Using reinforcement learning to understand the emergence of "intelligent" eye-movement behavior during reading

The eye movements of skilled readers are typically very regular (K. Rayner, 1998). This regularity may arise as a result of the perceptual, cognitive, and motor limitations of the reader (e.g., limited visual acuity) and the inherent constraints of the task (e.g., identifying the words in their correct order). To examine this hypothesis, reinforcement learning was used to allow an artificial "agent" to learn to move its eyes to read as efficiently as possible. The resulting patterns of simulated eye movements resembled those of skilled readers and suggest that important aspects of eye-movement behavior might emerge as a consequence of satisfying the constraints that are imposed on readers. These results also suggest novel interpretations of some contentious empirical results, such as the fixation duration costs associated with word skipping (R. Kliegl & R. Engbert, 2005), and theoretical assumptions, for example the familiarity check in the E-Z Reader model of eye-movement control (E. D. Reichle, A. Pollatsek, D. L. Fisher, & K. Rayner, 1998).     

Sticks and Stones May Break My Bones but Words Relate to Adult Physiology? Child Abuse Experience and Women's Sympathetic Nervous System Response while Self-Reporting Trauma

Child abuse predicts a wide range of long-lasting deleterious outcomes, including disruptions in the biological systems central to emotion arousal and regulation. However, little is known about the specific ways in which child abuse affects adulthood sympathetic reactivity and recovery. This study investigated the association between child abuse experience and adult skin conductance level and habituation in 85 at-risk women as they completed a self-report trauma questionnaire. Childhood emotional abuse was independently associated with blunted skin conductance habituation over the course of survey completion after controlling for other abuse subtypes and current trauma symptoms. These results suggest that women emotionally abused as children experience prolonged emotional arousal and poor physiological regulation of emotion in response to reminders of traumatic experiences.     

Mnemonic differences and similarities across opposing social groups: The linguistic conflict at the University of Leuven as a case study

In the 1960s, a historical event occurred at one of Europe's most prestigious universities: The Dutch-speaking students forced the French-speaking students to relocate and establish their own university. We compared the extent to which members of each social group developed elaborate memories of the events surrounding the conflict and whether they were associated with differences in rehearsal type (media, conversational, rumination) and initiating conditions (importance, political engagement, and negative/positive emotions). All participants were university students at the time of the conflict. We found that Dutch-speakers exhibited more elaborate memories compared to French-speakers and that importance was associated with elaborate memories only for the Dutch-speakers. However, positive emotions appear to be critical in the formation of elaborate memories across the social groups. We found no such associations for negative emotions. We discuss these results in terms of the social/cognitive processes transcending social group membership in understanding how individuals remember past conflicts.     

Does European citizenship breed xenophobia? European identification as a predictor of intolerance towards immigrants

The European Union is generally perceived as endorsing universalistic and multi‐cultural values. However, social identity and self‐categorization theories predict that, when certain conditions are met, a negative relation between ingroup identification and tolerance towards outgroup members should be observed. We argue that the creation of the status of ‘Citizen of the Union’ in Maastricht may contribute to meeting those conditions and therefore to increase intolerance towards resident foreigners. If that is the case, a paradoxical situation could emerge, in which people's levels of tolerance towards foreigners would contradict group values. We examined the relations between values associated with Europe, European and national identification, and tolerance towards foreigners through a survey study with a—non‐representative—sample of undergraduate French‐speaking Belgian students. Results show that Europe was generally associated with humanistic values. But they also reveal that strong European identifiers tended to express more xenophobic attitudes than weak European identifiers, whilst national identification was not related with such attitudes. Copyright © 2002 John Wiley & Sons, Ltd.     

Limits of rhythm perception

To what extent are listeners sensitive to the time intervals separating non-consecutive events in sound sequences? The subjects of Experiment 1 were presented with sequences of 20 identical tones in which the 10 odd-numbered tones or the 10 even-numbered tones made up an isochronous sub-sequence (with a periodicity of 0.5-1 s) whereas the other tones, acting as distractors, occurred at random moments. Such sequences appeared to be very difficult to discriminate from sequences without any timing regularity, which revealed a lack of perceptual sensitivity to their "second-order" intervals. Experiment 2 employed repetitive sequences in which the first-order intervals (separating consecutive tones) took two possible values, forming a ratio that subjects had to classify as larger or smaller than 2. The results of this experiment suggest that subjects were able to make use of second-order intervals in their task, but mainly due to the predictable nature of the sequences; the relative positions of subjective accents (Povel & Essens, 1985) had no significant effect on performance. It is concluded that the perception of subtle timing details in "ordinary" music may rest on nothing more than a sensitivity to the relations between first-order intervals (within a given auditory stream).     

Traffic Offences: Planned or Habitual? Using the Theory of Planned Behaviour and habit strength to explain frequency and magnitude of speeding and driving under the influence of alcohol

This study addresses the socio‐cognitive determinants of traffic offences, in particular of speeding and drinking and driving. It has two aims: (1) to test the hypothesis of a direct effect of habits on offences (i.e., independent of intentions) by employing a specific measure of habits (i.e., the SRIH) and (2) to analyse the offences by taking account of three distinct parameters: Frequency, usual magnitude (i.e., the most frequent deviation from the law) and maximal magnitude (i.e., the greatest deviation occasionally adopted) in order to represent more accurately the variability of the offending behaviours. A total of 642 drivers replied to a questionnaire. The results corroborate the idea that intention and habit are distinct and direct determinants of offences. The use of the SRIH dismisses the criticisms made with regard to the measure of past behaviour. The distinction between the three behavioural parameters proves to be relevant, as their determinants are not exactly similar. Finally, attitude and subjective norm had direct effects on the maximal magnitude and/or on the frequency of the offence. The discussion concerns the contribution of this study to the analysis of offences as well as its limitations and addresses the theoretical plausibility of the direct effects of attitude and the subjective norm.     

Multimedia mug books: how multi should the media be?

The impact of allowing witnesses to choose the type of cues presented in multimedia mug books was explored in two experiments. In Experiment 1, participants viewed a videotaped crime and attempted to identify the perpetrator from one of three types of mug books: (a) dynamic‐combined—participants could choose to follow static mug shots with a computerized video clip combining three types of dynamic cues: the person walking, talking, and rotating; (b) dynamic‐separable—participants could limit the types of dynamic cues presented; and (c) static—just the static mug shot was presented. The dynamic‐separable condition produced significantly fewer false positive foil identifications than the static condition. Within the dynamic‐separable condition, voice was the most preferred cue. Experiment 2 explored the contribution of the individual cues. Participants attempted identifications from single dynamic cue mug books where only one type of cue was presented if a participant chose additional information. It was found that providing individual cues did not improve performance over the static mug book control. Based on the potential danger of witnesses choosing to rely on single dynamic cues, it was suggested that multimedia mug books should present dynamic cues in combination. Copyright © 2000 John Wiley & Sons, Ltd.