A multinomial model for short-term priming in word identification

The elimination of positive priming with increasing prime duration reflects a transition from perceptual fluency to disfluency rather than bias against primed words

With immediate repetition priming of forced choice perceptual identification, short prime durations produce positive priming (i.e., priming the target leads to higher accuracy, while priming the foil leads to lower accuracy). Many theories explain positive priming following short duration primes as reflecting increased perceptual fluency for the primed target (i.e., decreased identification latency). However, most studies only examine either accuracy or response times, rather than considering the joint constraints of response times and accuracy to properly address the role of decision biases and response caution. This is a critical oversight because several theories propose that the transition to negative priming following a long duration prime reflects a decision strategy to compensate for the effect of increased perceptual fluency. In contrast, the nROUSE model of Huber and O'Reilly (2003) explains this transition as reflecting perceptual habituation, and thus a change to perceptual disfluency. We confirmed this prediction by applying a sequential sampling model (the diffusion race model) to accuracy and response time distributions from a new single item same-different version of the priming task. In this way, we measured strategic biases and perceptual fluency in each condition for each subject. The nROUSE model was only applied to accuracy from the original forced-choice version of the priming task. This application of nROUSE produced separate predictions for each subject regarding the degree of fluency and disfluency in each condition, and these predictions were confirmed by the drift rate parameters (i.e., fluency) from the response time model in contrast to the threshold parameters (i.e., bias).

Using continual flash suppression to investigate cognitive aftereffects

In a naming task with visually presented words, Prioli and Kahan (2015) reported that negatively valenced words were identified more slowly than neutral words in a condition with continual flash suppression (CFS), which involves showing the dominant eye changing Mondrian patterns, delaying awareness for the word shown to the other eye. However, when these same words were shown to both eyes (i.e., no CFS), negatively valenced words were identified more quickly. The authors hypothesized that the negative word deficit with CFS reflects greater habituation (i.e., a cognitive aftereffect) that accrues for negative words before the observer becomes aware of the word. However, aftereffects typically occur in response to a target stimulus that follows an adapting stimulus, rather than in response to a single stimulus that is initially processed without awareness. Thus, it is not immediately obvious that the explanation provided by Prioli and Kahan is adequate. Here I report a simulation using a model that was previously applied to cognitive aftereffects, demonstrating that their account can explain this crossover interaction. These results suggest that CFS may be a useful technique for studying cognitive aftereffects without concern for conscious decision strategies.

Testing an associative account of semantic satiation

How is the meaning of a word retrieved without interference from recently viewed words? The ROUSE theory of priming assumes a discounting process to reduce source confusion between subsequently presented words. As applied to semantic satiation, this theory predicted a loss of association between the lexical item and meaning. Four experiments tested this explanation in a speeded category-matching task. All experiments used lists of 20 trials that presented a cue word for 1s followed by a target word. Randomly mixed across the list, 10 trials used cues drawn from the same category whereas the other 10 trials used cues from 10 other categories. In Experiments 1a and 1b, the cues were repeated category labels (FRUIT-APPLE) and responses gradually slowed for the repeated category. In Experiment 2, the cues were nonrepeated exemplars (PEAR-APPLE) and responses remained faster for the repeated category. In Experiment 3, the cues were repeated exemplars in a word matching task (APPLE-APPLE) and responses again remained faster for the repeated category.

Multinomial Processing Tree Models

Multinomial processing tree (MPT) models have become popular in cognitive psychology in the past two decades. In contrast to general-purpose data analysis techniques, such as log-linear models or other generalized linear models, MPT models are substantively motivated stochastic models for categorical data. They are best described as tools (a) for measuring the cognitive processes that underlie human behavior in various tasks and (b) for testing the psychological assumptions on which these models are based. The present article provides a review of MPT models and their applications in psychology, focusing on recent trends and developments in the past 10 years. Our review is nontechnical in nature and primarily aims at informing readers about the scope and utility of MPT models in different branches of cognitive psychology.

Computer simulations of the ROUSE model: an analytic simulation technique and a comparison between the error variance-covariance and bootstrap methods for estimating parameter confidence

This article provides important mathematical descriptions and computer algorithms in relation to the responding optimally with unknown sources of evidence (ROUSE) model of Huber, Shiffrin, Lyle, and Ruys (2001), which has been applied to short-term priming phenomena. In the first section, techniques for obtaining parameter confidence intervals and parameter correlations are described, which are generally applicable to any mathematical model. In the second section, a technique for producing analytic ROUSE predictions is described. Huber et al. (2001) averaged many stochastic trials to obtain stable behavior. By appropriately weighting all possible combinations of feature states, an alternative analytic version is developed, yielding asymptotic model behavior with fewer computations. The third section ties together these separate techniques, obtaining parameter confidence and correlations for the analytic version of the ROUSE model. In doing so, previously unreported behaviors of the model are revealed. In particular, complications due to local minima are discussed, in terms of both variance-covariance analyses and bootstrap sampling analyses.

Multinomiale Verarbeitungs- baummodelle in der Sozialpsychologie

Zusammenfassung: Multinomiale Verarbeitungsbaummodelle (MVB-Modelle) werden in der kognitiven Psychologie seit über 15 Jahren angewendet. Der Vorteil der Überführung einer Theorie in ein MVB-Modell ist zunächst in der mathematisch exakten Formulierung zu sehen; derart formulierte Hypothesen gewinnen an Überprüfbarkeit und empirischem Gehalt. Darüber hinaus erlaubt ein validiertes MVB-Modell die Interpretation seiner Parameter als Wahrscheinlichkeiten der zugrunde liegenden Prozesse und damit die getrennte Erfassung dieser Prozesse. In der Sozialpsychologie wurden MVB-Modelle unter anderem in der Forschung zur illusorischen Korrelation, zum Inkongruenzvorteil im Gedächtnis für schemarelevantes Material, und zu sozialer Kategorisierung eingesetzt und haben in diesen Bereichen entscheidend zum Erkenntnisgewinn beigetragen. Eine methodische Weiterentwicklung ermöglicht die Modellierung von Parameterheterogenität und eröffnet damit neue Möglichkeiten der Anwendung von MVB-Modellen in der Sozialpsychologie.

Confusion and Compensation in Visual Perception: Effects of Spatiotemporal Proximity and Selective Attention

The authors investigated spatial, temporal, and attentional manipulations in a short-term repetition priming paradigm. Brief primes produced a strong preference to choose the primed alternative, whereas long primes had the opposite effect. However, a 2nd brief presentation of a long prime produced a preference for the primed word despite the long total prime duration. These surprising results are explained by a computational model that posits the offsetting components of source confusion (prime features are confused with target features) and discounting (evidence from primed features is discounted). The authors obtained compelling evidence for these components by showing how they can cooperate or compete through different manipulations of prime salience. The model allows for dissociations between prime salience and the magnitude of priming, thereby providing a unified account of "subliminal" and "supraliminal" priming.

High-level effects of masking on perceptual identification

The extent to which visual form versus higher-level information is used to identify briefly flashed words is assessed in a perceptual identification task. In this task, a word is briefly flashed, post-masked, and a decision is made between two alternatives. The availability of visual (e.g., case or color) and higher-level information (e.g., abstract letter codes, phonology, and meaning) was manipulated by varying the information that discriminates the alternatives. Performance was better with higher-level than with visual information when pattern post-masks were used, but the reverse occurred without masking. The authors conclude that both higher-level and visual-form information can be used to identify words with the strategy depending on the information available at the time of the choice.

Bias in masked word identification: unconscious influences of repetition priming

The beneficial influence of a prior study episode on subsequent identification of a word includes a large bias component, revealed in the forced-choice variant of the masked word identification test. In that type of test, subjects show a preference for a studied probe over a nonstudied probe, regardless of which one matches the masked target word. The forced-choice test was used in the present experiments to test the possibility that this bias effect is due to conscious recollection. Results show that bias was strongly attenuated (1) by changes in modality between study and test, and (2) under certain conditions, by using a conceptually driven study task. The bias effect was found only when probes were orthographically similar to one another, as predicted by the counter model (Ratcliff & McKoon, 1997). These results provide strong evidence that the bias effect is not mediated by conscious recollection.

Mechanisms of source confusion and discounting in short-term priming: 1. Effects of prime duration and prime recognition

Huber, Shriffrin, Lyle, and Ruys (2001) measured short-term repetition priming effects in perceptual identification with two-alternative forced-choice testing. There was a preference to choose repeated words following passive viewing of primes and a preference against choosing repeated words following active responding to primes. In this present study, we explored conditions of prime processing that produce this pattern of results. Experiment 1 revealed that increased prime duration under passive viewing instructions produces the active priming pattern. Experiment 2 assessed memory for primes: With poor recognition of primes, there was a strong preference for repeated words; however, with good recognition of primes, this preference was eliminated. These results are modeled by a computational theory of optimal decision making, responding optimally with unknown sources of evidence (ROUSE). In ROUSE, a preference for repeated words results from source confusion between primes and choice words. A reversal in the direction of preference arises from the discounting of words known to have also appeared as primes.