EARSHOT: A Minimal Neural Network Model of Incremental Human Speech Recognition

Despite the lack of invariance problem (the many-to-many mapping between acoustics and percepts), human listeners experience phonetic constancy and typically perceive what a speaker intends. Most models of human speech recognition (HSR) have side-stepped this problem, working with abstract, idealized inputs and deferring the challenge of working with real speech. In contrast, carefully engineered deep learning networks allow robust, real-world automatic speech recognition (ASR). However, the complexities of deep learning architectures and training regimens make it difficult to use them to provide direct insights into mechanisms that may support HSR. In this brief article, we report preliminary results from a two-layer network that borrows one element from ASR, long short-term memory nodes, which provide dynamic memory for a range of temporal spans. This allows the model to learn to map real speech from multiple talkers to semantic targets with high accuracy, with human-like timecourse of lexical access and phonological competition. Internal representations emerge that resemble phonetically organized responses in human superior temporal gyrus, suggesting that the model develops a distributed phonological code despite no explicit training on phonetic or phonemic targets. The ability to work with real speech is a major advance for cognitive models of HSR.     

How Experts Adapt Their Gaze Behavior When Modeling a Task to Novices

Domain experts regularly teach novice students how to perform a task. This often requires them to adjust their behavior to the less knowledgeable audience and, hence, to behave in a more didactic manner. Eye movement modeling examples (EMMEs) are a contemporary educational tool for displaying experts’ (natural or didactic) problem-solving behavior as well as their eye movements to learners. While research on expert-novice communication mainly focused on experts’ changes in explicit, verbal communication behavior, it is as yet unclear whether and how exactly experts adjust their nonverbal behavior. This study first investigated whether and how experts change their eye movements and mouse clicks (that are displayed in EMMEs) when they perform a task naturally versus teach a task didactically. Programming experts and novices initially debugged short computer codes in a natural manner. We first characterized experts’ natural problem-solving behavior by contrasting it with that of novices. Then, we explored the changes in experts’ behavior when being subsequently instructed to model their task solution didactically. Experts became more similar to novices on measures associated with experts’ automatized processes (i.e., shorter fixation durations, fewer transitions between code and output per click on the run button when behaving didactically). This adaptation might make it easier for novices to follow or imitate the expert behavior. In contrast, experts became less similar to novices for measures associated with more strategic behavior (i.e., code reading linearity, clicks on run button) when behaving didactically.     

Inherited and Environmental Moderators of Mother-Child Behavioral Contingency and Contingent Negativity at 27 Months

Previous research has shown that as depressive symptomology increases, mothers tend to show withdrawn or harsh interaction patterns with their children, and the quality of these interaction patterns have subsequently been linked to child behavior problems. However, little research has examined bidirectional influences between mothers and their children, and how these moment-to-moment contingencies differ based on heritable and environmental characteristics. We used data from the Early Growth and Development Study a prospective adoption study to examine how adoptive mothers’ depressive symptoms and children’s heritable tendencies for negative affectivity interact to predict the quality of mother-child interactions at child age 27-months. Results detected two distinct dyadic interaction patterns. The first was a withdrawn interaction style and was observed in children with a high heritable tendency for negative affect. In that style, mother and child interactions were not contingent upon each other, suggesting a lack of joint engagement. The second was a volatile interaction style, observed in children with a low heritable tendency for negative affect. In these cases, mother and child interactions were highly contingent but negative. Our findings demonstrate essential differences in how dyadic interaction patterns vary according to level of depressive symptomology and heritable tendency for negative affect.     

Examples,practice problems,or both? Effects on motivation and learning in shorter and longer sequences

Research suggests some sequences of examples and problems (i.e., EE, EP) are more effective (higher test performance) and efficient (attained with equal/less mental effort) than others (PP, sometimes also PE). Recent findings suggest this is due to motivational variables (i.e., self-efficacy), but did not test this during the training phase. Moreover, prior research used only short task sequences. Therefore, we investigated effects on motivational variables, effectiveness, and efficiency in a short (Experiment 1; four learning tasks; n = 157) and longer task sequence (Experiment 2; eight learning tasks; n = 105). With short sequences, all example conditions were more effective, efficient, and motivating than PP. With longer sequences, all example conditions were more motivating and efficient than PP, but only EE was more effective than PP. Moreover, EE was most efficient during training, regardless of sequence length. These results suggest that example study (only) is more effective, efficient, and more motivating than PP.     

Learning through observing: Effects of modeling truth‐ and lie‐telling on children’s honesty

The current study examined the influence of observing another's lie‐ or truth‐telling – and its consequences – on children's own honesty about a transgression. Children (N = 224, 5–8 years of age) observed an experimenter (E) tell the truth or lie about a minor transgression in one of five conditions: (a) Truth‐Positive Outcome – E told the truth with a positive outcome; (b) Truth‐Negative Outcome – E told the truth with a negative outcome; (c) Lie‐Positive Outcome – E lied with a positive outcome; (d) Lie‐Negative Outcome – E lied with a negative outcome; (e) Control – E did not tell a lie or tell the truth. Later, to examine children's truth‐ or lie‐telling behavior, children participated in a temptation resistance paradigm where they were told not to peek at a trivia question answer. They either peeked or not, and subsequently lied or told the truth about that behavior. Additionally, children were asked to give moral evaluations of different truth‐ and lie‐telling vignettes. Overall, 85% of children lied. Children were less likely to lie about their own transgression in the TRP when they had previously witnessed the experimenter tell the truth with a positive outcome or tell a lie with a negative outcome.     

Using the UTAUT2 model to explain public acceptance of conditionally automated (L3) cars: A questionnaire study among 9,118 car drivers from eight European countries

We investigated public acceptance of conditionally automated (SAE Level 3) passenger cars using a questionnaire study among 9,118 car-drivers in eight European countries, as part of the European L3Pilot project. 71.06% of respondents considered conditionally automated cars easy to use while 28.03% of respondents planned to buy a conditionally automated car once it is available. 41.85% of respondents would like to use the time in the conditionally automated car for secondary activities. Among these 41.85%, respondents plan to talk to fellow travellers (44.76%), surf the internet, watch videos or TV shows (44%), observe the landscape (41.70%), and work (17.06%). The UTAUT2 (Unified Theory of Acceptance and Use of Technology) was applied to investigate the effects of performance and effort expectancy, social influence, facilitating conditions, and hedonic motivation on the behavioural intention to use conditionally automated cars. Structural equation analysis revealed that the UTAUT2 can be applied to conditional automation, with hedonic motivation, social influence, and performance expectancy influencing the behavioural intention to buy and use a conditionally automated car. The present study also found positive effects of facilitating conditions on effort expectancy and hedonic motivation. Social influence was a positive predictor of hedonic motivation, facilitating conditions, and performance expectancy. Age, gender and experience with advanced driver assistance systems had significant, yet small (<0.10), effects on behavioural intention. The implications of these results on the policy and best practices to enable large-scale implementation of conditionally automated cars on public roads are discussed.     

The Process Through Which an Advocacy Intervention Resulted in Positive Change for Battered Women Over Time

A prior experimental evaluation of a community-based advocacy program for women with abusive partners demonstrated positive change in the lives of women even 2 years postintervention (C. M. Sullivan & D. I. Bybee, 1999). The current study explored the complex mediational process through which this change occurred, using longitudinal structural equation modeling and formal tests of mediation. As hypothesized, the advocacy intervention first resulted in women successfully obtaining desired community resources and increasing their social support, which enhanced their overall quality of life. This improvement in well-being appeared to serve as a protective factor from subsequent abuse, as women who received the intervention were significantly less likely to be abused at 2-year follow-up compared with women in the control condition. Increased quality of life completely mediated the impact of the advocacy intervention on later reabuse. Discussion places advocacy for women in the context of other efforts that are needed to build an effective community response to preventing intimate violence against women.     

Beyond bruises and broken bones: the joint effects of stress and injuries on battered women's health

We investigated the joint mediating effects of injuries and stress on the relationship between abuse and women's health. A community sample of 397 women, half of whom had been assaulted by an intimate partner within the prior 6 months, was interviewed about their experience of intimate partner violence, injuries, stress, income, depression, and physical health problems. Structural equation modeling techniques confirmed the complex model of hypothesized indirect effects of abuse on women's physical health problems through injuries, stress, and depression. Stress accounted for 80% of the indirect effect of abuse on women's physical health. Its direct effect on physical health was somewhat larger than its indirect effect through depression, but both processes played a key role in determining the effect of abuse on women's physical health problems. Furthermore, abuse was a stronger predictor of women's stress than was poverty. Implication and future research are discussed.     

The competition-among-relations-in-nominals theory of conceptual combination: implications for stimulus class formation and class expansion

One way in which new concepts are added to the conceptual system is through conceptual combination. The competition-among-relations-in-nominals (CARIN) theory (Gagné & Shoben, 1997) proposes that conceptual combination involves specifying a thematic relation (e.g., noun MADE OF modifier) to link the constituent concepts (e.g., chocolate and bee). This theory claims that relations have different strengths for various concepts that correspond to how often a modifier and relation have been paired in previous encounters with combined concepts and that this relational knowledge strongly affects the ease with which combined concepts can be formed. A mathematical model that incorporates key claims of the theory is presented, and empirical findings that are relevant to evaluating the CARIN theory are reviewed. The parallels between the CARIN theory and approaches to stimulus class formation are also discussed.     

Facilitative orthographic neighborhood effects: the SERIOL model account

A large orthographic neighborhood (N) facilitates lexical decision for central and left visual field/right hemisphere (LVF/RH) presentation, but not for right visual field/left hemisphere (RVF/LH) presentation. Based on the SERIOL model of letter-position encoding, this asymmetric N effect is explained by differential activation patterns at the orthographic level. This analysis implies that it should be possible to negate the LVF/RH N effect and create an RVF/LH N effect by manipulating contrast levels in specific ways. In Experiment 1, these predictions were confirmed. In Experiment 2, we eliminated the N effect for both LVF/RH and central presentation. These results indicate that the letter level is the primary locus of the N effect under lexical decision, and that the hemispheric specificity of the N effect does not reflect differential processing at the lexical level.