What Matters: A Framework Review (2017)
- Now That Schools Are Promoting Broader Definitions of Success, How Do We Measure Progress? (2019)
Stay in-the-know with innovations in learning by signing up for the weekly Smart Update.