This resource provides a brief introduction to designing your course assessment strategy based on your course learning objectives and introduces best practices in assessment design. It also addresses important issues in grading, including strategies to curb cheating and grading methods that reduce implicit bias and provide actionable feedback for students.
In this document, assessments refer to all the ways students’ learning can be measured. This includes summative assessments such as tests and papers, but also formative assessments such as a survey to gauge understanding of course concepts.
Table of Contents:
- Crafting effective assessments
- Encouraging academic integrity
- Grading fairly
- Resources & further reading
Crafting effective assessments
Tie assessments to the course learning objectives. To determine what kinds of assessments to use in your course, consider what you want the students to learn to do and how that can be measured. When designing an overall plan, it is important to begin with the end in mind.
Consider what type of assessments best fit your learning objectives. For example, a case study is appropriate for measuring students’ ability to apply skills to a new situation, while a multiple choice exam is better for testing their understanding of concepts. This table of assessment choices from Carnegie Mellon University can help you think about the alignment of learning objectives and types of assessments.
Rethink traditional assessment to enhance the learning experience. At the end of a learning unit or module, summative assessments are frequently employed to measure students’ learning. These assessments are usually graded, cumulative in design and take the form of a midterm exam, research paper or final project. Consider replacing a traditional assessment with an authentic assessment situated in a meaningful, real-world context or modifying existing assessments to “do” the subject instead of recalling information. Here are some high-level questions for to get you started:
- Does this assessment replicate or simulate the contexts in which adults are “tested” in the workplace, civic life or personal life?
- Does this assessment challenge students to use what they’ve learned in solving or analyzing new problems?
- Does this assessment provide direct evidence of learning?
- Is this assessment realistic? Have students been able to practice along the way?
- Does this assessment truly demonstrate success and mastery of a skill students should have at the end of your course?
Further considerations for authentic assessment design are available in this guide from University of Illinois.
In practice, authentic assessments look different by discipline and level of the course. A good starting point is to research common examples of alternative assessments, but consider researching approaches in your discipline. There are also ways to improve traditional assessments such as quizzes to be a measure of true learning instead of memorization.
Our page on Alternative Strategies for Assessment and Grading outlines some options for creating assessment activities and policies which are learning-focused, while also being equitable and compassionate. The suggestions are loosely grouped by expected faculty time commitment.
Tailor learning by assessing previous knowledge. At the beginning of a learning unit or module, use a diagnostic assessment to gain insight into students’ existing understanding and skills prior to beginning a new concept. Examples of diagnostic assessments include: discussion, informal quiz, survey or a quick write paper (see this list for more ideas).
Use frequent informal assessments to monitor progress. Formative assessments are any assessments implemented to evaluate progress during the learning experience. When possible, provide several low-stakes opportunities for students to demonstrate progress throughout the course. Formative assessments provide five major benefits: (1)
- Students can identify their strengths and weaknesses with a particular concept and request additional support during the learning unit.
- Instructors can target areas where students are struggling that should be addressed either individually or in whole class activities before a more high-stakes assessment.
- Formative assessments can be reviewed and evaluated by peers which provides additional opportunities to learn, both for the reviewer and the student being reviewed.
- Informal, low-stakes assessments reduce student anxiety.
- A more frequent, immediate feedback loop can make some assessments (like graded quizzes) less necessary.
Examples include quick assessments like polls which can make large classes feel smaller or more informal, or end-of-class reflection questions on the day’s content. This longer list of low-stakes, formative assessments can help you find methods that work with your content and goals.
Use rubrics when possible. Students are likely to perform better on assessments when the grading criteria are clear. Research suggests that assessments designed with a corresponding rubric lead to an increased attention to detail and fewer misunderstandings in submitted work. (2) If you are interested in creating rubrics, Arizona State University has a detailed guide to get started.
|How Rubrics Help Students
How Rubrics Help Instructors
|Improve student performance by clearly showing
the student how their work will be evaluated
and what is expected.
|Encourage the instructor to clarify their criteria in specific terms.
|Help students become better judges of the quality of
their own work.
|Provide objectivity and consistency in grading student work.
|Provide students with more informative feedback about their strengths and areas that need improvement.
|Provide useful feedback to the instructor regarding the effectiveness of instruction.
Break up larger assessments into smaller parts. Scaffolding major or long-term work into smaller assignments with different deadlines gives students natural structure, helps with time and project management skills and provides multiple opportunities for students to receive constructive feedback. Students also benefit from scaffolding when:
- Rubrics are provided to assess discrete skills and evaluate student practice via smaller pre-assignments.
- The stakes are lowered for preliminary work.
- Opportunities are offered for rewrite or rework based on feedback.
Use practices that promote inclusive assessment design. Take inventory of the explicit and implicit norms and biases of your course assessments. For example, are your assessment questions phrased in a way that all students (including non-native English speakers) can be successful? Do your course assessments meet basic accessibility standards, including being appropriate for students with visual or hearing needs?
Encouraging academic integrity
The Duke Community Standard embraces the principle that “intellectual and academic honesty are at the heart of the academic life of any university. It is the responsibility of all students to understand and abide by Duke’s expectations regarding academic work.” (3) Learning the rules of legitimacy in academic work is part of college education, so the topic of cheating and plagiarism should be embraced as part of ongoing discussion among students, and faculty instructors should remind students of this obligation throughout their courses.
Include a statement about cheating and plagiarism in your syllabus. Remind students that they must uphold the standards of student conduct as an obligation of participating in our learning community. This can be reinforced before important assessments as well. Studies have shown that when students have to manually agree to the Honor Pledge prior to submitting an assignment (either online or in person), they are less likely to cheat.(4)
Specify where training is available. Because of their cultural or academic experiences, some students may not be familiar with what constitutes plagiarism in your course. Students can use library resources to learn more about plagiarism and take the university’s plagiarism tutorial.
Include specific guidelines for collaboration, citation and the use of electronic sources for every assessment. For example, it may be necessary to define what kinds of online sources are considered cheating for your discipline (for example, online translators in language courses) or help students understand how to cite correctly.
Provide ongoing feedback to reduce the temptation to cheat. Students are more likely to seek short cuts when they don’t know how to approach a task. Requiring students to turn in smaller parts of a paper or project for feedback and a grade before the final deadline can lessen the risk of cheating. Having multiple milestones on larger assessments reduces the stress of finishing a paper at the last minute or cramming for a final exam.
Ask questions that have no single right answer. The most direct approach to reduce cheating is to design open-ended assessment items. When writing test or quiz questions ask yourself: could this answer be easily discovered online? If so, rewrite your question to elicit more critical thinking from your students.
Open-ended assessments can take the form of case studies, projects, essays, podcasts, interviews or “explain your work” problem sets. Students can provide examples of course concepts in a novel way. They can record themselves explaining the idea to someone else or make a mind map of related events or ideas. They can present their solutions to real-world scenarios as a poster or a podcast. If you choose to conduct an exam, designing questions that ask students to decide which concepts or equations to apply in a scenario, rather than testing recall, may make the most sense for many courses. You could include an oral exam component where students explain their work for a particular problem.
Minimize opportunities for cheating in tests and quizzes online. If you offer quizzes or tests through Sakai, there are several steps that you can take to reduce cheating, plagiarism or other violations:
- Sakai tests include a pledge not to violate the Duke Community Standard. You could also have this printed at the top of a physical test.
- Limit time. Set a time limit that gives students enough time to properly progress through the activity but not so much that unprepared students can research every question.
- Randomize question or answer order. When you randomize (or shuffle) your test or quiz questions, all students will still receive the same questions but not necessarily in the same order. This strategy is particularly useful when you have a large question pool and choose to show a few questions at a time. When you randomize the answers to a question, all students will still receive the same answers but not necessarily in the same order.
- Use large question pools. Pools allow you to use the same question across multiple assessments or create a large number of questions from which to pull a random subset. For example, you could develop (or repurpose) 30 questions in a pool and have Sakai randomly choose 15 of those questions for each student’s assessment.
- Hide correct answers and scores until the test or quiz is closed. This can prevent students from sharing questions and answers with peers during the assessment period.
- Require an explanation of the student’s answer. Require a rationale for their answer either as a short text question or perhaps a voice recording.
Duke has chosen not to implement a proctoring technology. When thinking about proctoring, keep in mind how implementing such policies and technologies might affect our ability to create equitable student-centered learning experiences. Several issues of student well-being and technological constraints you might want to keep in mind include:
- Student privacy: In an online setting, proctoring services essentially bring strangers into students’ homes or dorm rooms — places students may not be comfortable exposing. Additionally, often these services record and store actions of students on non-Duke servers and infrastructure. This makes proctoring services problematic for the in-class setting as well. These violations of privacy perpetuate inequity through the use of surveillance technologies.
- Technology access: If testing is online all students may not have the same access to technology (e.g., external webcams) for proctoring.
- Accessibility: Proctoring software can create more barriers for students who need accommodations.
- Unease: Proctoring reinforces a surveillance aspect to learning, which impacts student performance.
Start with clear instructions, a direct assignment prompt and transparent grading criteria. Explicit instructions reduce confusion and the number of emails that you may receive from your students requesting clarification on an assignment. Your assignment instructions should detail:
- Length requirements
- Formatting requirements
- Expectations of style, voice and tone
- Acceptable structure for reference citations
- Due date(s)
- Technology requirements needed for the assignment
- Description of the measures used to evaluate success
Offer meaningful feedback and a timely response when grading. There are many ways to provide feedback to students on submitted work. Regardless of the grading strategy and tool that you choose, there are a few best practices to consider when providing student feedback:
- Feedback should be prompt. Send feedback as soon as possible after the assignment to give students an adequate amount of time to reflect before moving on to the next assignment.
- Feedback should be equitable. Rubrics can help ensure that students are receiving consistent feedback for similar work.
- Feedback should be formative. Meaningful feedback focuses on students’ strengths and shares constructive areas to further develop their skills. It is not necessary to correct all errors if patterns can be pointed out.
We recommend avoiding curves for both individual assignments and final course grades. There are several downsides to curves that will negatively impact your pedagogy:
- Curves lower motivation to learn and incentivize cheating
- Curves create barriers to an inclusive learning environment
- Curves also “often result in grades unrelated to content mastery” (Jeffrey Schinske and Kimberly Tanner)
Rather than using curves, you can introduce feedback strategies that allow students to improve their performance on future assessments by revising submitted work or reflecting on study habits.
Create customized rubrics to grade assignments consistently. Rubrics can reduce the grading burden over the long-term for instructors and increase the quality of the work students create. A well-designed rubric:
- Provides clear criteria for success that help students produce better work and instructors to be consistent with grading.
- Points out specific areas for students to address in future assignments.
- Allows for consistency in grading and more meaningful feedback.
Grade students anonymously. Blind grading removes any potential positive or negative bias when reviewing an individual’s work. The main assessment tools at Duke, Sakai and Gradescope, have easy controls for implementing anonymous grading.
Use a grade book that is visible to students. Students should have online access to their grades throughout the semester. It is not necessary to post their cumulative course grade at all times, but seeing the individual items is important. Knowing how they are doing reduces student stress before big assessments. An open and up-to-date grade book provides opportunities for students and instructors to address issues in a timely manner. It allows students to correct any omissions by the instructor and instructors have an immediate sense of which students are struggling as well.
Resources & further reading
Best Practices for Inclusive Assessment (Duke University)
What are inclusive assessment practices? (Tufts University)
Sequencing and Scaffolding Assignments (University of Michigan)
Blind Grading (Yale University)
Using Rubrics (Cornell University)
How to Give Your Students Better Feedback with Technology (Chronicle of Higher Education)
- The Many Faces of Formative Assessment (International Journal of Teaching and Higher Education)
- A Review of Rubric Use in Higher Education (Reddy, Y, et al, Assessment and Evaluation in Higher Education)
- Duke Community Standard
- The Impact of Honor Codes and Perceptions of Cheating on Academic Cheating Behaviors, Especially for MBA Bound Undergraduates (O’Neill H., Pfeiffer C.)