Earlier this week I wrote that multimodal learning can work. And for learning a foreign language, too, we have known for some time that it can help not only to hear the word and the translation, but also, for example, to see an image or make a gesture. But what works better? Is it primarily the combination of different information sources, or does the movement itself offer an additional benefit?

A new study by Paula Johnstone and colleagues in Learning and Instruction examined these questions. The researchers also wanted to know whether previous positive results for gestures hold up when students practice less intensively, for example. After all, much earlier research used dozens of repetitions per word, which is not always realistic in an educational context, as many language teachers will recognise…

The study included 52 German-speaking students from teacher training programs. They learned 24 words from an artificial language: twelve concrete and twelve abstract words. They covered each word ten times over five days.

The students were taught the words in three ways. In the first condition, the participants heard only the foreign word and the German translation. In the second condition, students received the words with a relevant image. And in the third condition, they saw a relevant gesture and performed it themselves. The words presented in each condition varied among students.

Next, they were tested on how much they still remembered after eight days, two months, and six months. The researchers used three tests: free recall, translation from the native language to the new language, and vice versa.

The first finding is clear: both images and gestures helped with memorisation. Interestingly, this applied primarily to abstract words. For concrete words, the researchers found no significant differences between the three conditions.

The effects for abstract words were also not negligible. For example, with free recall, the difference compared to listening only was d = .50 for gestures and d = .60 for images. The researchers also found similar benefits in the translation tests.

But perhaps even more interesting: making gestures oneself yielded no additional benefit compared to viewing an image. That difference was not significant at any of the three measurement points.

I initially thought 52 participants were too small a sample to detect such differences. But no, that does not seem to be the most likely explanation here. First, every participant went through all the conditions, allowing participants to serve as their own controls. Moreover, the researchers had calculated in advance that 52 participants would provide approximately 81.5% power to detect an effect of approximately d = .33.

More importantly, they conducted additional Bayesian analyses. These provided strong evidence that the difference between gestures and images was smaller than the practically relevant difference the researchers had previously determined. Therefore, this is not simply a matter of the absence of a statistically significant result.

Does this mean gestures cannot add value? No. Earlier research in which students practised words much more frequently sometimes found gestures to be more advantageous than images. It is therefore possible that the specific motor added value only emerges after much more practice and repetition.

The study naturally has limitations. You could already deduce several from what I described above. To list them:

• The participants were 52 young adult students,
• They learned artificial words
• Not yet mentioned: everything was presented auditorily and tested orally.

For this reason, the authors warn that we do not know whether the results apply to written vocabulary or other learning contexts.

However, I do think this research offers an interesting nuance. When practice is limited, multimodal support seems more important than gestures for learning foreign vocabulary. This means a relevant image can be just as effective as performing a gesture. I also find this a useful insight for classroom practice. Showing an image generally seems much simpler to me than having an entire class repeatedly perform gestures.