The meaning inference has been found to be an effective tool to promote the learners’ comprehension and vocabulary learning while reading a text. Moreover, such promotion has been facilitated by using hypermedia glosses that provided some information about the word in different multimedia formats. The present study embedded the clues for meaning inference in hypermedia gloss and called it the contextualized hypermedia gloss. Thus, the purpose was to study the effectiveness of the contextualized hypermedia glosses on reading comprehension, vocabulary recall, and recognition. To do this, the glosses were delivered via Telegram application to 76 intermediate students in the University of Mazandaran. They were assigned into four groups receiving three formats of contextualized gloss: text, picture, and picture and text, although the control group received noncontextualized gloss in textual dictionary definition format. The results of the comparisons between groups indicated that providing direct access to the meaning was significantly better for immediate vocabulary recall (p < 0.05) than the contextualized pictorial group, providing the indirect access to meaning. Moreover, the comparison of test scores in the four groups showed no significant difference in delayed recognition and reading comprehension tests.
