Results based on pre-and post- auditory tests (to see if subjects could hear the long-short distinction) turned out to be a mixed bag: "The overall effect of hand gesture on learning of segmental phonology is limited."
In some contexts it seemed to work: " . . .observing the syllabic-rhythm hand gesture (of the instructor) yielded the most balanced improvement between word-initial and word- final vowels and between slow and fast speaking rates." What did not seem to work as well (or at all) was when subjects just " . . . produced the moraic*-rhythm gesture (along) with the instructor."
Library of Congress
An earlier blog post looked at a number reasons why kinaesthetic-only techniques (those that are not haptic) may not work--that is using a gesture, like the "waving hand" in this study. Probably the most important factor is the potentially unsystematic use of the gesture, especially for highly visual and emotionally "conservative" learners.
That was an important early discovery in our haptic work, which involves anchoring all gestures with touch in various ways on the stressed syllable of a word. Any number of students told us unequivocally that unless (a) the gesture moved through something close to the precise, same track in the visual field and (b) "felt the same" in their bodies each time that it was used by the instructor or themselves, they found the procedure at best irrelevant, at worst very disconcerting.
The Hirata et al (2014) study not only used "unanchored" gesture, it used the same long or short gesture(s) for signalling length, regardless of the vowel. That is not unlike having students stretch a rubber band on long vowels (Gilbert, 2012), a technique that gets across the concept of vowel length very well but probably does little to transfer that idea into ongoing production in speech.
With apologies to G.K. Chesterton: (Unanchored) kinaesthetic teaching of pronunciation has not been tried and found wanting here; it has, not surprisingly, been found inconsistent and unsystematic. But a touch of "haptic" might have made a very significant impact. Keep in touch.
*For more on the concept of mora and how it affects syllable length, see the succinct wikipedia note.
Hirata, Y., Kelly, S., Huang, J., Manansalaa, M. (2014). Effects of Hand Gestures on Auditory Learning of Second-Language Vowel Length Contrasts, Journal of Speech, Language, and Hearing Research 57: 2090–2101, December 2014.