Title: Increasing learning in a college physics course with timely use of short multimedia summaries
Authors: Spencer Dunleavy, Greg Kestin, Kristina Callaghan, Logan McCarty, and Louis Deslauriers
First author’s institution: Columbia University and Harvard University
Journal: Physical Review Physics Education Research, 18 010110 (2022)
Most introductory STEM courses require students to learn large amounts of complex information at a relatively high rate, often with a few sections of a textbook chapter a class period. At the same time, instructors are often pressured to cover a wide variety of content, stuck in a balancing act between introducing all the context and ensuring students learn it. Yet, there’s only so much information the brain can process at a time.
Unsurprisingly, some efforts to help students learn focus on making it easier to understand the content by breaking it into smaller, digestible chunks, a strategy appropriately called “chunking.”
One way to do so is to create short summaries of the key points of a lecture or lesson. While prior work has shown that both student- and instructor-generated summaries are effective for improving understanding of concepts, instructor-generated summaries were better for transfer of knowledge tests.
There’s no reason that these summaries have to be written however. Multimedia summaries could be just as effective. Today’s study puts those ideas to the test, finding that under the right conditions, 1-minute multimedia summaries of key ideas can help students learn the material and organize their knowledge.
To conduct their study, the authors selected students from a calculus-based, second semester introductory physics course at Harvard University, covering electricity and magnetism aimed at life science and pre-med students. The course was taught in an active-learning style with a mix of group work, instructor feedback, and discussion.
The idea of 1-minute video summaries had been previously introduced in the first semester course 5 years prior to the study and they had been introduced to the second semester course 3 years before the study. Most of the major topics in the course had a video about them, and most students reported looking at at least one video during the semester.
To conduct their study and limit the effects of previous lessons, the authors chose to study what students learned from a video about thin optical lenses, as understanding thin optical lenses requires limited knowledge of previous topics in the course. The study was also conducted after the material was covered in class but before the videos were available to the class. An example of what the video contained is shown below in Figure 1.
All the students in the course were invited to participate in the study but only 148 or about two-thirds of the class did. The researchers split the students into three groups. All groups received annotated lecture and discussion notes about the topic. However, group A was also provided access to the 1-minute video about thin optical lenses, group B was also provided with access to the video in addition to a short written summary of the key points, and group C was not given any additional resources. Based on the similarity of exam scores on the two midterms already given in class between the groups, the researchers assumed the groups had similar levels of physics preparation and ability.
The students were then given 20 minutes to review the materials they had been provided. Once those 20 minutes had ended, the researchers gave students 7 minutes of instructions and examples of how to create a concept map, 18 minutes to create the concept map, and 20 minutes to complete a test of learning. The process is summarized in figure 2.
After all the students were finished, two reviewers graded the concept maps on a scale of 3-9 in three areas, content knowledge, relational knowledge, and accuracy, each graded on a scale of 1-3.
The tests were graded on a scale of 0-30 with partial credit on items possible.
When looking at the results, the researchers found that the videos did help students learn the concepts better. Looking at the concept map results, students in group A and B did better than students in group C. Students in group A earned a median score of 8.25 out of 9, students in group B earned a median score of 8, and students in group C earned a median score of only 6.5. The differences were statistically significant.
Likewise, the test of learning showed a similar trend. Students in group A earned a median of 20.75 out of 30, students in group B earned a median score of 23, and students in group C earned a median score of 9. Again, the differences were statistically significant. The results are summarized in Figure 3.
While groups A and B outperformed group C, the differences between groups A and B were not statistically significant, suggesting that the text summary didn’t add anything additional compared to the videos.
While the study was only conducted on a single lesson, the results show promise that short video summaries after class can improve student learning and their organizing of the material. However, the researchers emphasize that the summaries should be provided after lecture because that is when students are trying to make sense of what they’ve learned. Course surveys tended to agree with the authors in that most students who watched the videos reported doing so after the lecture but before starting homework.
For instructors who wish to develop videos for their own class, the authors provided a list of six recommendations. First, instructors should break down the big ideas into core concepts and provide connections between them.
Second, instructors should release the videos when students are most overwhelmed by the content, meaning after lecture.
Third, the videos should be short so the amount of time spent on them is negligible compared to the rest of the course, hence the authors’ use of single minute videos.
Fourth, instructors should take the perspective of a student when making the video rather than providing the expert’s perspective. That can include consulting with previous students to learn how they approached the problems and what the common mistakes will be.
Fifth, instructors should limit the distractions in the videos and focus on the material rather than cool visuals or sound effects.
Finally, instructors should use contrasting cases to illustrate the key points. This means varying the context so students can understand how different properties of the problem affect the steps of the solution, both in terms of what stays the same and what changes. For example, in pushing a box up or down a ramp, the gravitational force always points down the ramp but the direction of friction depends on which way the box is pushed.
Figures used under CC BY 4.0
I am a postdoc in education data science at the University of Michigan and the founder of PERbites. I’m interested in applying data science techniques to analyze educational datasets and improve higher education for all students