One of his studies is an investigation of Japanese learners' vocabulary notebooks, which shows that "these learners have difficulty identifying high frequency words... and view all words they do not know as equally important." But in fact some words are quite important (will appear again often) and some are not (will appear rarely, and in situations where contextual interpretation is adequate). So learners need their teachers' help to know which words are worth writing in their notebooks or otherwise focusing on.But do teachers necessarily know this themselves?
Another of McCrostie's studies shows that "native speaker intuitions of word frequency appear limited to differentiating between very frequent and very infrequent words, with [ESL] teachers performing no better than first-year university undergraduates."
The problem is that there are important needs distinctions, not just between the high and low frequency zones, but also in the medium frequency zone. For example, take imperialism and imperious (from the Rex Murphy text on the VP-BNC/COCA page). Neither word is particularly uncommon to an educated English speaker. However, imperialism is a 5k word (probably worth learning for an academic learner) while imperious is a 10k word and could not be defined out of context by many educated native speakers.
What can be done to make teachers better helpers in this area? The Freq Trainer program is one idea for raising broad-category frequency awareness. Stay tuned for some research results.
What is "good performance" on the Freq Trainer quiz? Being frequency aware is partly about having explicit knowledge of word frequencies, as will show up the first time "Check" is clicked. But being aware is also about the ability to tap implicit frequency knowledge through thoughtful second guesses. For example, if cripple is first classed 6-10k, but this is wrong, then should it go up (to 11-15k) or down (to 3-5k or even 1-2k)? (The correct answer is down, to 3-5k.)
As the designer of this game, its first user, and a researcher who works with frequency a lot, I normally score 70% on the first click and then can wrap up in three tries (regardless of quiz size).
References
McCrostie, J. (2007a). Examining learner vocabulary notebooks. ELT Journal 61(3):246-255.
McCrostie, J. (2007b). Investigating the accuracy of teachers word frequency intuitions. RELC Journal 38(1): 53-66.