Testing plays three roles in education. First, it serves a motivational role by holding students accountable for their work.
Second, testing serves an assessment function, not only for the purpose of assigning grades (“summative assessment”) but also for providing feedback to students to guide their learning (“formative assessment”). There is, in fact, no formally recognized definition of formative assessment. Perhaps because of its broad and uncertain definition, it remains uncertain how efficacious formative assessment is in improving student learning. Despite this, formative assessment is common in modern educational practice, particularly in hybrid learning paradigms.
Third and finally, summative testing intrinsically improves learning. The latter is called the “testing effect.” The testing effect is known to affect not only the retention and recall of knowledge, but also that of manual skills.
With the uncertainty surrounding the effectiveness of formative assessment, it remains an open question as to whether it (frequent low- or no-stakes testing) or the testing effect (frequent high-stakes testing) is more effective in promoting learning.
We compared formative assessment to summative assessment via their effects on learning in two sections of a course in human physiology for biomedical engineering students. One section of the course (control) used weekly quizzes between each of four exams, with students receiving the higher of two scores – either the exam score, or the average score of the quizzes. The other section of the course (experimental) used frequent, low-stakes, primarily formative assessments to help students gauge their own learning between each of four exams. Learning outcomes were assessed through a physiology concept inventory administered on the first and last days of the course, and through a subset of questions on each of the four exams that were common between the two course sections. The data showed a trend toward higher overall exam scores and post-course retention and recall in the section taught using only summative assessments compared to the section that used formative assessment. The differences, however, were not significant except when retention-recall was considered in isolation from comprehension. These data suggest that well-structured formative assessments can perform nearly as well in inducing the testing effect as frequent, higher-stakes formative assessments.
Are you a researcher? Would you like to cite this paper?
Visit the ASEE document repository at
for more tools and easy citations.