Learning through visuals

Including visuals in lessons helps students better explain what they learn, but doesn’t necessarily lead to stronger test performance.

Two recent papers authored or co-authored by Logan Fiorella, associate professor in the Mary Frances Early College of Education’s Department of Educational Psychology, explore the effects of provided visuals for students on their explanations of concepts.

One study illustrates how two different types of visuals, concept maps and images, serve as a scaffold, while another shows the possibility for visuals to serve as a crutch.

Explanations through relationships

In one paper, “Effects of instructor-provided visuals on learner- generated explanations,” Fiorella and educational psychology alumna Shelbi Kuhlmann (M.A. ’18, Ph.D. ’20) compare different types of provided visuals and the types of explanations students generate from them.

Students were divided into three groups and learned a lesson on the nervous system with different provided visuals. One group received two concept maps, the second received two images, and the third did not receive a visual.

The study looked at three ways information is conveyed: hierarchical relationships, or how concepts are contained within broader concepts; structural relationships, the physical appearances of objects and their position in space; and temporal relationships, how events occur over time.

Kuhlmann and Fiorella found that the different visuals led to differences in the types of relationships students made in their explanations. The maps group created more hierarchical relationships, while both the maps and images groups generated more temporal relationships compared to the text-only group.

Knowing what kinds of visuals best facilitate different relationships can help instructors decide what visuals to include in lessons.

“Different types of visuals emphasize different types of relationships in your explanations, so it has implications, not just that visuals can help you explain, but which kinds of visuals and how you should design those visuals to support different types of relationships,” he said.

Scaffold vs. crutch

Another study authored by Fiorella, “Learning by explaining after pauses in video lectures: Are provided visuals a scaffold or a crutch?” finds potential pitfalls with the role of visuals on test performance.

Here, students were divided into three groups and shown a five-part video about the human kidney. The video was paused between each part, and one group was provided with both the screenshot and the transcript from that portion of the video to study. The other two groups were prompted to write an explanation for provided questions between each section: one group was provided a screenshot from the video as a visual aid, while the other was not.

“The group that got the visuals did explain better than the explain-only condition without a visual,” Fiorella said. “They generated better explanations during learning when the visuals were there.”

On a post-test without access to the visuals, both explain groups performed better than the group who studied the lesson without explaining. Although those who explained with access to visuals generated more complex explanations, they did not perform significantly better on the test than the group who explained without visuals.

This difference suggests that the visuals may have served as a crutch for students that did not necessarily help them internalize and retain the information in the long run.

Generating stronger explanations

Despite their differences, both studies demonstrate that students benefit from explaining what they’ve learned.

“Both are finding that explaining is beneficial, and both are finding that visuals can help you explain more effectively,” Fiorella said.

Fiorella said prior studies show the benefits of students generating their own visuals and then receiving external visuals as feedback. A balance of provided and student-generated visuals may help assess how visuals help students learn and retain information long-term.

“We need to help students internalize the visualization and not just rely on the fact that there’s an external visual provided to them,” Fiorella said. “So, it might serve as a scaffold if you have it there at first, but then remove it and make sure that students are not relying on it and they can independently generate the same types of inferences.”