As the Age of AI unfolds at an astounding rate, questions like this continue to emerge focused on the power of AI to catalyze and accelerate innovation. Yet, too often, we answer these questions through a lens of reform – making the current structures and systems better, instead of transformation – changing the system. A reform-oriented answer here would accept far too many of the traditional definitions and methodologies that we hold within the assessment field. That reform-oriented answer might suggest faster and efficient ways to measure traditional academic knowledge, and maybe even propose approaches to testing beyond multiple choice. However, if we do not interrogate what we measure, how we measure, and when we measure on a deeper level as we think beyond the “test”, we will squander an opportunity to transform our education system toward one that is engaging, relevant, and effective for every young person.
What We Measure
We currently assess a scope of K-12 academic standards that have remained the same for over a century. In 1894 the Committee of Ten introduced the vision of commencement-level success as defined through a high school diploma. Although our workforce and society have evolved significantly, requiring an expanded set of skills and knowledge, we have never revisited the diploma at scale. Until now. Growing demand for durable skills from community, industry and higher education, has led to almost half of our states developing Portraits of a Graduate. These Portraits include core academic skills, as well as durable skills like curiosity, critical thinking, adaptability, and collaboration, which are grounded in the science of human development and predictive of multiple measures of success from high school and college attainment, to happiness. As we imagine how AI can effectively capture learning and growth, it is important to align on the purpose of our education system in service of this broader set of skills and knowledge because these constructs cannot be measured within the traditional paradigm of assessment. Prioritizing durable skills means evolving beyond the traditional concept of mastery. “Mastering” collaboration is not the same as mastering algebraic knowledge. There are indeed core skills required for collaboration that an individual can develop and demonstrate like goal-setting and active listening. However, effective collaboration does not simply rely on a number of individual skills, but contextual factors as well. How long has the team been working together? Is there a power dynamic to navigate? What levels of prior knowledge and expertise exist on the team? Successful capture of learning and growth connected to collaboration will need to include these contextual factors in order for the insight and feedback to be meaningful, which presents myriad implications for how we measure.
How We Measure
Our traditional assessment infrastructure holds a number of assumptions about knowledge and skill acquisition as a fixed and final achievement. With regard to durable skills in particular, AI will provide the opportunity to capture and present learning and growth along a trajectory, painting a more meaningful picture of development in multiple content areas and contexts. Imagine feedback on effective collaboration in both chemistry class and orchestra, with insights from both that will inform application and transfer to other contexts. In addition, when it comes to a student’s trajectory of growth, we also rely on normative scores that compare students to a “normal” trajectory. All of this reinforces the ages and stages (i.e., grades K-12) structure of our education system – a standardized and rigid approach to child and adolescent development, which science continues to reject. Whether it applies to reading or recovering from trauma, individuals do not follow one “normal” trajectory; rather, multiple pathways for development have been demonstrated within the research (Stafford-Brizard, Cantor & Rose, 2017). The potential for AI to capture and represent these multiple pathways to inform instruction and assessment at scale will make personalized learning grounded in the science of human development possible.
When We Measure
Finally, a truly transformative approach to assessment will end the concept of testing as an event. This doesn’t suggest that high stakes, high reward milestones like the completion of a complex project or public presentation of work should not exist. However, if the purpose of testing is to gauge understanding of a concept or demonstration of a skill, then AI can support the capture of this through artifacts and behaviors drawn from authentic contexts. Measurement of a student’s learning and development can occur while it is happening and demonstrated. Blurring the experience of learning and assessment will mean an end to the disruption and wasted time of educators who are traditionally forced to make space for summative and high stakes assessment. Reclaiming this time back means unlocking hundreds of hours for effective teaching and learning. As important, for the student, removing the high stakes assessment event means reductions in anxiety and stress.
The biggest opportunity we face in capturing evidence of learning and growth beyond the test, is not simply innovative assessment methods, but the potential to place the vast share of our resources toward learning experiences that make school the supportive, joyful, and purposeful place it should be – a place that prepares youth to thrive and lead beyond commencement.