Mean SRTs with standard errors across NH (left panel) and HI (right panel) listeners measured by Rana and Buchholz (2018) [57] as a function of masker sensation level (dB SL, using the individual SRT in quiet as reference). The two vocoded-speech maskers were either co-located with the speech target in front of the listener (“co-loc”, blue) or simulated on each side of the listener (“separ”, red). Predicted SRTs and performance statistics are displayed for the model vicente2020.