Emotional reasoning is emotionally-based cognition operating on subjective terms, independent of rationality, and using feelings as criteria. In Study 1, 113 participants, 29 men and 84 women, focussed either on their feelings or on the reasons for their choices as they made a series of decisions in a card-playing game. Contrary to predictions, participants who focussed on their feelings were less likely to make optimal decisions in the game. This study thus provided no evidence that emotional reasoning can reach optimal conclusions. In a second study, 96 participants, 35 men and 61 women, made a series of decisions to cooperate or compete with an unseen partner in a Prisoner's Dilemma game. Participants who focussed on their feelings were less likely to compete than participants who focussed on reasons for making each decision, so that they did less well in the short term, but significantly better over the long term, than participants in the Reasons condition. Participants who described themselves as highly rational were also less likely to do well in the game. This study demonstrates that emotional reasoning can be more effective than rational decision-making.

Three experiments investigated whether morphological constituents influence word processing during reading. Individuals read sentences containing free stem, bound stem, and pseudo-prefixed words. In Experiments 1 and 2, a parafoveal display change manipulation indicated that the morphological constituents of a prefixed word are not available for preprocessing in the parafovea as reading times on the target word did not differ for prefixed versus pseudo-prefixed words. Interestingly, parafoveal preview of word initial and word final letters resulted in an equivalent amount orthographic facilitation for all word types. In Experiment 3, a fast priming manipulation indicated that morphological priming effects for prefixed words are obscured during sentence processing. However, the form of the prime did facilitate subsequent word processing for all three word types. The results suggest that English prefixed words are accessed via their whole word form, as there was no evidence of morphological decomposition for prefixed words during sentence processing.

This dissertation addresses the general question of how children's phonological categories in perception differ from adults' phonological categories. A well-known characteristic of adults' phonological categories is the existence of trading relations and integration among cues to a phonological category. In trading relations, less evidence from one cue to a phonological category can be offset by more evidence from another cue. When cues are integrated by the listener, the cues are not used independently; instead the perceptual value of one cue depends on the value of another. Little research has previously been carried out into the development of trading relations and integration in children's speech perception. This dissertation tests two possible factors affecting the development of adultlike trading relations: the distribution of cues in the dialect or language variety being learned, and the development of integration. The results of identification experiments with adult and four- to six-year-old child listeners from different dialects (Standard American English, African American English, and Australian English) suggest that children's trading relations develop to approximate those of the adults in their dialect group. Children's trading relations differ, however, in that children make relatively heavier use of one cue to a contrast. The results of a discrimination experiment suggest that integration of vowel duration and first formant (F1) cues to final stop voicing in Australian English develop gradually in childhood. The implications of this finding are discussed with respect to gesturalist, auditorist, and associative views of trading relations. The finding suggests that whether or not integration for some trading relations results from the recovery of a gesture or the experience of an auditory property, integration for some subtypes of trading relations, such as context effects, may develop as a result of learning from experience with patterns of acoustic covariation among cues to phonological categories.

This dissertation investigates the ways in which speech perception is guided by the expectation that the stimulus is an utterance in the perceiver's language, with a particular focus on how phonotactics affects the interpretation of acoustically ambiguous segments. A model is proposed in which phonological grammar, expressed here as a system of ranked and violable constraints within the framework of phonological Optimality Theory, is used to select among competing candidate parses of the acoustic input. This grammar-based theory is contrasted with two grammarless alternative accounts of perception: the connectionist network TRACE, which derives phonotactic perceptual effects from the lexicon, and a statistical theory based on transitional probabilities. Experimental evidence is presented to show (1) that English listeners' judgments of vowels and of consonant clusters disfavor configurations which are grammatically illegal in the language, (2) that the dispreference for illegal configurations is far stronger than that for configurations which are legal but have zero frequency, and (3) that it is due to a response dependency, rather than to auditory or other stimulus factors, and cannot be explained by foreign-language exposure. Two experiments with Japanese listeners find that (1) the lexical stratum membership of nonsense words can produce a phonotactic perceptual effect, (2) that the triggering and target segments can be up to three segments distant, and (3) that the stratum-phonotactic effect is larger than a word-superiority effect obtained with the same listeners and paradigm. These results are shown to be consistent with the grammar-based model, but inconsistent with the two grammarless alternatives. Analysis of the three models reveals that the shortcomings of the alternatives is due to their inability to abstract over phoneme classes and larger linguistic structures. It is concluded that the mechanisms of speech perception have access to a full-fledged phonological competence.

This work presents an investigation of how prosodic information is used in natural language processing and how prosody should be incorporated into models of sentence comprehension. It is argued that the processing system builds a prosodic representation in the early stages of processing, and is guided by this prosodic representation through multiple stages of analysis. Specifically, the results of four sentence comprehension experiments demonstrate that prosodic phrasing influences syntactic attachment decisions, focus interpretation, and the availability of contextual information in the resolution of lexical ambiguity. Two explicit hypotheses of how prosodic structure is used in processing are proposed to account for these effects: one which accounts for effects of phonological phrasing on syntactic processing decisions and a second which accounts for effects of intonational phrasing on semantic/pragmatic interpretation. Three sources of evidence are provided in support of the central claim that the processor must build and use a prosodic representation from the early stages of processing. First, an experiment on the resolution of prepositional phrase attachment ambiguity demonstrates that syntactic attachment decisions are influenced by the overall pattern of phonological phrasing in utterance, and not simply by prosodic boundaries located at the point of syntactic ambiguity. Thus, the effects of a single kind of prosodic element, at a single level in the prosodic hierarchy, must be accounted for with respect to the larger prosodic structure. A second experiment shows that the interpretation of focus is dependent on both the pattern of pitch accents in the utterance and the pattern of prosodic phrasing, establishing that different kinds of prosodic elements in the prosodic structure are used jointly in processing decisions. Two additional experiments, one on the interpretation of context-sensitive adjectives and a second on the resolution of within-category lexical ambiguity, demonstrate that phonologically distinct levels of prosodic phrasing have separable effects on language processing. Taken together, the four experiments suggest that prosody has a much broader role in sentence comprehension than previously recognized, and that models of sentence processing should be modified to incorporate prosodic structure.

Research on the relationship between fear and persuasion led to a proliferation of conflicting results. The purpose of this project was to develop and test hypotheses regarding how fear might impact the persuasion process delineated by the Elaboration Likelihood Model. Fear could direct message recipients into the central or peripheral route by motivating or distracting them from extensive message processing. Fear could also serve as a peripheral cue. It was hypothesized that fear arousal relevant to a persuasive message would motivate subjects to carefully process the message, therefore, central route processing would occur and the amount of persuasion would be based on message quality. It was further hypothesized that fear arousal irrelevant to a persuasive message would distract subjects from attending to the message so that persuasion would occur in the peripheral route and not be based on message quality. Fear arousal was also expected to act as a peripheral cue, enhancing persuasion by its mere presence. A 3 (fear arousal: relevant, irrelevant, none) x 3 (message quality: strong, weak, minimal) x 2 (topic: heart disease, peptic ulcers) design was used. A secondary goal of this study was to develop a methodology to arouse fear separate from a persuasive message to avoid the confounding variable problem present in other fear appeal research. This was successfully accomplished. As predicted, there was a marginally significant effect demonstrating that relevant fear resulted in a greater disparity between strong and weak messages than no fear arousal for one of the topics. Contrary to predictions, irrelevant fear arousal did not result in smaller differences in persuasion when compared with no fear arousal. There was a marginally significant effect that relevant fear arousal produced greater intentions than no fear arousal in the minimal message condition for one of the topics. The results provided partial support for the hypotheses that fear can motivate extensive message processing and can serve as a peripheral cue. There was no evidence that irrelevant fear distracted from extensive processing. Implications of these results and possibilities for future research are discussed.

The purpose of this exploratory study was to investigate Chinese junior college students' condom use intentions to prevent HIV/AIDS. The following research question was addressed: How is knowledge about HIV/AIDS risk, attitudes, subjective norms, self-efficacy, prevention expectancies, and prior condom use related to condom use intention among college students? Questionnaires and semi-structured interviews were employed in this study. Variables for the survey were selected from the Theory of Reasoned Action and the Self-Efficacy Theory. A questionnaire was modified and translated from an English instrument and back-translated to ensure cultural relevance and sensitivity. In addition, twenty students were purposely selected for interviewing among those who indicated they had had sexual encounters. Of all the students in the survey sample, 29% males and 9% females admitted having sexual experiences. Data analysis was based on the subsample (N = 302) who had sexual experience. The subjects' knowledge of AIDS-related risk behavior produced an average score of 68% correct responses based on true-false questions. Knowledge, however, was not significantly related to any of the other variables. Logistic regression analysis was employed to analyze students' intentions for condom use to prevent HIV/AIDS. For all students, prior condom use was the most important determinant of intention for future condom use. For male students, the estimated odds ratio of prior condom use and self-efficacy were significantly associated with intention to use condoms. This finding suggests that the actual practice to master the skill of condom use is critical. For female students, prior condom use and attitudes significantly predicted the intention for condom use. Themes from the interview data were students' general HIV/AIDS knowledge, sources of sex education, perceived risk of acquiring HIV/AIDS, strategies for managing risk, discussion content and the influences of parents, siblings, friends, and sexual partners on students' personal norms. The implication from the interview data was that prevention of pregnancy was the most important factor for students to use condoms, rather than fear of contracting STDs or HIV/AIDS. Thus, suggestions for education programs are discussed.

Auditors are often required to justify their judgments to more senior members of their firms. However, the documentation required to justify a judgment can differ. While it is possible that different justification requirements may produce different effects on judgment processes, little is known about the relative effectiveness of different justification techniques. This study uses a control environment assessment task to examine the effects of alternative justification techniques on auditors' (reviewees') judgments, on the information transferred from the reviewee to the reviewer, and on the resulting judgment of the reviewer. Professional auditors were given a detailed case concerning a client's control environment which was based on an actual audit in which fraud was present. Auditors were assigned to be either a reviewee or reviewer based on their experience level. Reviewees were divided into three justification groups: supporting, balanced, or component. They were asked to make a series of assessments relating to the control environment's ability to prevent fraud, and to justify those assessments using their respective justification requirements. Reviewers were paired with a reviewee and asked to review that individual's work. Reviewers received the reviewee's judgment and justification, along with the information contained in the detailed case, and made the same series of control environment assessments as reviewees. The main variables examined were decision accuracy and the amount and type of evidence documented in the reviewee's justification. The results indicated that reviewees in the component group documented a larger evidence set in their justifications than did reviewees in the balanced and supporting groups, and that the component justification reviewees were the least accurate in their fraud risk assessments. Reviewer results tended to mirror those of reviewees, though comparisons between component and supporting reviewers were not as strong. One potential explanation for the underperformance of component group subjects concerns the dilution, effect. The substantial amount of evidence documented by component reviewees may have diluted the impact of key information that pointed to possible fraud problems. These results suggest that, when performing a reasonably complex evaluative task, the way in which reviewees are required to justify their assessments can affect both the accuracy of their assessments and the information they document as part of their justifications. Given that both the reviewee's assessment and justification are later passed on for review, different justification techniques can affect the accuracy of reviewers' assessments. Implications for audit practice and for decision making research are discussed.

The present study investigates whether consonant and vowel information make different contributions at the early stage of visual word recognition. In linguistics, the distinction between consonants and vowels has been formalized in the modern theory of phonology, which assumes separate representation and processing of consonants and vowels (e.g., Clements & Keyser, 1983; Durand, 1990; Goldsmith, 1990). In cognitive psychology, such a distinction has been suggested generally in the notion of “islands of reliability”, regarding consonant information as more reliable backbones of word processing at the early stage (Carr & Pollatsek, 1995; Perfetti & McCutchen, 1992; Brown & Besner, 1987). This notion has been more specifically formulated in the two cycles model of phonology assembly (Berent and Perfetti, 1995). The purpose of the present study was to investigate the relative contribution of consonants and vowels in word recognition in silent reading through two different experimental paradigms. In the missing letter paradigm used in Experiments 1 and 2, fixation times on a word during reading were examined as a function of the condition and duration of the missing letter. If consonants have an earlier advantage over vowels in word processing, then missing a consonant should yield a greater cost (i.e., longer fixation times) than missing a vowel. In the fast priming paradigm used in Experiment 3, the relation of the prime and target words was manipulated in terms of consonant and vowel similarity. The consonant-vowel distinction was then estimated in the different priming effects as a function of prime-target relation. The results from the three experiments in this dissertation, in line with the findings of Berent and Perfetti (1995), strongly indicate that there is a clear temporal distinction between consonants and vowels in their contribution to word identification in normal reading. Consonants plays a more important role than vowels in the early stage of word processing. This was demonstrated in Experiments 1 and 2 by the longer fixation time on the target word when the consonant was missing than when the vowel was missing for a brief duration (30 ms) from the onset of the word, and was demonstrated in Experiment 3 by the shorter fixation time on the target word when the prime word was more similar to the target word in terms of consonant similarity than in terms of vowel similarity at the 30 ms prime duration. Further, the results from Experiment 3 suggest that lexicality plays a significant role in the consonant-vowel distinction: the early consonant-vowel distinction was observed only when the prime was a high frequency word.

Previous research has found that when a word like “clam” is presented at the foveal level, its transposed-letter neighbor “calm” is also activated. This activation of multiple lexical candidates causes interference in naming and lexical decision tasks. Here, four eye-tracking experiments were conducted to explore the nature of transposed-letter (TL) neighborhood effects within the context of normal silent reading. Experiments 1 and 2 addressed the processing of target words that have a transposed-letter neighbor (e.g., angel, angle) in comparison to target words that do not have a transposed-letter neighbor (e.g., alien, slope ). Furthermore, Experiment 2 manipulated the sentence context leading up to the target word to explore whether semantic constraints can attenuate neighborhood interference effects. The results indicated that readers do take longer to process words that have a TL neighbor than words that do not have a TL neighbor. This interference effect, however, disappeared when the beginning of the sentence was constraining such that only one of the two members of the TL pair was likely. While both Experiments 1 and 2 explored effects at the foveal level, Experiments 3 and 4 explored the parafoveal processing of transposed-letter neighbors by employing an eye-movement-contingent boundary change paradigm. In Experiment 3, readers received a parafoveal preview of a TL target word that was either (1) identical to the target word (e.g., calm as the preview of calm), (2) a TL-neighbor (e.g., clam) or (3) a substituted-letter (SL) control (e.g., chem). In Experiment 4, a further set of parafoveal preview conditions was explored. Across both experiments, readers' fixation durations on the target words were significantly longer when the parafoveal previews were substituted-letter nonwords than when they were TL neighbors, suggesting that TL neighbors (when presented in the parafovea) facilitate word recognition, rather than inhibit processing. Collectively, these experiments indicate that TL neighborhood interference effects do occur in normal silent reading, but these effects occur late and are influenced by sentence constraints.