Start with the end in mind
In the backward design model, you start with the end in mind. You ask, “What do the learners need to be able to do or know (demonstrate)”? You then create the assessment that will measure whether the goals are achieved.
Authentic Assessment
The best evidence is often an authentic assessment. Authentic assessments apply the skills and information presented during the course in ways that are similar to tasks performed in the real work (in the workforce). The is an older cartoon of a job applicant responding to the question, “What can you do” with “I can take tests”. Multiple-choice tests are useful for measuring recall and recognition. Test questions can be designed to assess high-order thinking and critical reasoning. For example, an Excel exam can offer a spreadsheet with an error or errors and the learner needs to identify the correct answer and what formula is needed to correct the mistake(s). Or for example, a scenario may be presented and a few follow-up questions refer back to the scenario. An advantage of multiple-choice exams is that they may be machine-graded. That reserves the instructor’s time for other feedback and forms of evaluation. It also allows learners to move forward and not wait on an exam to be graded before proceeding. Using Mastery Paths (Canvas), once an auto-graded assessment is completed, learners are directed to appropriate remediation or next steps. That said, often auto-graded exams, tests, and quizzes do not require higher-order thinking. They are, therefore, not examples of authentic assessment.
Vary Assessment Formats
Consider a mix of system-graded assessments and instructor-graded in your course. Also, consider guided practice in your course. When students practice on their own without instructor feedback, the activity can lack quality. It not only may be inefficient, it may be ineffective. Offer the learners multiple and diverse (several formats) opportunities to apply and reflect on what they are learning.
Test Security
Institutions and/or students spend a lot of money attempting to prevent cheating. Unfortunately, remote exam proctoring of high-stake exams often requires technology. This process adds a layer of challenge to the assessment in addition to demonstrating the acquisition of content knowledge. It adds a skills and access to equipment issue for learners. One thing to consider in your course is creating assessments on which it is difficult to cheat. Below are a few assessment design tips:
- Open-book exams:
- students may use whatever resources they have at their disposal
- open-book exam may contain multiple-choice questions
- you could ask students to apply concepts from class to novel real-world scenarios that they come up with
- you could use a program like Turnitin to check answers
- when possible, use formula-based questions, for which an algorithm can generate unique versions of the question for each student
- use application-based questions that require more than simply looking up an answer
- avoid using publisher materials, since answers to these questions can typically be easily located online
- create a large test bank, with a subset of questions chosen at random, and in random order, so that no two students receive the same exam (create a departmental question pool)
- Avoiding multiple-choice exams
- Consider oral exams over Zoom, or recorded
- Creative projects completed over the course of the semester
References