In most speech intelligibility studies that use open response sets, subjects listen only once to a particular stimulus. During daily communication, the utterance and the concurrent noise are rarely fixed hence presenting a stimulus only once seems a highly relevant approach to speech perception. However in forensic applications one tends to have recordings of the distorted speech, hence the possibility for replay. I am wondering whether anyone ever looked at the effects of replay on speech intelligibility. Does it help to replay the same speech fragment over and over again? If yes, when does this effect saturate? Are listeners able to judge when additional replay no longer improves intelligibility, or do they merely become more convinced in their (possibly erroneous) responses?
Your ideas, remarks and in particular references are warmly appreciated.