2.3. Test conditions

Masked SRTs, defined as the SNR between the target and both masker sentences that produces 50% correct recognition of target keywords in sentences, were adaptively measured using a procedure similar to the coordinate response measure test from Brungart et al. (2001). Two target keywords (randomly selected from the number and color categories) were embedded in a five-word carrier sentence uttered by the male target talker. The name “John” was used to cue listeners to the target sentence. All words in the target and masker sentences were randomly selected to be different (mutually exclusive). During testing, listeners were asked to identify the number and color keywords from a closed set of 10 responses; the responses for the name, verb, and object categories were greyed out. Recognition of the target keywords was measured in the presence of two masker sentences (two-talker masker). One masker sentence was produced by a different male talker and by a female talker. The relative levels of the male and female masker speech streams were systematically changed from −9 to +9 dB in 3 dB steps (i.e., −9, −6, −3, 0, 3, 6, and 9 dB). Masked SRTs were also measured in the presence of one male talker and one female talker alone (one-talker masker) as a control condition.