Response Set Size Research Articles

The fluctuating masker benefit for normal-hearing and hearing-impaired listeners with equal audibility at a fixed signal-to-noise ratio.

Normal-hearing (NH) listeners can extract and integrate speech fragments from momentary dips in the level of a fluctuating masker, yielding a fluctuating-masker benefit (FMB) for speech understanding relative to a stationary-noise masker. Hearing-impaired (HI) listeners generally show less FMB, suggesting a dip-listening deficit attributable to suprathreshold spectral or temporal distortion. However, reduced FMB might instead result from different test signal-to-noise ratios (SNRs), reduced absolute audibility of otherwise unmasked speech segments, or age differences. This study examined the FMB for nine age-matched NH-HI listener pairs, while simultaneously equalizing audibility, SNR, and percentage-correct performance in stationary noise. Nonsense syllables were masked by stationary noise, 4- or 32-Hz sinusoidally amplitude-modulated noise (SAMN), or an opposite-gender interfering talker. Stationary-noise performance was equalized by adjusting the response-set size. Audibility was equalized by removing stimulus components falling below the HI absolute threshold. HI listeners showed a clear 4.5-dB reduction in FMB for 32-Hz SAMN, a similar FMB to NH listeners for 4-Hz SAMN, and a non-significant trend toward a 2-dB reduction in FMB for an interfering talker. These results suggest that HI listeners do not exhibit a general dip-listening deficit for all fluctuating maskers, but rather a specific temporal-resolution deficit affecting performance for high-rate modulated maskers.

Evaluative Processing of Food Images: Longer Viewing for Indecisive Preference Formation.

The well-known gaze cascade hypothesis proposes that as people look longer at an item, they tend to show an increased preference for it. However, using single food images as stimuli, we recently obtained results that clearly deviated from the general proposal that the gaze both expresses and influences preference formation. Instead, the pattern of data depended on the self-determination of exposure duration as well as the type of evaluation task. In order to disambiguate how the type of evaluation determines the relationship between viewing and liking we conducted the present follow-up study, with a fixed response set size as opposed to the varying set sizes in our previous study. In non-exclusive evaluation tasks, subjects were asked how much they liked individual food images. The recorded response was a number from 1 to 3. In exclusive evaluation tasks, subjects were asked for each individual food image to give one of three response options toward a limited selection: include it, exclude it, or defer the judgment. When subjects were able to determine the exposure duration, both the non-exclusive and exclusive evaluations produced inverted U-shaped trends such that the polar ends of the evaluation (the positive and negative extremes) were associated with relatively short viewing times, whereas the middle category had the longest viewing times. Thus, the data once again provided firm evidence against the notion that longer viewing facilitates preference formation. Moreover, the fact that non-exclusive and exclusive evaluation produced similar inverted U-shaped patterns suggests that the response set size is the critical factor that accounts for the observations here versus in our previous study. When keeping the response set size constant, with an equal opportunity to observe inverted U-shaped patterns, the findings are suggestive of a role for the level of decisiveness in determining the length of viewing time. For items that can be categorically identified as positive or negative, the evaluations are soon completed, with relatively brief viewing times. The prolonged visual inspection for the middle category may reflect doubt or uncertainty during the evaluative processing, possibly with an increased effort of information integration before reaching a conclusion.

The fluctuating-masker benefit for normal-hearing and hearing-impaired listeners with equal audibility at a fixed signal-to-noise ratio

While normal-hearing (NH) listeners demonstrate better speech intelligibility for fluctuating-masker than for stationary-noise conditions, hearing-impaired (HI) listeners generally show little or no fluctuating-masker benefit (FMB). This result has been interpreted in terms of suprathreshold deficits (e.g., reduced spectral or temporal resolution or distorted stream-segregation cues) that limit “dip-listening.” However, reduced FMB for HI listeners might instead be attributable to audibility limitations or to differences between the signal-to-noise ratios (SNRs) at which NH and HI listeners are tested. This study examined this issue by equalizing stationary-noise performance to allow measurements at a common SNR, equalizing audibility, and presenting identical signals to pairs of NH and HI listeners. Audibility was equalized using linear gain, low-pass filtering (4 kHz) and intensity filtering to remove speech-signal elements below the HI audiometric threshold. Nonsense-syllable identification performance in stationary noise was equalized by adjusting the response set size. Stationary-noise trials (adapting set size) were interleaved with fluctuating-masker trials (adapting SNR), ensuring stable stationary-noise performance throughout the test. Fluctuating maskers included low- and high-rate modulated noise, speech-modulated noise, and an interfering-talker condition. Results were assessed to determine whether and under which conditions the HI listeners demonstrated reduced FMB not attributable to SNR or audibility effects.

Set-size procedures for controlling variations in speech-reception performance with a fluctuating masker

Adaptive signal-to-noise ratio (SNR) tracking is often used to measure speech reception in noise. Because SNR varies with performance using this method, data interpretation can be confounded when measuring an SNR-dependent effect such as the fluctuating-masker benefit (FMB) (the intelligibility improvement afforded by brief dips in the masker level). One way to overcome this confound, and allow FMB comparisons across listener groups with different stationary-noise performance, is to adjust the response set size to equalize performance across groups at a fixed SNR. However, this technique is only valid under the assumption that changes in set size have the same effect on percentage-correct performance for different masker types. This assumption was tested by measuring nonsense-syllable identification for normal-hearing listeners as a function of SNR, set size and masker (stationary noise, 4- and 32-Hz modulated noise and an interfering talker). Set-size adjustment had the same impact on performance scores for all maskers, confirming the independence of FMB (at matched SNRs) and set size. These results, along with those of a second experiment evaluating an adaptive set-size algorithm to adjust performance levels, establish set size as an efficient and effective tool to adjust baseline performance when comparing effects of masker fluctuations between listener groups.

Goal-independent mechanisms for free response generation: Creative and pseudo-random performance share neural substrates

To what extent free response generation in different tasks uses common and task-specific neurocognitive processes has remained unclear. Here, we investigated overlap and differences in neural activity during musical improvisation and pseudo-random response generation. Brain activity was measured using fMRI in a group of professional classical pianists, who performed musical improvisation of melodies, pseudo-random key-presses and a baseline condition (sight-reading), on either two, six or twelve keys on a piano keyboard. The results revealed an extensive overlap in neural activity between the two generative conditions. Active regions included the dorsolateral and dorsomedial prefrontal cortices, inferior frontal gyrus, anterior cingulate cortex and pre-SMA. No regions showed higher activity in improvisation than in pseudo-random generation. These findings suggest that the activated regions fulfill generic functions that are utilized in different types of free generation tasks, independent of overall goal. In contrast, pseudo-random generation was accompanied by higher activity than improvisation in several regions. This presumably reflects the participants' musical expertise as well as the pseudo-random generation task's high load on attention, working memory, and executive control. The results highlight the significance of using naturalistic tasks to study human behavior and cognition. No brain activity was related to the size of the response set. We discuss that this may reflect that the musicians were able to use specific strategies for improvisation, by which there was no simple relationship between response set size and neural activity.

An Assessment of Behavioral Dynamic Information Processing Measures in Audiovisual Speech Perception

Research has shown that visual speech perception can assist accuracy in identification of spoken words. However, little is known about the dynamics of the processing mechanisms involved in audiovisual integration. In particular, architecture and capacity, measured using response time methodologies, have not been investigated. An issue related to architecture concerns whether the auditory and visual sources of the speech signal are integrated “early” or “late.” We propose that “early” integration most naturally corresponds to coactive processing whereas “late” integration corresponds to separate decisions parallel processing. We implemented the double factorial paradigm in two studies. First, we carried out a pilot study using a two-alternative forced-choice discrimination task to assess architecture, decision rule, and provide a preliminary assessment of capacity (integration efficiency). Next, Experiment 1 was designed to specifically assess audiovisual integration efficiency in an ecologically valid way by including lower auditory S/N ratios and a larger response set size. Results from the pilot study support a separate decisions parallel, late integration model. Results from both studies showed that capacity was severely limited for high auditory signal-to-noise ratios. However, Experiment 1 demonstrated that capacity improved as the auditory signal became more degraded. This evidence strongly suggests that integration efficiency is vitally affected by the S/N ratio.

Effects of response-set size on error-related brain activity

To study the effect of response-set size on action monitoring processes, the error-related negativity (Ne/ERN), the correct-related negativity (Nc/CRN), and behavioral indicators of action monitoring were compared across three groups of participants performing a two-choice, a four-choice, or an eight-choice version of the flanker task. After controlling for differential contribution of stimulus-related activity to response-locked averages resulting from large differences in response times across conditions, response-set size had strong effects on Ne/ERN and Nc/CRN. With increasing response-set size, the Ne/ERN amplitude decreased, but the Nc/CRN amplitude increased. Moreover, post-error behavioral adjustments were impaired with an increasing response-set size. These results suggest that action monitoring severely suffers when response-set size is increased. Implications of these findings for present theories of Ne/ERN and Nc/CRN are discussed.

Syntactic and strategic forces in picture naming: gender retrieval in blocked priming experiments

One of the core issues in research on the processing of grammatical gender in language production is whether it is represented as an abstract node or whether it is an inherent property of the noun. Interpretation of the relevant data is often complicated by the fact that they could theoretically concern either gender competition or determiner competition. In the present study, participants named pictures in pre-defined sets, using gender-marked demonstrative determiners. We independently varied the number of determiners and the number of gender classes in a given set. There were clear effects of the number of determiners (response set size). Evidence for effects of gender was less conclusive. Time course analyses revealed that a potential effect of gender is very subtle and highly susceptible to strategies that participants develop in the course of the experiment.

Application of Fuzzy Signal Detection Theory to the Discrimination of Morphed Tank Images

The effect of response set size on performance on a detection task was evaluated using both fuzzy and traditional signal detection theory. Fuzzy categories of stimuli were created using morphing software to blend profile images of American (M1A1) and Iraqi (T55) tanks to different degrees. These combinations were used to create static images varying from 100% T55 to 0% T55 (100% MIA1). Participants were asked to indicate the degree to which each image did not resemble an American tank. Consistent with previous research, results indicated that the FSDT model conforms to the normality assumption of traditional SDT. In addition, forcing observers to make binary decisions impaired performance relative to multi-category response sets in the FSDT analysis but not the traditional analysis. However, there were more model convergence failures in the FSDT analysis relative to the traditional analysis, mostly associated with conditions in which there were 100 response categories.

Deconfounding Serial Recall

Immediate recall of nine-digit lists was examined with a method designed to disentangle three factors: input serial position, output position, and response set size (the number of items yet to be recalled). Recall began at Input Serial Position 1, 4, or 7 and included either three consecutive items (in partial recall) or all items from the cued point to the end of the list and then continuing from the beginning in a circular fashion (in whole recall). Lists were spoken or printed and were sometimes temporally grouped. Specially selected comparisons demonstrated that (1) the large primacy effect in serial recall occurs mostly because of output interference, without which larger recency effects are seen instead; (2) benefits of mnemonic grouping are dependent on stimulus grouping mainly for auditory stimuli; and (3) auditory superiority effects stem from a greater resistance of acoustic memory to output interference. We offer an integration of results from serial recall and other memory tasks and caution against modeling serial recall in isolation.

Set size and repetition in the picture–word interference paradigm: implications for models of naming

Caramazza and Costa ( Cognition 75 (2000) B51) reported results which demonstrate that a semantically related word distractor interferes in picture naming even when it is not in the response set and there is no possibility for mediated interference. They interpreted the results to be problematic for the model of lexical access proposed by Levelt, Roelofs, and Meyer ( Behavioral and Brain Sciences 22 (1999) 1). Roelofs ( Cognition 80 (2001 , this issue) 283–290 argues that those results are not inconsistent with Levelt et al.'s model when certain new assumptions about the mechanism of lexical selection are considered. Here we show that even with these assumptions the model still makes the wrong predictions. We report new results which demonstrate that the semantic interference and facilitation effects that are obtained respectively in the basic-level and category-level naming variants of the picture–word interference paradigm are not the result of response set size and response repetitions.

The roles of stimulus and response set size in the identification and categorisation of unidimensional stimuli

Dual mapping theories (e.g., Atkinson, Bower, & Crothers, 1965; Lacouture & Marley, 1991, 1995) propose that psychophysical judgment involves two component processes: stimulus representation and response selection. This empirical and theoretical paper studies the relative contributions of these two component processes in determining performance in the identification and categorisation of unidimensional stimuli. We extend Lacouture and Marley's (1995) mapping model of absolute identification to categorisation, and demonstrate that the extension predicts that the number of response categories has a much larger effect than the number of stimuli in determining various aspects of the relevant data, including the so-called bow effect. This prediction is supported by the results of three experiments in which the number of stimuli and/or the number of response categories was manipulated.

On random generation and the central executive of working memory.

Four experiments explore participants' attempts to generate random sequences. Experiment 1 showed that oral random number generation is strongly affected by both response speed and response set size, in contrast to a random key-pressing task in Expt 2. Expt 3 confirmed differences between output modalities in the set size effect, and Expt 4 indicated that an important source of difficulty in producing random numbers orally lies in the requirement to represent candidate choices. Across experiments, data show a strong tendency on the part of participants to suppress response repetitions, an effect which decays over intervening responses. Whilst consistent with the possibility that random generation taps some executive functions, findings suggest the need to expand current models of attentional control in working memory to account for distinct constraints.

Explorations of Cohen, Dunbar, and McClelland's (1990) connectionist model of Stroop performance.

The J. D. Cohen, K. Dunbar, and J. L. McClelland (1990) model of Stroop task performance is used to model data from a study by D. H. Spielder, D. A. Balota, and M. E. Faust (1996). The results indicate that the model fails to capture overall differences between word reading and color naming latencies when set size is increased beyond 2 response alternatives. Further empirical evidence is presented that suggests that the influence of increasing response set size in Stroop task performance is to increase the difference between overall color naming and word reading, which is in direct opposition to the decrease produced by the Cohen et al. architecture. Although the Cohen et al. model provides a useful description of meaning-level interference effects, the qualitative differences between word reading and color naming preclude a model that uses identical architectures for each process, such as that of Cohen et al., to fully capture performance in the Stroop task.

Random Generation of Numbers: A Search for Underlying Processes

To explicate the mechanisms which support attempts at random number production, two experiments explore the role of memory systems in random sequencing. In the first, subjects produced written random number sequences with two response vocabularies, and subjects were also required to estimate how often they had chosen particular values in their response set. Memory for responses was found to correlate with the quality of random generation, and sequences were less random with a larger response repertoire. However, memory performance did not account for the effect of response set size, suggesting different aetiologies for these effects. In a second experiment, subjects were given a concurrent memory load during random generation, and a concurrent task requiring avoidance of certain response values. Performance declined under dual-task conditions, and the avoidance requirement particularly impaired subjects' ability to inhibit prepotent responses. Findings are integrated into a view of random generation as a complex task with multiple performance constraints.

Response selection deficits in frontal excisions

We compared the performance of patients with frontal excisions, patients with temporal excisions and controls in tasks involving speeded choice responses in which a number of variables were manipulated including: perceptual difficulty, stimulus and response set-size, associative complexity, and spatial stimulus-response compatibility. Response times were sensitive to all manipulations but did not show any group differences. The error rates of the three groups were equally affected by perceptual difficulty and response set-size but frontals were preferentially affected by spatial S-R compatibility, associative complexity, and the number of stimuli per response. The results are consistent with a basic deficit in response selection processes which could underly many problems produced by frontal lesions.

Segmental intelligibility of synthetic speech produced by rule

This paper reports the results of an investigation that employed the modified rhyme test (MRT) to measure the segmental intelligibility of synthetic speech generated automatically by rule. Synthetic speech produced by ten text-to-speech systems was studied and compared to natural speech. A variation of the standard MRT was also used to study the effects of response set size on perceptual confusions. Results indicated that the segmental intelligibility scores formed a continuum. Several systems displayed very high levels of performance that were close to or equal to scores obtained with natural speech; other systems displayed substantially worse performance compared to natural speech. The overall performance of the best system, DECtalk--Paul, was equivalent to the data obtained with natural speech for consonants in syllable-initial position. The findings from this study are discussed in terms of the use of a set of standardized procedures for measuring intelligibility of synthetic speech under controlled laboratory conditions. Recent work investigating the perception of synthetic speech under more severe conditions in which greater demands are made on the listener's processing resources is also considered. The wide range of intelligibility scores obtained in the present study demonstrates important differences in perception and suggests that not all synthetic speech is perceptually equivalent to the listener.

Response-set size effects in recall and recognition

A retroactive inhibition design was employed to study the role of rules in paired-associate learning and retention. Subjects first learned one of three unmixed lists of paired nonsense syllables. A rule applied to all items in two of the lists: Pair members changed first letter (rhymed) or changed middle letter. In the third list, pair members were not related. A common second list was learned, and either a first-list recall test or a four-alternative forced-choice recognition test followed. First-list learning and recall were superior for lists with rules. Recognition, which did not show a ceiling effect, was not superior for lists with rules when both target and distractors followed a rule. The results supported the hypothesis that rules facilitate retrieval rather than storage, and that they do so to the extent that they reduce response-set size.

Rules and Paired-Associate Retention

The present study sought to determine whether the superior retention enjoyed by lists with rules is a function of superior storage of individual pairs or superior retrieval resulting from the reduction in response-set size. Fifty-four students participated as a requirement for introductory psychology. They were randomly assigned to conditions and individually tested. A retroactive inhibition design was used with all lists made up of 12 CVC word pairs. Each student learned one of t h e e first lists. For two of the lists a general rule applied, either each pair rhymed, e.g., PAN-FAN, or changed middle letter, e.g., PAN-PIN. The third list consisted of unrelated pairs, e.g., PAN-WIG. Stimuli in the three lists were the same and responses were similar in ThorndikeLorge frequency. Anticipation learning at 2:2-sec. rate with a 4-sec. intertrial interval was taken to 7/12 correct. Three different orders of the pairs were used, and all lists had the same sequence of orders. subjects were informed of the rule applying to their list. All subjects received 12 anticipation trials on a common second list of unrelated pairs, containing the first list stimuli and responses unrelated to those in the first lists. There followed a test of retention of the first list in which the entire set of stimuli was presented on a sheet of paper. Half the subjects had to recall each response and half had to select it from two alternatives, both of which followed the rule for the rule lists or were unrelated for the unrelated list. Mean crials to first-list criterion were 3.44 (middle-letter change), 5.33 (rhyme), and 9.78 (unrelated). These means were significantly different (P2,51 = 19.66, p < .01). Number of responses recalled yielded a significant F2.24 of 13.13 ( 9 < .01). The means were 10.00 (middle-letter change), 9.22 (rhyme), and 5.44 (unrelated). Little difference was observed in recognition, with means ranging from 11.33 ro 12.00. These results support the response-restriction hypothesis which maintains that rules are beneficial to the extent that they restrict the number of possible responses to the stimulus. In first list learning and in recall rules did indeed prove helpful, with the more restrictive rule (middle-letter change) proving most effective. As expected, rules were not helpful in recognition. Since both alternatives satisfied the rule, response-set size was the same for all. W e may conclude, then, that rules aid retrieval by response restriction rather than storage of individual pairs. This conclusion can only be tentative, however, for the negligible difference obtained among means for recognition may have been imposed by a ceiling effect. Accepted December 17, 1981. '

Response-Set Size and Paired-Associate Learning

4 groups of 15 students each learned a single unmixed list of 13 pairs for five anticipation trials followed by a free-recall test. The lists contained pairs of nonsense syllables that rhymed, changed middle letter, reversed the letter order, or were unrelated. All lists involving rules were superior to the list of unrelated pairs. In order of increasing effectiveness the rules were rhyming, middle-letter change, letter reversal. Performance was inversely related to response-set size. The more restrictive rules resulted in increased free recall of the pairs. Results supported the hypothesis that rules are beneficial to the extent that they reduce response-set size.