In many English reading tests, hard words have short notes at the bottom of the page. These notes are called glosses. Do they help us remember words?

Researchers studied 42 studies. Readers with glosses learned about 45 percent of the new words. Readers without glosses learned only about 27 percent. Glosses with pictures or sound were not better than glosses with only words. So a short, simple note is enough.

A gloss is a short explanation of a difficult word, usually placed in the margin or at the bottom of a page. A 2020 meta-analysis by Yanagisawa, Webb and Uchihara combined 42 studies with 3,802 learners. People who read glossed texts learned 45.3 percent of the target words on immediate tests and 33.4 percent on delayed tests, compared with 26.6 and 19.8 percent for people who read without glosses.

Multiple-choice glosses, which ask readers to choose the correct meaning, worked best, and glosses in the learners' first language worked better than English ones. But adding pictures or audio did not make a significant difference. Even with glosses, however, many words were forgotten later, so learners need to meet them again.

Glosses, the brief definitions or translations printed beside difficult words, are so familiar in reading tests that few learners think about them. Applied linguists, however, have studied them for decades. In 2020, Yanagisawa, Webb and Uchihara published a meta-regression of 42 studies involving 3,802 participants.

Learners who read glossed texts recalled or recognized 45.3 percent of target words immediately and 33.4 percent at delayed testing, compared with 26.6 and 19.8 percent for unglossed reading, roughly a 1.7-fold advantage. Multiple-choice glosses, which require readers to evaluate options, were the most effective, whereas in-text glosses and end-of-text glossaries were the least effective.

First-language glosses outperformed second-language glosses, but textual, pictorial and auditory glosses did not differ significantly. A separate meta-analysis found that adding a second mode to a textual gloss helped, but adding a third did not. Glosses also support comprehension, and an earlier meta-analysis reported a medium effect on reading comprehension. The findings are not entirely consistent, though. Another meta-analysis found that the advantage of first-language glosses was clearest for beginners and on immediate vocabulary tests. In practice, simple, well-placed glosses combined with later review seem most sensible.