Answer comparison in automated question answering

In the context of Question Answering (QA) on free text, we assess the value of answer comparison and information fusion in handling multiple answers. We report improvements in answer re-ranking using fusion on a set of location questions and show the advantages of considering candidates as allies rather than competitors. We conclude with some observations about answer modeling and evaluation methodology, arising from a more recent experiment with a larger set of questions and a greater diversity of question types and candidates.     

Cross-examination of the expert witness: Do hostile tactics affect impressions of a simulated jury?

Ninety-one introductory psychology students viewed a videotape of a simulated personal damage trial in which a psychologist expert witness testified about a plaintiff's psychological problems resulting from trauma. The defense attorney's cross-examination tactics, employed against the expert witness, were varied in the videotape. A 2 × 2 factorial design was used, with leading versus nonleading questions and a hostile versus nonhostile attorney. An interaction effect was obtained, in which a hostile attorney using leading questions and a nonhostile attorney using nonleading questions were perceived as least effective.     

Characteristics of male partners of adult female incest survivors

The interpersonal style of male partners of adult female incest survivors was investigated. Two groups of men (partners of incest survivors, n=30; and a control group, n=30) completed the Narcissistic Personality Inventory and the Depressive Experiences Questionnaire. A multivariate analysis of variance failed to support clinically-based hypotheses (a) that partners of incest survivors would exhibit a greater sense of entitlement in interpersonal relationships and (b) that a larger number of men abused in childhood would be in relationship with incest survivors than those not abused. Results point to a need for further controlled quantitative studies of the characteristics of male partners of female incest survivors. This article is based on the Master's Thesis of the first author submitted to the Graduate School of Texas Woman's University.     

Buffalo Creek Survivors in the Second Decade: Comparison with Unexposed and Nonlitigant Groups1

This study investigated the second-decade effects of the Buffalo Creek dam collapse and flood that occurred in West Virginia in 1972. One hundred twenty-one survivors who had participated in an earlier lawsuit against the coal company that built the collapsed dam were compared to 78 nonlitigant survivors on self-reported symptoms, clinical ratings, and diagnoses. A nonexposed sample from a geographically and culturally similar neighborhood was investigated as well. Findings showed no differences between the litigant and nonlitigant survivor groups. The survivor groups together showed higher rates of anxiety, depression, and hostility symptoms and diagnoses than the nonexposed sample. The findings were discussed in the context of the nature of the traumatic event, social and cultural influences on recovery, and the constellation of symptoms which differentiated the groups.     

Progamming time in serial tapping responses as a function of pathway constraint

Summary This study varied the accuracy demand within a linear series of targets to investigate the effect of movement-pathway constraints on response-programming time. Sidaway, Christina, and Shea (1988) have suggested that constraints placed upon movement initiation by the demand for response precision may play an important role in determining the length of the programming process. By varying the subtended angles of a series of three targets, this experiment tested the specific prediction of Sidaway et al. that programming time may be a function of the target, within a line of targets, that subtends the smallest angle at the start position. It is this target that demands the greatest precision in the movement pathway. Subjects participated in a series of conditions in which the size and placement of the target that imposed the maximal constraint was varied. In each condition the subjects were required to strike a series of three targets with a stylus in a simple reaction-time paradigm. Analysis of the reaction-time results revealed a significant effect of size of constraint, but no effect of position of constraint. Analysis of the movementtime data dispelled movement-duration and movement-velocity interpretations of the results and intimated a possible online trajectory-correction process.Portions of these data were first presented at the Annual Meeting of the Psychonomic Society in Atlanta, Georgia, November 1989.     

Explicit syllable and phoneme segmentation in the young child

To write a language, one must first abstract the unit to be used from the acoustic stream of speech. Writing systems based on the meaningless units, syllables and phonemes, were late developments in the history of written language. The alphabetic system, which requires abstraction of the phonemic unit of speech, was the last to appear, evolved from a syllabary and, unlike the other systems, was apparently invented only once. It might therefore be supposed that phoneme segmentation is particularly difficult and more difficult, indeed, than syllable segmentation. Speech research suggests reasons why this may be so. The present study provides direct evidence of a similar developmental ordering of syllable and phoneme segmentation abilities in the young child. By means of a task which required preschool, kindergarten, and first-grade children to tap out the number of segments in spoken utterances, it was found that, though ability in both syllable and phoneme segmentation increased with grade level, analysis into phonemes was significantly harder and perfected later than analysis into syllables. The relative difficulties of the different units of segmentation are discussed in relation to reading acquisition.