Made available in DSpace on 2016-04-04T18:29:27Z (GMT). No. of bitstreams: 1 Sebastiao Elyseu Junior 2.pdf: 1309648 bytes, checksum: 780dfab0e8a1fc758aa3bbf94ec05941 (MD5) Previous issue date: 2006-05-12 / This work deals with the building and validation of a scale of ownership of self-report, based on clinical experience and in an ethological ownership theory, that postulates it as a instinctive tendency of the individual to achieve and/or mantain the dominion of one or more figures of ownership by means of instinctive patterns of ownership behaviour, regulating the provision for self usufruct and/or of the offspring. This first version was composed of 25 items, related to the ownership of the two general types of figures of ownership: people and things. These items allow the subject to indicate the degree of intensity of the answers, in a 4-point Likert scale type (1-weak, 2-medium, 3- strong, 4-excessive). The phase of building included the semantical and theoretical analysis of the items by university students and judges, respectively. The level of agreement among five independent judges in relation to its theoretical pertinence was 90% and Kappa = 0,88, being, thus, high. The sample of the test was constituted of university students (N= 296, 140 male and 156 female) from the areas of humanities, exacts and biologics, aged 18-28. The phase of validation involved the reliability valuation, the construct validity and the criteria validity. The reliability study, made by the Spearman rank correlation coefficient (N= 84, 38 male and 46 female), indicated that the secure and anxious ownership patterns and envious pattern, for people figure, and the secure and anxious ownership patterns, for thing figure, presented acceptable temporal stability in the retest (p< 0,01), four months later. The study of construct validity, made by factorial analysis by the method of analysis of principal components and Oblimin rotation, resulted in three interpretable factors (21,88% of variance), that evaluate the following dimensions: F1 - Ownership security People, F2 - Ownership security Things, and F3 - Ownership detachment. The good internal consistency of the scale was revealed by an Alpha of 0,70. As measure of external criteria, for the study of criteria validity, semi-structured clinical interviews were carried out in a smaller sample (N= 30, 15 male and 15 female). A valuation of agreement among three independent judges, the author included, about the categorisation and quantification of the answers of four interviews, of 69,69% and, by the Kappa coefficient, of 0,51, 0,43 and 0,73 for each pair of judges, was considered acceptable, suggesting the reliability of the criteria. The Spearman rank correlation coefficients, between the scale and interviews, indicated good validity (p< 0,01) for the secure (rs = 0,65) and anxious (rs = 0,73) ownership patterns and envious pattern (rs = 0,65), when they refer to the figure of person or animal. When they refer to the figure of things, the coefficients were a little more modest, but significant (p< 0,01) for the secure (rs = 0,51) and jealous (rs = 0,54) ownership patterns. A descriptive analysis of the distribuition of the referred patterns, in the general sample and by sex, has also been carried out. / Trata-se da constru??o e valida??o de uma escala de posse de auto-relato, baseada na experi?ncia cl?nica e numa teoria etol?gica da posse, que a postula como uma tend?ncia instintiva de o indiv?duo obter e/ou manter o dom?nio de uma ou mais figuras de posse, por meio de padr?es instintivos de comportamento de posse, regulando a provis?o para usufruto pessoal e/ou da prole. Essa primeira vers?o ficou composta por 25 itens, referentes ? posse dos dois tipos gerais de figura de posse: pessoas e coisas. Esses itens permitem ao sujeito indicar o grau de intensidade das respostas, numa escala do tipo Likert de quatro pontos (1-fraca, 2-m?dia, 3-forte, 4-demasiada). A fase de constru??o incluiu a an?lise sem?ntica e te?rica dos itens por universit?rios e juizes, respectivamente. O ?ndice de acordo entre cinco juizes independentes sobre a sua pertin?ncia te?rica foi de 90% e Kappa = 0,88, sendo, portanto, alto. A amostra do teste foi constitu?da por universit?rios (N= 296, 140 homens e 156 mulheres) das ?reas de humanas, exatas e biol?gicas, entre 18 e 28 anos de idade. A fase de valida??o envolveu a estimativa de precis?o, a validade de constructo e a validade de crit?rio. O estudo de precis?o, feito pelo coeficiente de correla??o por postos de Spearman, com uma amostra (N= 84, 38 homens e 46 mulheres), indicou que os padr?es de posse segura e ansiosa e padr?o invejoso, para a figura de pessoa, e os padr?es de posse segura e ansiosa, para a figura de coisa, apresentaram estabilidade temporal aceit?vel no reteste (p< 0,01), quatro meses depois. O estudo de validade de constructo, feito por an?lise fatorial pelo m?todo de an?lise de componentes principais e rota??o Oblimin, resultou em tr?s fatores interpret?veis (21,88% de vari?ncia), que avaliam as seguintes dimens?es: F1 - Seguran?a da posse - Pessoas, F2 - Seguran?a de posse - Coisas, e F3 - Desprendimento da posse. A boa consist?ncia interna da escala foi revelada por um Alpha de 0,70. Como medida de crit?rio externo, para o estudo de validade de crit?rio, foram realizadas entrevistas cl?nicas semi-estruturadas numa amostra menor (N= 30, 15 homens e 15 mulheres). A estimativa de acordo entre tr?s juizes independentes, inclu?do o autor, sobre a categoriza??o e quantifica??o das respostas de quatro entrevistas, de 69,69%, e, pelo coeficiente Kappa, de 0,51, 0,43 e 0,73 para cada par de juizes, foi considerada aceit?vel, sugerindo a fidedignidade do crit?rio. Os coeficientes de correla??o por postos de Spearman, entre a escala e as entrevistas, indicaram boa validade (p< 0,01) para os padr?es de posse segura (rs = 0,65) e ansiosa (rs = 0,73) e padr?o invejoso (rs = 0,65), quando se referem ? figura de pessoa ou animal. Quando se referem ? figura de coisas, os coeficientes foram um pouco mais modestos, mas significantes (p< 0,01) para os padr?es de posse segura (rs = 0,51) e ciumenta (rs = 0,54). Uma an?lise descritiva da distribui??o dos referidos padr?es, na amostra geral e por sexo, tamb?m foi feita.