Auditory intensity perception: successive versus simultaneous, across-channel discriminations

Separating the contributions of primary and unwanted cues in psychophysical studies

A fundamental issue in the design and the interpretation of experimental studies of perception relates to the question of whether the participants in these experiments could perform the perceptual task assigned to them using another feature, or cue, than that intended by the experimenter. An approach frequently used by auditory- and visual-perception researchers to guard against this possibility involves applying random variations to the stimuli across presentations or trials so as to make the "unwanted" cue unreliable for the participants. However, the theoretical basis of this widespread practice is not well developed. In this article, we describe a 2-channel model based on general principles of psychophysical signal detection theory, which can be used to assess the respective contributions of the unwanted cue and of the primary cue to performance or thresholds measured in perceptual discrimination experiments involving stimulus randomization. Example applications of the model to the analysis of results obtained in representative studies from the auditory- and visual-perception literature are provided. In several cases, the results of the model-based analyses indicate that the effectiveness of the randomization procedure was less than originally assumed by the authors of these studies. These findings underscore the importance of quantifying the potential influence of unwanted cues on the results of psychophysical experiments, even when stimulus randomization is used.

Psychometric functions for pure-tone frequency discrimination

The form of the psychometric function (PF) for auditory frequency discrimination is of theoretical interest and practical importance. In this study, PFs for pure-tone frequency discrimination were measured for several standard frequencies (200-8000 Hz) and levels [35-85 dB sound pressure level (SPL)] in normal-hearing listeners. The proportion-correct data were fitted using a cumulative-Gaussian function of the sensitivity index, d', computed as a power transformation of the frequency difference, Δf. The exponent of the power function corresponded to the slope of the PF on log(d')-log(Δf) coordinates. The influence of attentional lapses on PF-slope estimates was investigated. When attentional lapses were not taken into account, the estimated PF slopes on log(d')-log(Δf) coordinates were found to be significantly lower than 1, suggesting a nonlinear relationship between d' and Δf. However, when lapse rate was included as a free parameter in the fits, PF slopes were found not to differ significantly from 1, consistent with a linear relationship between d' and Δf. This was the case across the wide ranges of frequencies and levels tested in this study. Therefore, spectral and temporal models of frequency discrimination must account for a linear relationship between d' and Δf across a wide range of frequencies and levels.

Psychophysical reverse correlation with multiple response alternatives

Psychophysical reverse-correlation methods such as the "classification image" technique provide a unique tool to uncover the internal representations and decision strategies of individual participants in perceptual tasks. Over the past 30 years, these techniques have gained increasing popularity among both visual and auditory psychophysicists. However, thus far, principled applications of the psychophysical reverse-correlation approach have been almost exclusively limited to two-alternative decision (detection or discrimination) tasks. Whether and how reverse-correlation methods can be applied to uncover perceptual templates and decision strategies in situations involving more than just two response alternatives remain largely unclear. Here, the authors consider the problem of estimating perceptual templates and decision strategies in stimulus identification tasks with multiple response alternatives. They describe a modified correlational approach, which can be used to solve this problem. The approach is evaluated under a variety of simulated conditions, including different ratios of internal-to-external noise, different degrees of correlations between the sensory observations, and various statistical distributions of stimulus perturbations. The results indicate that the proposed approach is reasonably robust, suggesting that it could be used in future empirical studies.

On the choice of adequate randomization ranges for limiting the use of unwanted cues in same-different, dual-pair, and oddity tasks

A major concern when designing a psychophysical experiment is that participants may use a stimulus feature (cue) other than that intended by the experimenter. One way to avoid this problem is to apply random variations to the corresponding feature across stimulus presentations to make the unwanted cue unreliable. An important question facing experimenters who use this randomization (roving) technique is how large the randomization range should be to ensure that the participants cannot achieve a certain proportion correct by using the unwanted cue, while at the same time avoiding unnecessary interference of the randomization with task performance. Previous researchers have provided formulas for the selection of adequate randomization ranges in yes-no and multiple-alternative forced choice tasks. In this article, we provide figures and tables that can be used to select randomization ranges that are better suited to experiments involving a same-different, dual-pair, or oddity task.

Limiting unwanted cues via random rove applied to the yes-no and multiple-alternative forced choice paradigms

When a random rove is used in a perceptual task to control the influence of an unwanted cue that may confound the decision strategy of primary interest, the effectiveness of the rove is determined by its range. Green [Profile Analysis (Oxford University Press, Oxford, 1988)] provided a formula which allows experimenters to determine the roving range required to ensure that the listeners relying on the unwanted cue cannot exceed a pre-defined percentage of correct responses in a two-interval, two-alternative forced-choice experiment. Here, Green's analysis is extended to the yes-no and m-alternative, forced-choice paradigms (m>2).

Broadband auditory stream segregation by hearing-impaired and normal-hearing listeners

PURPOSE

To investigate the effects of hearing loss on auditory stream segregation of broadband inharmonic sounds.

METHOD

Auditory stream segregation by listeners with normal and impaired hearing was measured for 6-component inharmonic sounds ("A" and "B") using objective and subjective methods. Components in the A stimuli ranged between 1000 and 4000 Hz, whereas B stimuli were generated at the same frequency ratio but scaled upward in frequency relative to the A stimuli. In Experiment 1, streaming was measured by having listeners detect a delay inserted into a sequence of A and B stimuli (A_B_A_B_...) for B stimuli with different frequencies. In Experiment 2, streaming was measured using an ABA_ABA_... sequence, and the frequency of the B stimulus decreased until listeners reported that they could "no longer hear two separate streams."

RESULTS

Experiment 1 indicated no significant differences between groups in the size of the just detectable delay and no significant interactions between group and the scaling factor between the B and A stimuli. Experiment 2 revealed no significant differences in streaming abilities between normal-hearing and hearing-impaired groups.

CONCLUSIONS

Overall, results indicate that listeners with normal and impaired hearing have similar auditory streaming abilities for broadband inharmonic complex stimuli.

Likelihood ratio, optimal decision rules, and correct response probabilities in a signal detection theoretic, equal-variance Gaussian model of the observer in the 4IAX paradigm

This article provides a synthetic account of the likelihood ratio, optimal decision rules, and correct response probabilities in a signal detection theoretic model of the observer in the dual-pair comparison, or four-interval AX (4IAX), paradigm. The model assumes a static sampling process, resulting in four, equal-variance normally distributed (i.e., Gaussian) observations on each trial First, a likelihood ratio equation allowing for an arbitrary degree of correlation between observations is provided. Specific solutions for the cases of independent and highly correlated observations are then derived. It is shown that these solutions, and the associated decision rules, correspond to those provided independently in earlier publications. A modified 4IAX paradigm involving, as a standard, an additional stimulus (C) located medially between the A and the B stimuli is also considered. It is shown that the optimal (static, equal-variance, Gaussian) decision model for this paradigm is unaffected by correlation between observations and is equivalent to the standard 4IAX with highly correlated observations. Finally, we consider how, under the considered (static, equal-variance, Gaussian) model, the proportion of correct responses in the different versions of the 4IAX paradigm is related to d', and a solution for the case of independent observations is provided.

Comparing F0 discrimination in sequential and simultaneous conditions

In an influential study, Carlyon and Shackleton [J. Acoust. Soc. Am. 95, 3541-3554 (1994)] measured listeners' performance (d') in fundamental-frequency (F0) discrimination between harmonic complex tones (HCTs) presented simultaneously in different spectral regions and compared their performance with that found in a sequential-comparison task. In this Letter, it is suggested that Carlyon and Shackleton's analysis of the simultaneous-comparison data did not adequately reflect their assumption that listeners were effectively comparing F0's across regions. A reanalysis consistent with this assumption is described. The new results suggest that under the assumption that listeners were effectively comparing F0 across regions, their performance in this task was substantially higher than originally estimated by Carlyon and Shackleton, and in some conditions much higher than expected from the performances measured in a traditional F0-discrimination task with sequential HCTs. Possible explanations for this outcome, as well as alternative interpretations, are proposed.

The effect of onset asynchrony on profile analysis by normal-hearing and hearing-impaired listeners

The effect of onset asynchrony on discrimination of spectral shape was evaluated for hearing-impaired and normal-hearing listeners. Stimuli were the sum of four tones equally spaced on a logarithmic frequency scale. The standard stimulus had tones of equal amplitude, and the signal stimulus had two adjacent components increased in level, and the other two components decreased in level. Thresholds for discrimination between the standard and signal stimuli were measured as a function of an onset asynchrony among the components of 0, 50, and 200 ms. Hearing-impaired and normal-hearing listeners had similar thresholds when the stimulus components were widely spaced in frequency, but hearing-impaired listeners had much higher thresholds for narrowly spaced components. Excitation pattern analyses indicated that listeners may use spectral peaks in the stimulus rather than the change in excitation across the full stimulus bandwidth for spectral shape discrimination tasks. Increasing temporal asynchrony of components resulted in increased thresholds for both groups of listeners to a greater extent in the wide span than the narrow span. Reduced effects of onset asynchrony in the narrow span suggest that spectral resolvability of components plays an important role in the processing of onset asynchrony across frequency.

A comparison of psychophysical procedures for level-discrimination thresholds

Five different psychophysical procedures were used to measure level-discrimination (also called intensity discrimination) thresholds for 1-kHz tones at two levels (30 and 90 dB SPL) and two durations (10 and 500 ms). The procedures were the classic transformed up-down staircase method with a two-alternative forced-choice (2AFC) paradigm (UPD), 15- and 50-trial implementations of the method of maximum likelihood (MML) with a cued yes-no paradigm, and 18-trial implementations of ZEST using both cued yes-no and 2AFC paradigms. Results obtained from nine normal listeners show that estimates of level-discrimination thresholds for the four conditions are similar across all five procedures when different points of convergence are accounted for. The variance of threshold estimates within listener and condition was smallest for UPD, largest for the MML with 15 trials, and statistically indistinguishable among the others. The sweat factors ranged from 5.5 for MML with 50 trials to about 1.4 for UPD and ZEST. Simulations show that ideal performance of procedures may be far from real-life experience and that these deviations are likely to depend on complex interactions between listener behavior and parameter choices used for implementing the procedures. Therefore, empirical verification is important for judging the effectiveness of psychophysical procedures.

A comparison of the effects of differences in temporal gating and ear of presentation on profile discrimination

The aim of the study was to examine the effects of differences in temporal gating and ear of presentation (both separately and in combination) on listeners' ability to detect an increment in the level of a 1 kHz component (the target) relative to that of four spectrally flanking components. The flanking components were always presented to the listeners' right ear, while the target component was either presented to the same ear (monaural presentation) or to the left ear (dichotic presentation). Similarly, the target and flanking components were either gated on and off at the same time (synchronous presentation), or else the target component began 100 ms before and terminated 100 ms after the four flanking components (asynchronous presentation). On average, thresholds were lowest in the synchronous, monaural condition, and highest in the two asynchronous conditions. Ear differences alone did result in elevated thresholds for most listeners. However, combining differences in gating and ear of presentation produced thresholds that were indistinguishable from those obtained when gating differences alone were employed. These results are consistent with the suggestion that differences in temporal gating lead to more complete segregation of concurrent frequency components than differences in spatial location.

A brief delay decreases perceived similarity and improves discrimination

A same-different discrimination between the colors of 2 halves of a circle was poorer when the halves were viewed simultaneously than when the delay between viewing them was 33 ms. The simultaneous discrimination also produced a poor performance on different stimuli relative to same stimuli in comparison to the successive discrimination. Poor-on-different performance should indicate high perceived similarity. The simultaneous colors were also judged as more similar. The delay did not alter the physical similarity of the colors. Thus, the delay may have decreased assimilation (perceived similarity) and thereby improved discrimination, in support of the hypothesis that perceived similarity frequently affects discrimination.

Decision strategies of hearing-impaired listeners in spectral shape discrimination

The ability to discriminate between sounds with different spectral shapes was evaluated for normal-hearing and hearing-impaired listeners. Listeners detected a 920-Hz tone added in phase to a single component of a standard consisting of the sum of five tones spaced equally on a logarithmic frequency scale ranging from 200 to 4200 Hz. An overall level randomization of 10 dB was either present or absent. In one subset of conditions, the no-perturbation conditions, the standard stimulus was the sum of equal-amplitude tones. In the perturbation conditions, the amplitudes of the components within a stimulus were randomly altered on every presentation. For both perturbation and no-perturbation conditions, thresholds for the detection of the 920-Hz tone were measured to compare sensitivity to changes in spectral shape between normal-hearing and hearing-impaired listeners. To assess whether hearing-impaired listeners relied on different regions of the spectrum to discriminate between sounds, spectral weights were estimated from the perturbed standards by correlating the listener's responses with the level differences per component across two intervals of a two-alternative forced-choice task. Results showed that hearing-impaired and normal-hearing listeners had similar sensitivity to changes in spectral shape. On average, across-frequency correlation functions also were similar for both groups of listeners, suggesting that as long as all components are audible and well separated in frequency, hearing-impaired listeners can use information across frequency as well as normal-hearing listeners. Analysis of the individual data revealed, however, that normal-hearing listeners may be better able to adopt optimal weighting schemes. This conclusion is only tentative, as differences in internal noise may need to be considered to interpret the results obtained from weighting studies between normal-hearing and hearing-impaired listeners.

Adaptive procedures in psychophysical research

As research on sensation and perception has grown more sophisticated during the last century, new adaptive methodologies have been developed to increase efficiency and reliability of measurement. An experimental procedure is said to be adaptive if the physical characteristics of the stimuli on each trial are determined by the stimuli and responses that occurred in the previous trial or sequence of trials. In this paper, the general development of adaptive procedures is described, and three commonly used methods are reviewed. Typically, a threshold value is measured using these methods, and, in some cases, other characteristics of the psychometric function underlying perceptual performance, such as slope, may be developed. Results of simulations and experiments with human subjects are reviewed to evaluate the utility of these adaptive procedures and the special circumstances under which one might be superior to another.

The Weber fraction versus perceived similarity

The Weber fraction type of discrimination outcome means that stimuli of high magnitude are difficult to discriminate. Simultaneous and successive discriminations between two long (high magnitude) lines and also between two short (low magnitude) lines were carried out. Over-all performance on the successive discrimination between the two long lines was much poorer than on the three other discriminations. This discrimination also produced the poorest performance on different stimuli relative to same stimuli. The latter result should indicate that the two successive long lines were perceived as most similar. These and additional results support the conclusion that over-all discrimination performance and perceived similarity were consistently associated even though they were not consistently associated with length and time of occurrence (simultaneous versus successive). Nevertheless, the association of discrimination with perceived similarity has received minimal attention in the psychophysics area.

Different auditory filter bandwidth estimates based on profile analysis, notched noise, and hybrid tasks

Auditory filter bandwidths were estimated in three experiments. The first experiment was a profile-analysis experiment. The stimuli were composed of sinusoidal components ranging in frequency from 200 to 5000 Hz. The standard stimulus was the sum of equal-amplitude tones, and the signal stimulus had a power spectrum that varied up-down ... up-down. The number of components ranged from four to 60. Interval-by-interval level randomization prevented the change in level of a single component from reliably indicating the change from standard to signal. The second experiment was a notched-noise experiment in which the 1000-Hz tone to be detected was added to a noise with a notch arithmetically centered at 1000 Hz. Detection thresholds were estimated both in the presence of and in the absence of level randomization. In the third, hybrid, experiment a 1000-Hz tone was to be detected, and the masker was composed of equal-amplitude sinusoidal components ranging in frequency from 200 to 5000 Hz. For this experiment, thresholds were estimated both in the presence and absence of level variation. For both the notched-noise and hybrid experiments, only modest effects of level randomization were obtained. A variant of Durlach et al.'s channel model ["Towards a model for discrimination of broadband signals," J. Acoust. Soc. Am. 80, 63-72 (1986)] was used to estimate auditory filter bandwidths for all three experiments. When a two-parameter roex(p,r) filter weighting function was used to fit the data, bandwidth estimates were approximately two to three times as large for the two detection tasks than for the profile-analysis task.

Across-channel intensity discrimination in the presence of an interferer

Listeners' thresholds for discriminating changes in the relative intensities of two octave-spaced pure tones were measured in three conditions using a 2AFC procedure. In the baseline condition the tone pair was presented alone, while in the two interference conditions the tone pair was accompanied by four additional tones, with the frequency separation between adjacent components of the resultant six-component complex being one octave. In the interference conditions the flanking components were either gated synchronously with the target pair, or began 200 ms ahead. The level of the flanking components relative to the target was randomized on each presentation. The overall stimulus level was also randomized on each presentation to reduce the effectiveness of within-channel comparisons. Threshold elevations in the synchronous condition relative to the baseline ranged from approximately 5 to 17 decibels. By contrast, in the asynchronous condition only one listener showed significant (although substantially reduced) interference. The results complement previous observations that across-channel intensity comparisons are poorer between components that begin or end at different times, and are qualitatively consistent with the hypothesis that temporal misalignment promotes the perceptual segregation of simultaneous frequency components.

Relative intensity comparisons between a tone and spectrally remote noise: effects of onset asynchrony

The present experiment investigated the effect of onset asynchrony on listeners' ability to make relative intensity comparisons between a 1-kHz tone and a band of noise high-pass filtered at 3 kHz. In the synchronous condition, the tone and noise were gated on and off simultaneously. In the two asynchronous conditions, either the tone was gated on before the noise or vice versa, both stimuli terminating simultaneously. In the sequential condition, the offset of the tone coincided with the onset of the noise. The task of the six listeners was to indicate in which of the two presentation intervals the level of the tone was incremented relative to that of the noise. To deter the use of strategies based on successive, within-channel level comparisons, the overall level of the stimuli was randomized on each presentation. For all listeners thresholds were lowest in the synchronous condition, and highest in the sequential condition, the difference ranging from approximately 7 to 18 dB (signal re: pedestal amplitude). Furthermore, five of the six listeners had significantly lower thresholds in the noise-leading condition than in the tone-leading condition, the average difference across listeners being approximately 5 dB. The results are discussed in terms of auditory grouping and the possible strategies available to the listeners.

Profile analysis with an asynchronous target: evidence for auditory grouping

Green and Dai [in Auditory Physiology and Perception, edited by Y. Cazals, L. Demany, and K. Horner (Pergamon, Oxford, 1992)] reported a series of experiments which suggested that listeners' ability to perform simultaneous across-frequency comparisons of intensity is severely impaired when the target and flanking components begin or end at different times. The present experiment sought to replicate the effect of onset asynchrony and included an additional condition in which the leading portion, of the asynchronous target component was accompanied by a pair of "captor" tones. The intended purpose of the captor tones was to promote a perceptual organization in which the leading and synchronous portions of the asynchronous target component were grouped with different auditory objects. For all six listeners an asynchrony of 320 ms raised thresholds substantially relative to the synchronous onset condition, the magnitude of the increase ranging between 6 and 16 dB. By contrast, the mean elevation of threshold in the presence of the captor was only 3 dB, although for all listeners thresholds were still greater than for the synchronous onset condition. The results support the view that the deleterious effect of onset asynchrony on profile analysis performance is due to the operation of auditory grouping principles.

Detectability of duration and intensity increments in melody tones: a partial connection between music perception and performance

Two experiments demonstrate positional variation in the relative detectability of, respectively, local temporal and dynamic perturbations in an isochronous and isodynamic sequence of melody tones, played on a computer-controlled piano. This variation may reflect listeners' expectations of expressive performance microstructure (the top-down hypothesis), or it may be due to psychoacoustic (pitch-related) stimulus factors (the bottom-up hypothesis). Percent correct scores for increments in tone duration correlated significantly with the average timing profile of pianists' expressive performances of the music, as predicted specifically by the top-down hypothesis. For intensity increments, the analogous perception-performance correlation was weak and the bottom-up factors of relative pitch height and/or direction of pitch change accounted for some of the perceptual variation. Subjects' musical training increased overall detection accuracy but did not affect the positional variation in accuracy scores in either experiment. These results are consistent with the top-down hypothesis for timing, but they favor the bottom-up hypothesis for dynamics. The perception-performance correlation for timing may also be viewed as being due to complex stimulus properties such as tonal motion and tension/relaxation that influence performers and listeners in similar ways.

A two-filter model for frequency discrimination

In this paper, we explore a two-filter model, the simplest version of multi-channel models for frequency discrimination of simple tones. According to this model, frequency discrimination is based on a change in the relative output levels of two auditory filters, one centered below and the other above the frequency of the tone. This idea can explain the experimental results that frequency discrimination is relatively unaffected by randomization of stimulus level. Moreover, it suggests a close relationship between the ability of listeners to perform frequency discrimination of simple tones and spectral-shape discrimination of two-tone complexes. The ability of three listeners to perform these two tasks was measured at six frequencies (from 0.25 to 8 kHz). The results from the spectral-shape-discrimination task were used to predict frequency-difference limens. There was a high correlation between obtained and predicted values.