The problem openness creates for tests
Many traditional assessments reward remembering facts under time pressure, with no reference material available. That made sense when facts were hard to look up. When information is freely available to anyone with a connection, a test of recall measures something narrower than before, and a take-home task that can be answered by a quick search shows little about what a learner understands.
Open education pushes the question: what do we actually want to know about a learner, and how could we see it? The answers tend to point toward tasks that require using knowledge rather than reciting it. This article looks at three common approaches (open-book assessment, portfolios and public work) and the practical issues around each. Evidence on how well each works is mixed, and none is a universal fix.
Open-book and open-resource assessment
In an open-book exam, students may consult notes, textbooks or other materials. The aim is to test application and reasoning instead of memory.
What it can do well. Questions can ask students to analyze a new case, compare approaches, or justify a choice, which is closer to real work. Students often spend less effort on rote memorizing and more on understanding how ideas connect.
What to watch. Poorly written open-book questions are simply recall questions in disguise; students then burn time hunting for answers. Good questions are usually harder to write and take longer to answer. Students who have not practiced using their materials efficiently may be at a disadvantage, so a practice paper helps. Where internet access is allowed, there is also a risk of collaboration or outsourcing, which is a reason to combine a written task with a short conversation or a follow-up question.
Portfolios
A portfolio is a collection of work over time, usually with the learner’s own reflection on what the pieces show and how they changed. It might include drafts, finished essays, project reports, code, recordings or photographs.
Strengths. Portfolios show progress, not just a final snapshot. They give learners a record they can reuse for applications. Writing a reflection, about what went well and what they would do differently, encourages self-assessment and links to ideas about thinking about your own thinking.
Limits. They take time to assemble and to mark. Standards need to be explicit, or marking becomes inconsistent. Learners with fewer resources at home may produce less polished work through no fault of their own, so criteria should focus on learning, not presentation polish. And a portfolio can reveal personal information, so students should control what is included.
Public and authentic work
Another approach asks students to produce something for a real audience: a revised open textbook chapter, a public explainer, a translation, a data set, or a community project. This is sometimes called authentic assessment, because the task resembles real-world work and the audience extends beyond the marker.
Potential benefits. Students often take more care when others will see the result. The output can have value beyond the course, and learners gain a body of work they can point to.
Concerns that deserve attention.
- Privacy and safety. Publishing under a real name creates a lasting public record. Younger learners especially should not be required to share personal information. Offer options such as pseudonyms, a private space, or a choice of audience.
- Consent. Learners should know in advance how their work will be shared, and under what license.
- Pressure. Some students will find public exposure stressful. Alternatives should be available without penalty.
- Fairness. An audience’s response, such as likes or comments, should never be part of the grade.
- Quality and accuracy. Public contributions need review so that errors are not published unchallenged.
Changing what is graded
Rethinking tasks usually also means rethinking criteria. A few shifts are common.
- From a single mark to feedback on specific strengths and next steps.
- From one high-stakes exam to several smaller, lower-stakes tasks that count in total.
- From ranking students against one another to judging them against stated standards, as in competency-based education, where learners move on when they show mastery.
- From individual-only work to a mix, with clear ways to identify each person’s contribution to a group task.
- From the final product only to the process: planning notes, drafts, and responses to feedback.
Frequent low-stakes quizzes are one way to combine these shifts with retrieval practice, which many studies associate with better retention, though the size of the effect varies with the subject and design.
Practical issues
Workload. Richer assessment generally means more marking time. Rubrics, peer review, and well-designed checklists help, but a teacher with large classes may need to limit the number of complex tasks. Realistic planning matters.
Consistency. When different teachers grade portfolios or projects, comparing samples and agreeing on standards reduces unfair differences.
Accessibility. Any assessment should be usable by learners with disabilities, using accessible formats and reasonable adjustments, without extra burden on the student.
Recognition. Institutions and exam bodies have their own rules. A teacher inside a formal program should check what is permitted before changing a required exam. Open badges or similar records may document achievement outside formal grades, though acceptance differs.
Online proctoring and monitoring. Some institutions use software to watch students during exams. These tools raise privacy and fairness concerns, and many educators prefer redesigning the task so that surveillance is less necessary.
A cautious conclusion for teachers
Openness changes what is easy to find and what is easy to share, so it is reasonable to ask assessments to reflect that. Start small: convert one recall-heavy test into an application task, add a reflective note to one assignment, or let students choose between a private and a public version of a project. Collect feedback from students, check whether the results tell you something new, and adjust. There is little evidence that any single format is best for everyone, and the right mix depends on your subject, your learners and the rules you work under.