Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Computer-generated characters, commonly used as avatars and agents, present a broad range of appearance and types. The growing body of research on avatars and agents does not yet provide a categorical and dimensional framework for understanding people’s perception of computer characters. Two studies were conducted to explore people’s categorization of computer characters and dimensions used in the categorization. In Study 1, the participants (N = 40) sorted the images of 70 computer-generated characters into groups. Cluster analysis revealed that human and nonhuman emerged as two large types. Group labels provided by the participants were analyzed to help interpret subcategories within each of the two clusters. In Study 2, a different sample of participants (N = 273) rated the same 70 characters on a set of attributes. The ratings were used in regression analysis to help interpret the multidimensional scaling analysis of the sorting data in Study 1. Four dimensions were identified: humanness, graphic details and stylization, good-natured vs. bad-natured, and gender. The results provide an initial categorical and dimensional framework for further research on perception and effects of computer characters.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: This presentation focuses on a recent case in which 52 gays were prosecuted in Egypt for their alleged debauchery. It will focus on three interrelated aspects of the case. First, it provides a detailed description of the internal organization of the written ruling considered as a text. Second, it offers an in-depth analysis of excerpts of the transcribed investigations which were conducted by the public prosecutor on some of the accused people. Third, it focuses on the categorizations processes and devices operating at the many levels of both the written ruling and the transcribed investigations. As a whole, this presentation will contribute to the study of the possible effects of the embedment of the discourse addressing moral issues in a judicial setting. In that sense, it will shed some light on what I call procedural constraint and legal relevance, i.e. the orientation of members of a legal encounter toward the pre-allocated structure of the judicial proceedings and toward the necessity to produce characterizations that may sound acceptable by their over-hearing and over-reading audiences. It will show that the grammar of morality is thoroughly practical and closely related to the setting in which categorizations are produced, independent from the commonly heard postulates concerning the so-called Arab Islamic culture. Finally, it will show how much law and morality are practically intertwined and contingently produced, i.e. how the rule of law is locally achieved.

Publication Type: Individual Poster
Abstract: David Rakison and colleagues examined the use of object parts to from categories (e.g., Rakison & Butterworth, 1998). They tested 14- to 24-month-olds with a combination of animal, vehicle and hybrid animal-vehicle toys, using a sequential touching task and found that infants’ apparent use of parts to form categories decreased with age: 14-month-olds always formed categories on the basis of parts, 18-month-olds sometimes formed categories on the basis of parts and sometimes on the basis of taxonomic structures, whereas 24-month-olds always formed categories on the basis of taxonomic kind (bodies). We examined the extent to which 18-month-olds would adapt their categorization of normal and hybrid stimuli as a function of context.

Fifty-two infants were tested using sequential touching on a set of 8 toys that could be partitioned as containing (1) the global level of animal, (2) the basic level of car, (3) objects having wheels, (4) objects having legs, (5) hybrid objects, or (6) as normal objects. Infants were randomly assigned to one of two prior context conditions in which they were shown some human dolls and trucks. In the Partonomic condition, the experimenter began by pointing out parts of the toys to the infants by bending the dolls at the waist and legs or by turning the trucks upside down and spinning their wheels. In the Taxonomic condition, the experimenter introduced the toys by “walking” the dolls (without bending the dolls at the waist or their legs) or by “driving” the trucks along the table.

Touch patterns were analyzed by fitting a finite mixture model to the data (Thomas and Dahlin; 2001). Many individual infants were found to categories in multiple ways. We then tabulated the number of infants that could be described as categorizing by parts, by taxonomic category, or by both. In the Taxonomic condition 22 categorized taxonomically, 2 categorized by parts, and 1 categorized both. In the Partonomic condition, 2 categorized taxonomically, 10 categorized by parts, and 0 categorized both. Categorization strategy was contingent on the familiarization context (Chi-squared (2) = 21.0, p<. 0001). The distribution of categorizers in the Partonomic context differed from chance, Chi-squared (2) = 14.0, p<. 001), with most of these infants forming categories on the basis of parts. The distribution of categorizers in the Taxonomic condition also differed from chance (Chi-squared (2) = 29.4, p<. 0001), with most of these infants forming categories on the basis of taxonomic information.

Publication Type: Individual Poster
Abstract: In a study investigating infant categorization, Slater and Morison (1987) presented newborns sequentially with several exemplars of a given shape in a fixed-trial familiarization procedure. When this was followed by visual preference test trials, newborns did not show a significant difference in looking time for either the novel or the familiar stimulus. Thus habituation does not reveal any evidence of form categorization in newborns. We, in essence, replicated this study using an operant sucking paradigm. The operant sucking paradigm facilitates the study of newborn perceptual abilities by permitting a precise and simple way of obtaining habituation-like data (e.g. Aldridge et al., 2001). As in the habituation paradigm, the infant becomes familiar with the repeated presentation of a stimulus. Differential attention to the novel stimulus demonstrates discrimination and visual recognition memory. However, whereas habituation refers to the passive decrease in responsiveness that occurs when a stimulus becomes familiar, the operant sucking procedure relies on the active self-presentation of the stimuli contingent on sucking behavior. This distinction may be the reason we did find evidence of form categorization by newborns in the operant study. We found that newborns could categorize circles versus squares, and triangles versus crosses. Our claim depends on a between-subject analysis, focusing on novelty preference rather than on a decrement in response as is found in habituation. There was no within-subject evidence of habituation but the between-subject novelty effect was significant. Other well-known studies have relied on a similar statistical analysis (eg., Eimas et al.,1971).
To untangle the discrepancy in results found in the aforementioned research, we conducted a study of pure habituation. Newborns were repeatedly presented with a simple geometric shape (a cross, a circle, a square, or a triangle) for a 25-second period with by a 3-second interstimulus interval. Newborn-looking during the stimulus presentation periods was measured over the duration of the study. There was no evidence at all of habituation over the 14.53 minutes of the study. Yet, Slater, Morison, and Rose (1983) found that infants with an average age of 3 days readily habituated to each of the four stimuli. The major question that we must address is how to reconcile the incompatible conclusions that the different indicators suggest. Our newborns were only hours old and it may be that they are unlikely to be bored by anything we show them. Habituation is, after all, a manifestation of boredom.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: By conducting an experiment in South Korea and America, this study incorporated social categorization and cross-cultural factors into the third-person effect research. The greater third-person effect on out-group members, compared to in-group members, was only applicable to Americans. South Koreans did not show a greater third-person effect on out-group members than on in-group members. This discrepancy of social categorization explanations between South Koreans and Americans was explained by cross-cultural characteristics. South Koreans’ strong collectivistic culture weakens the tendency to overestimate a negative message on others. Americans’ third-person effect on comparison targets was always greater than South Koreans’ third-person effect.