-
[show abstract]
[hide abstract]
ABSTRACT: There is a growing chorus of voices in the scientific community calling for greater openness in the sharing of raw data that lead to a publication. In this commentary, we discuss the merits of sharing, common concerns that are raised, and practical issues that arise in developing a sharing policy. We suggest that the cognitive science community discuss the topic and establish a data-sharing policy.
Topics in Cognitive Science 01/2013; 5(1):214-221. · 2.88 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Four experiments examined listeners' segmentation of ambiguous schwa-initial sequences (e.g., a long vs. along) in casual speech, where acoustic cues can be unclear, possibly increasing reliance on contextual information to resolve the ambiguity. In Experiment 1, acoustic analyses of talkers' productions showed that the one-word and two-word versions were produced almost identically, regardless of the preceding sentential context (biased or neutral). These tokens were then used in three listening experiments, whose results confirmed the lack of local acoustic cues for disambiguating the interpretation, and the dominance of sentential context in parsing. Findings speak to the H&H theory of speech production (Lindblom, 1990), demonstrate that context alone guides parsing when acoustic cues to word boundaries are absent, and demonstrate how knowledge of how talkers speak can contribute to an understanding of how words are segmented.
Journal of Memory and Language 05/2012; 66(4):509-529. · 2.73 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The power law (y =ax
−b) has been shown to provide a good description of data collected in a wide range of fields in psychology. R. B. Anderson and
Tweney (1997) suggested that the model’s data-fitting success may in part be artifactual, caused by a number of factors, one
of which is the use of improper data averaging methods. The present paper follows up on their work and explains causes of
the power law artifact. A method for studying the geometric relations among responses generated by mathematical models is
introduced that shows the artifact is a result of the combined contributions of three factors: arithmetic averaging of data
that are generated from a nonlinear model in the presence of individual differences.
Memory & Cognition 04/2012; 28(5):832-840. · 1.92 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Words can be pronounced in multiple ways in casual speech. Corpus analyses of the frequency with which these pronunciation variants occur (e.g., Patterson & Connine, 2001) show that typically, one pronunciation variant tends to predominate; this raises the question of whether variant recognition is aligned with exposure frequency. We explored this issue in words containing one of four phonological contexts, each of which favors one of four surface realizations of word-medial /t/: [t], [ʔ], [ɾ], or a deleted variant. The frequencies of the four realizations in all four contexts were estimated for a set of words in a production experiment. Recognition of all pronunciation variants was then measured in a lexical decision experiment. Overall, the data suggest that listeners are sensitive to variant frequency: Word classification rates closely paralleled production frequency. The exceptions to this were [t] realizations (i.e., canonical pronunciations of the words), a finding which confirms other results in the literature and indicates that factors other than exposure frequency affect word recognition.
Journal of Phonetics 07/2011; 39(3):304-311. · 1.41 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: An ideal experiment is one in which data collection is efficient and the results are maximally informative. This standard can be difficult to achieve because of uncertainties about the consequences of design decisions. We demonstrate the success of a Bayesian adaptive method (adaptive design optimization, ADO) in optimizing design decisions when comparing models of the time course of forgetting. Across a series of testing stages, ADO intelligently adapts the retention interval in order to maximally discriminate power and exponential models. Compared with two different control (non-adaptive) methods, ADO distinguishes the models decisively, with the results unambiguously favoring the power model. Analyses suggest that ADO's success is due in part to its flexibility in adjusting to individual differences. This implementation of ADO serves as an important first step in assessing its applicability and usefulness to psychology.
Psychonomic Bulletin & Review 02/2011; 18(1):204-10. · 2.61 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Speech is produced over time, and this makes sensitivity to timing between speech events crucial for understanding language. Two experiments investigated whether perception of function words (e.g., or, are) is rate dependent in casual speech, which often contains phonetic segments that are spectrally quite reduced. In Experiment 1, talkers spoke sentences containing a target function word; slowing talkers' speech rate around this word caused listeners to perceive sentences as lacking the word (e.g., leisure or time was perceived as leisure time). In Experiment 2, talkers spoke matched sentences lacking a function word; speeding talkers' speech rate around the region in which the function word had been embedded in Experiment 1 caused listeners to perceive a function word that was never spoken (e.g., leisure time was perceived as leisure or time). The results suggest that listeners formed expectancies based on speech rate, and these expectancies influenced the number of words and word boundaries perceived. These findings may help explain the robustness of speech recognition when speech signals are distorted (e.g., because of a casual speaking style).
Psychological Science 09/2010; 21(11):1664-70. · 4.43 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To comprehend spoken language, listeners need to find words from a continuous stream of speech sounds. Little work has explored whether there are reliable acoustic cues to word boundaries in conversational speech, which is highly reduced and under-articulated, potentially creating ambiguities at word boundaries. Segmentation may be even more difficult when the same segment repeats at a word boundary, ending the preceding word and beginning the following word (e.g., gas station). Segmentation in this environment was investigated by examining the production and perception of fricative s in semi-spontaneous speech. Twenty talkers produced sentences containing ambiguous two-word sequences with s between the two words. All sequences are interpretable in three ways (e.g., grow snails, gross snails, and gross nails) depending on how the frication is segmented. Acoustic analyses of the production data examined whether there are acoustic cues distinguishing the three versions of the ambiguous sequences. Listening experiments using the talkers' productions as stimuli evaluated the degree of ambiguity in the tokens and identified acoustic cues that listeners use to segment the two words. Results will be discussed in the context of theories of speech perception and word segmentation.
The Journal of the Acoustical Society of America 03/2010; 127(3):1956. · 1.55 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Discriminating among competing statistical models is a pressing issue for many experimentalists in the field of cognitive science. Resolving this issue begins with designing maximally informative experiments. To this end, the problem to be solved in adaptive design optimization is identifying experimental designs under which one can infer the underlying model in the fewest possible steps. When the models under consideration are nonlinear, as is often the case in cognitive science, this problem can be impossible to solve analytically without simplifying assumptions. However, as we show in this letter, a full solution can be found numerically with the help of a Bayesian computational trick derived from the statistics literature, which recasts the problem as a probability density simulation in which the optimal design is the mode of the density. We use a utility function based on mutual information and give three intuitive interpretations of the utility function in terms of Bayesian posterior estimates. As a proof of concept, we offer a simple example application to an experiment on memory retention.
Neural Computation 12/2009; 22(4):887-905. · 1.88 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Models of a psychological process can be difficult to discriminate experimentally because it is not easy to determine the values of the critical design variables (e.g., presentation schedule, stimulus structure) that will be most informative in differentiating them. Recent developments in sampling-based search methods in statistics make it possible to determine these values and thereby identify an optimal experimental design. After describing the method, it is demonstrated in 2 content areas in cognitive psychology in which models are highly competitive: retention (i.e., forgetting) and categorization. The optimal design is compared with the quality of designs used in the literature. The findings demonstrate that design optimization has the potential to increase the informativeness of the experimental method.
Psychological Review 08/2009; 116(3):499-518. · 7.76 Impact Factor
-
Mark A Pitt
[show abstract]
[hide abstract]
ABSTRACT: Spoken words undergo frequent and often predictable variation in pronunciation. One form of variation is medial /t/ deletion, in which words like center and cantaloupe are pronounced without acoustic cues indicative of syllable-initial /t/. Three experiments examined the consequences of this missing phonetic information on lexical activation. In Experiment 1, the Ganong Paradigm (W. F. Ganong, 1980) was used to measure the strength of activation of /t/-deleted variants, comparing labeling and response time results with their citation counterparts. Phonemic restoration was used in Experiment 2 to generalize the results. In Experiment 3, Experiment 1 was replicated with a large number of trials so that the time course of activation could be mapped. Results show that lexical influences on labeling begin sooner and reach a higher level for the citation than for the /t/-deleted variant, although the overall shapes of their activation profiles are similar.
Journal of Experimental Psychology Human Perception & Performance 07/2009; 35(3):896-910. · 3.06 Impact Factor
-
Mark A Pitt
[show abstract]
[hide abstract]
ABSTRACT: One account of how pronunciation variants of spoken words (center-> "senner" or "sennah") are recognized is that sublexical processes use information about variation in the same phonological environments to recover the intended segments (Gaskell & Marslen-Wilson, 1998). The present study tests the limits of this phonological inference account by examining how listeners process for the first time a pronunciation variant of a newly learned word. Recognition of such a variant should occur as long as it possesses the phonological structure that legitimizes the variation. Experiments 1 and 2 identify a phonological environment that satisfies the conditions necessary for a phonological inference mechanism to be operational. Using a word-learning paradigm, Experiments 3 through 5 show that inference alone is not sufficient for generalization but could facilitate it, and that one condition that leads to generalization is meaningful exposure to the variant in an overheard conversation, demonstrating that lexical processing is necessary for variant recognition.
Journal of Memory and Language 07/2009; 61(1):19-36. · 2.73 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Computational models are powerful tools that can enhance the understanding of scientific phenomena. The enterprise of modeling is most productive when the reasons underlying the adequacy of a model, and possibly its superiority to other models, are understood. This chapter begins with an overview of the main criteria that must be considered in model evaluation and selection, in particular explaining why generalizability is the preferred criterion for model selection. This is followed by a review of measures of generalizability. The final section demonstrates the use of five versatile and easy-to-use selection methods for choosing between two mathematical models of protein folding.
Methods in enzymology 02/2009; 454:287-304. · 1.90 Impact Factor
-
Advances in Neural Information Processing Systems 22: 23rd Annual Conference on Neural Information Processing Systems 2009. Proceedings of a meeting held 7-10 December 2009, Vancouver, British Columbia, Canada.; 01/2009
-
[show abstract]
[hide abstract]
ABSTRACT: A primary criterion on which models of cognition are evaluated is their ability to fit empirical data. To understand the reason why a model yields a good or poor fit, it is necessary to determine the data-fitting potential (i.e., flexibility) of the model. In the first part of this article, methods for comparing models and studying their flexibility are reviewed, with a focus on parameter space partitioning (PSP), a general-purpose method for analyzing and comparing all classes of cognitive models. PSP is then demonstrated in the second part of the article in which two connectionist models of speech perception (TRACE and ARTphone) are compared to learn how design differences affect model flexibility.
Cognitive Science 12/2008; 32(8):1285-303. · 2.38 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Smith and Minda (1998, 2002) argued that the response scaling parameter y in the exemplar-based generalized context model (GCM) makes the model unnecessarily complex and allows it to mimic the behavior of a prototype model. We evaluated this criticism in two ways. First, we estimated the complexity of the GCM with and without the yparameter and also compared its complexity to that of a prototype model. Next, we assessed the extent to which the models mimic each other, using two experimental designs (Nosofsky & Zaki, 2002, Experiment 3; Smith & Minda, 1998, Experiment 2), chosen because these designs are thought to differ in the degree to which they can discriminate the models. The results show that y can increase the complexity of the GCM, but this complexity does not necessarily allow mimicry. Furthermore, if statistical model selection methods such as minimum description length are adopted as the measure of model performance, the models will be highly discriminable, irrespective of design.
Psychonomic Bulletin & Review 01/2008; 14(6):1043-50. · 2.61 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Regressive place assimilation is a form of pronunciation variation in which a word-final alveolar sound takes the place of articulation of a following labial or velar sound, as when green boat is pronounced greem boat. How listeners recover the intended word (e.g., green, given greem) has been a major focus of spoken word recognition theories. However, the extent to which this variation occurs in casual, unscripted speech has previously not been reported. Two studies of pronunciation variation were conducted using a spontaneous speech corpus. First, phonetic labeling data were used to identify contexts in which assimilation could occur, namely, when a word-final alveolar stop (/t/, /d/, or /n/) was followed by a velar or labial consonant. Assimilation was indicated relatively infrequently, while deletion, glottalization, or canonical pronunciations were more often indicated. Moreover, lexical frequency was shown to affect pronunciation; high frequency lexical items showed more types of variation. Second, acoustic analyses showed that neither place of articulation cues (indicated by second formant variation) nor relative amplitude was sufficient to distinguish assimilated from deleted and canonical variants; only when closure duration was additionally taken into account were these three variant types distinguishable. Implications for theories of word recognition are discussed.
The Journal of the Acoustical Society of America 11/2007; 122(4):2340-53. · 1.55 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: We introduce a Fourier Transformation technique that enables one to derive closed-form expressions of performance measures (e.g., hit and false alarm rates) of simulation-based models of recognition memory. Application of the technique is demonstrated using the bind cue decide model of episodic memory (BCDMEM; Dennis & Humphreys, 2001). In addition to reducing the time required to test the model, which for models like BCDMEM can be excessive, asymptotic expressions of the measures reveal heretofore unknown properties of the model, such as model predictions being dependent on vector length.
Journal of Mathematical Psychology 06/2007; 51(3):198-204. · 1.70 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Vitevitch and Luce (1998) showed that the probability with which phonemes co-occur in the language (phonotactic probability) affects the speed with which words and nonwords are named. Words with high phonotactic probabilities between phonemes were named more slowly than words with low probabilities, whereas with nonwords, just the opposite was found. To reproduce this reversal in performance, a model would seem to require not merely sublexical representations, but sublexical representations that are relatively independent of lexical representations. ARTphone (Grossberg, Boardman, & Cohen, 1997) is designed to meet these requirements. In this study, we used a technique called parameter space partitioning to analyze ARTphone's behavior and to learn if it can mimic human behavior and, if so, to understand how. To perform best, differences in sublexical node probabilities must be amplified relative to lexical node probabilities to offset the additional source of inhibition (from top-down masking) that is found at the sublexical level.
Psychonomic Bulletin & Review 06/2007; 14(3):442-8. · 2.61 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Many models of spoken word recognition posit the existence of lexical and sublexical representations, with excitatory and inhibitory mechanisms used to affect the activation levels of such representations. Bottom-up evidence provides excitatory input, and inhibition from phonetically similar representations leads to lexical competition. In such a system, long words should produce stronger lexical activation than short words, for 2 reasons: Long words provide more bottom-up evidence than short words, and short words are subject to greater inhibition due to the existence of more similar words. Four experiments provide evidence for this view. In addition, reaction-time-based partitioning of the data shows that long words generate greater activation that is available both earlier and for a longer time than is the case for short words. As a result, lexical influences on phoneme identification are extremely robust for long words but are quite fragile and condition-dependent for short words. Models of word recognition must consider words of all lengths to capture the true dynamics of lexical activation.
Journal of Experimental Psychology Human Perception & Performance 11/2006; 32(5):1120-35. · 3.06 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To model behavior, scientists need to know how models behave. This means learning what other behaviors a model can produce besides the one generated by participants in an experiment. This is a difficult problem because of the complexity of psychological models (e.g., their many parameters) and because the behavioral precision of models (e.g., interval-scale performance) often mismatches their testable precision in experiments, where qualitative, ordinal predictions are the norm. Parameter space partitioning is a solution that evaluates model performance at a qualitative level. There exists a partition on the model's parameter space that divides it into regions that correspond to each data pattern. Three application examples demonstrate its potential and versatility for studying the global behavior of psychological models.
Psychological Review 02/2006; 113(1):57-83. · 7.76 Impact Factor
