Made available in DSpace on 2016-06-02T20:06:01Z (GMT). No. of bitstreams: 1 2197.pdf: 1979337 bytes, checksum: 0091ce4138b8a98277d9af0d5d2aa788 (MD5) Previous issue date: 2008-11-13 / Universidade Federal de Sao Carlos / In this dissertation we present a regression modelling proposal for modelling the market prices of urban batches at São Carlos city (SP), over the year of 2005. Usual regression modelling and survival techniques, with left censoring, are considered. A simulation study exames the coverage probabilities the asymptotic confidence for the parameters of the considered modelling. / Nesta disertação apresentaremos uma proposta de um modelo de equação de regressão representativa para a formação do valor de mercado dos lotes urbanos do município de São Carlos, SP, ano de 2005, visando à criação de Plantas de Valores Genéricos (PVG) utilizando as técnicas de: Modelos Lineares Usuais (erros normais e variância constante), estes amplamente utilizados, e a Análise de Sobrevivência com censura à esquerda. Após o ajuste, as duas metodologias são comparadas e testadas num estudo de simulação onde examinamos a probabilidade de cobertura de alguns parâmetros envolvidos na regressão.

<p>This dissertation investigated a psychoeducational intervention’s effectiveness in nurturing concern for others in adolescents with high intellectual abilities. The intervention was implemented at two research sites in a city in western India with 130 participants. Concern for others was conceptualized as an interplay of empathy, compassion, and prosocial behavior, and interrelationships among them were examined using correlational and regression analyses of self-reported survey data. Results indicated that prosocial behavior is positively associated with empathy (i.e., perspective taking and empathic concern) and other-compassion (i.e., compassion for others and compassion for other living beings) with correlations ranging from medium to strong (.46 ≤ <em>r</em> ≤ .79). Compassion for self, however, is not associated with prosocial behavior (<em>r</em> = .01) or any other key variables of having a concern for others (-.06 ≤ <em>r</em> ≤ .09). Compassion for others and perspective taking are the strongest predictors of prosocial behavior. Predicted self-reported prosocial behavior in girls is, on average, significantly greater than that in boys. Participants were then randomly assigned to treatment and control (delayed treatment) groups. The intervention’s effectiveness was evaluated using a convergent mixed-methods design by combining repeated-measures multivariate analysis of variance (RM-MANOVA) of self-reported survey data and thematic analysis of interview data. Classroom quality was perceived to be high. Moreover, on average, participants’ level of adherence to the intervention was high, especially for session attendance (93%), and self-reported home activity completion (89%). The RM-MANOVA results showed that the self-reported concern for others varied significantly over the time of participants’ participation in the intervention with a large effect (treatment group: ηp2 = .57; delayed treatment group: ηp2 = .47); however, the effects did not seem to sustain over the next three months. Univariate <em>post-hoc</em> analyses indicated significant differences with moderate effect size in prosocial behavior and compassion for others. Overall, the intervention was perceived to be a largely positive experience—appealing, meaningful, and supportive of belongingness—yet there is a scope for improvement, especially regarding active participation. Emotional regulation, mindful engagement, responsible communication, relationship building, kindness, and gratitude were described as key intervention affordances. The mixed-methods integration of results provided preliminary evidence for the effectiveness of the intervention. (IRB #1812021447)</p>