According to Tim Ferris,

If you train for recall, you get recognition automatically; if you train for recognition, recall is terrible, or as slow as molasses. (See The Time Ferris Show, 19 June 2014).

Even though it sounds intuitively plausible, is there actually any research to back up the statement that learning vocabulary for recall gives you recognition "for free"?

The question is relevant to the creation of flashcards (on paper or in an SRS): is creating flashcards for recall sufficient, or do you also need flashcards for recognition?

Since this questions asks for studies to support the answers, see the Resources for Researching Language Learning Questions.

  • It's not free because learning for recall is harder than recognition x2 or whatever.
    – SAH
    Dec 15, 2016 at 7:26
  • @SAH I know what you mean, but "for free" here means that you don't need to add recognition study when you're already doing recall study.
    – Tsundoku
    Dec 15, 2016 at 10:36
  • 1
    For people (like me) who are not 100% sure how recall vs recognition works, and what is the difference: interesting article about Memory Recognition and Recall in User Interfaces Jan 2, 2018 at 19:14

1 Answer 1


What is more, Glover’s (1989) research suggests that recall is more effective in learning than recognition, and that multiple-choice in web-based learning tends to reinforce recognition, not recall. According to constructivist assumption stressing the active construction of knowledge and situational activities (see Han, 1990), recall is a much more important intellectual activity than recognition because recall makes retrieving schemata an easier process for the learner. This coincides with the context where students construct their knowledge through the representation of specific knowledge internally and the interpretation of personal experiences. In the same context, recall plays a greater role in transferring knowledge than recognition, because recognition is a process of simply retrieving what the learner has obtained previously or what the learner has memorized (Han, 1990).

Viewed from the fact that recall and recognition are two different cognitive processes and that recall is closely related to the transfer of knowledge, it is unfavorable to adopt a multiple response mode that requires recognition in web-based learning. Graff (2003) suggests that introducing design principles that incorporate the learners’ cognitive style (i.e., holistic and analytic) with content might be instrumental in developing an effective instructional program. Thus, what is needed at this time is to test response modes that support active intellectual recall and the transfer of knowledge matching to the context. More specifically, it is the explicit response presented at the time when feedback is given that requires greater concern.

From "Improving Recall and Transfer Skills Through Vocabulary Building in WebBased Second Language Learning: An Examination by Item and Feedback Type" by Yun et al, Educational Technology & Society, 11.4 (2008), 158–172.

  • 1
    Would you be willing to elaborate on how this quote answers the question, preferably in your own words?
    – Hatchet
    Nov 16, 2016 at 1:23
  • @Hatchet I added emphasis.
    – fi12
    Nov 16, 2016 at 2:51
  • Could you add the full reference to the paper (authors, journal, year)? And possibly the full references for the papers cited in the quoted excerpt?
    – Tsundoku
    Nov 16, 2016 at 12:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.