Skip to main content
. 2015 Jun 8;129(4):473–490. doi: 10.1037/bne0000065

Figure 1.

Figure 1

Overview of behavioral methods. Ferrets were trained using a positive operant conditioning paradigm to report the perceived presence of a silent gap in a noise stimulus by nose-poking at one of two reward spouts at the periphery of a circular testing chamber (A). Trials were initiated by standing on a center platform, facing toward 0°, and nose-poking a central reward spout. During the training phase, “gap” stimuli were presented from a loudspeaker above the spout at −30°, whereas “no-gap”/continuous noise stimuli were presented from a loudspeaker located at +30°. During the testing phase, once the stimulus types (gap vs. no gap) were reliably paired with rewards at specific locations, all stimuli were presented from a loudspeaker at 0°, while responses were registered at water spouts positioned at ±30° (B). Example performance of one ferret during a session of the gap-detection task for the broadband noise stimulus (C). The percentage of correct responses for each gap length was used to construct psychometric functions, allowing estimation of the gap-detection threshold. Psychometric functions were transformed according to signal detection theory (Macmillan & Creelman, 2005) into sensitivity (d′), using the fitted psychometric functions, including empirical false alarm rates, on a session-by-session basis. Threshold was defined as the estimated gap length corresponding to a sensitivity of d′ = 1. The slope was computed as the gradient of the psychometric curve at threshold sensitivity (D). The asymptote of the curve was used to compute the maximum sensitivity, while the lapse rate was obtained experimentally. Example of gap-detection threshold progression as a function of session number for one animal (E). Data collected during training were not included in the analysis (T, dark gray block). Performance was stable over time, with the best three sessions extracted over the entire testing period (circles). Power spectra of the noise stimuli used in this experiment, comprising octave–bandpass narrowband noise (NBN) centered at 1 kHz (light gray), 4 kHz (midgray), or 16 kHz (dark gray) or flattened broadband noise (BBN) with a 30-kHz low-pass corner (black; F). Color scheme retained throughout figures.