Here’s an interesting way to incorporate collaboration in a quizzing strategy, with some pretty impressive results.
Beginning with the mechanics: students took three quizzes in an introductory pharmaceutical science course. First, they completed the quiz individually. After answering each question, they indicated how confident they were that their answer was correct—5 for absolutely certain and 1 for not knowing and guessing. Then for a period of time (length not specified in the article), they were allowed to collaborate with others seated near them on quiz answers. After that discussion, they could change their quiz answers, if they desired. At that point, they again rated their confidence in the correctness of the answers. Quiz answer sheets and confidence levels were then turned in. Immediately, correct quiz answers were revealed and once again students had the opportunity to discuss answers with each other.
An interesting scoring mechanism was used as well. Each correct answer was given a point, which was multiplied by the confidence rating assigned. With incorrect answers, the half a point off was multiplied by the confidence level and that amount deducted. If the question was unanswered, no points were added or deducted. The confidence scores were incorporated to encourage students to analyze their answers and confront how well they understood content needed to answer the question. They were deducted as a way to discourage guessing but to make the penalty smaller if a “guessed” answer was acknowledged.
Designing Better Quizzes: Ideas for Rethinking Your Quiz Practices
If your quiz strategies are becoming stale, this report is loaded with fresh ideas that will help you better understand why you are using quizzes, what you hope they will accomplish, and how they can best be used to facilitate student learning.
download your free copy »
An analysis of student answers revealed that answers were changed about 10 percent of the time, and 77 percent of them were changed in the direction of the right answer. The rest of the time a correct answer was changed to an incorrect one—suggesting a condition called “regressive collaboration.” In these cases, a more persuasive but not always correct student was able to convince another student to change from a correct to incorrect answer. However, the percentage of answers changed decreased significantly in the third quiz and the percentage of right answers changed to wrong decreased as well. Moreover, when students changed from a right answer to a wrong one, researchers think that may have precipitated even more analysis and insights.
Data show that student confidence in their answers increased as a result of the discussion with their peers. And feedback from students indicated their positive response to the approach. “The opportunity to discuss answers with my neighbors helped me learn during the quizzes” was agreed with by 86 percent of the students. Only 3 percent disagreed with the statement. The opportunity to discuss answers during the quiz was given to all students, but participation in this discussion was not required, and a small number of students chose not to participate in these exchanges.
The authors conclude, “It is reasonable to suggest that an approach in which assessment is viewed as a learning opportunity is likely to provide greater benefits to the student than one which seeks only to quantify what has been learned previously. In addition, the possibility of engaging in collaborative exchange is a more realistic approximation of real-life problem-solving, in which individuals are able to share their expertise in the solution of a problem or accomplishment of a task.” (p. 115)
Reference: Sainsbury, E. J. and Walker, R. A. (2008). Assessment as a vehicle for learning: Extending collaboration into testing. Assessment & Evaluation in Higher Education, 33 (2), 103-117.
