Shadowing means listening to English and repeating it right after the speaker, like a shadow. Many books say it is the best way to learn English. In a Japanese study, 43 university students practiced shadowing in nine lessons. All of them became better at catching English sounds.

But only the lower-level students got better scores on a listening test. So shadowing is good training for your ears, especially if sounds are hard for you.

Shadowing, repeating what you hear a moment after the speaker, began as a training method for simultaneous interpreters and became very popular in Japan. But what does research show? In a study published in 2016, Hamada taught shadowing to 43 Japanese university students over nine lessons.

Before and after the course, the students took a test of catching individual words and a high-school-level listening test. Both lower- and intermediate-level groups improved at catching sounds, but only the lower-level group improved on the listening test. Shadowing seems to train bottom-up processing, the skill of picking out words from connected speech. Evidence that it improves speaking is more limited, and many studies are small, short and have no comparison group. Learners who already hear sounds well may need other practice.

Shadowing, in which learners repeat speech a fraction of a second behind the speaker, originated in interpreter training and has become one of the most widely promoted techniques in Japanese English education. The research suggests it is a specialized tool rather than a cure-all.

In a study published in 2016, Hamada taught shadowing to 43 university students across nine textbook-based lessons and compared pre- and post-test results on a cloze dictation task, used as a measure of phoneme perception, and a high-school-level listening comprehension test. Both lower- and intermediate-level groups improved in phoneme perception, yet only the lower-level group made significant gains in comprehension.

The likely explanation is that shadowing forces attention to every sound, training bottom-up processing, which benefits learners whose main difficulty lies in decoding connected speech. This fits a 2021 meta-analysis showing moderate effects for decoding training in general. Several caveats apply, however. The study had no control group, its measure of phoneme perception is open to debate, and research on pronunciation or speaking gains has produced mixed results. Most studies involve small groups of Asian university students over a few weeks. Learners should therefore first identify whether sounds or vocabulary are their real obstacle.