Thesis (Ph. D.)--Miami University, Dept. of Psychology, 2004. / Title from second page of PDF document. Includes bibliographical references (p. 26-30).

Thesis (Ph. D.)--University of Connecticut, 2003. / Includes bibliographical references (leaves 104-112).

The study considered the prevalence of risky sexual behaviour in child headed families that usually affects the adolescents in South Africa. This study aimed at exploring the prevalence of risky sexual behaviour in child headed families, determining the causal factors of sexual promiscuous behaviour and effects of promiscuity. The area of investigation was Qora Village which is in a small town called Idutywa, Eastern Cape. This is an area that is characterised by poverty, illiteracy and unemployment. The sample was comprised of 50 participants and the study was conducted in two phases using mixed methods approach. Social exchange theory underpinned the study. Findings highlighted that these child headed adolescents had a role of caregivers and due to poverty they practiced risky sexual behaviour to bridge the economic gap. Furthermore, teenagers living in child headed families were likely to engage in risky sexual behaviour with people who are older than them; others are forced and coerced; while others perceive such act as a better option to bridge the economic gap that existed within their family backgrounds.

L’alcoolisation ponctuelle importante (API) est un phénomène mondial et à multiples facettes généralement défini comme une consommation de 4 boissons alcoolisées ou plus par occasion pour les femmes ou 5 verres ou plus par occasion pour les hommes. Les consommations excessives sous-tendent également de nombreux autres comportements à risque (violence inter et intra-personnelle, conduite en état d’ébriété, risques sexuels, etc.) et ont des effets nocifs avérés sur la santé (cancers, maladies cardiovasculaires, etc.). La triple nature de la consommation excessive d'alcool (biologique, psychologique et sociologique) englobe une grande variété de domaines allant de l'endocrinologie à la psychologie sociale et cognitive. Obtenir une vue d'ensemble du problème, cependant, est rendu difficile par le fait même que peu d'analyses évolutionnaires ont été suggérées. Ici, nous proposons d’aborder le problème sous l’angle de la théorie du signal coûteux.Quels types de signaux les buveurs excessifs envoient-ils ? À qui le signal est-il destiné ? Ces signaux sont-ils des parades nuptiales ou des menaces pour les compétiteurs afin d'assurer le succès reproductif ou maintenir un certain statut social ? Comment les facteurs contextuels peuvent-ils influencer les taux de consommation ? Explorer les causes, les corrélats et les prédicteurs de la consommation excessive d'alcool et sa nature interdisciplinaire devrait servir de point de départ pertinent pour ensuite révéler la nécessité d'un cadre évolutionnaire.Pour répondre à ces questions, j'ai commencé par élaborer une approche évolutionnaire de la consommation excessive d'alcool en tenant compte des données et des théories actuelles (Chapitre 1). J'ai ensuite mené plusieurs études de laboratoire (oculométrique notamment) et des expériences en ligne visant à évaluer les signaux inter- et intra-sexuels que les « binge drinkers » envoient aux autres (Chapitres 2 et 3). Sur la base de ces résultats, j'ai mis en place un programme de prévention sur le terrain visant à réduire l'attitude et les attentes positives de l'alcool chez les lycéens français ainsi qu'une réplication en ligne (Chapitre 4). Parallèlement, j'ai réalisé une large méta-analyse des comportements de consommation d'alcool chez les femmes et d'autres risques au cours du cycle menstruel pour comprendre les influences endocriniennes en jeu (Chapitre 5). Enfin, j’ai analysé la théorie d’histoire de vie de la consommation à risque au niveau populationnel à l'aide d'une modélisation multi-niveaux de la fréquence et de l'intensité des consommations excessives de 1997 à 2006 aux États-Unis (Chapitre 6). Dans leur ensemble, ces résultats s’inscrivent dans la perspective d’une approche évolutionnaire de la consommation excessive d'alcool et permettent d’envisager de créer des programmes de prévention micro-ciblés. / Binge drinking is a worldwide and multi-faceted phenomenon usually defined as having 4 or more alcoholic beverages per occasion for women or 5 or more drinks per occasion for men. Binge drinking also underlies many other risky behaviors (inter- and intra-personal violence, drunk driving, sexual risks, etc.) as well as negative health consequences (cancers, cardiovascular diseases, etc.). The three-fold nature of binge drinking (biological, psychological and sociological) encompasses a wide variety of areas ranging from endocrinology to social and cognitive psychology. Getting an overview of the problem, however, is made difficult by the very fact that few evolutionary analysis have been suggested. Here, we propose to frame the problem from a costly signaling perspective. What type of signals do the binge drinkers send? To whom is the signal directed? Are those signals courtship displays or threats to competitors in order to assure reproductive success or maintain social status? How can contextual factors influence drinking rates? Exploring causes, correlates and predictors of binge drinking and its interdisciplinary nature should serve as a relevant starting point to then reveal the necessity of an evolutionary framework.To address these questions, I first started by shaping an evolutionary-based approach to binge drinking taking into account current data and theories (Chapter 1). I then ran various laboratory studies (eye-tracking) and online experiments aiming at evaluating the mating and intra-sexual signals that binge drinkers send to others (Chapter 2 and 3). Based on those results, I implemented a field prevention program intended at reducing alcohol’s attitude and expectancies among French high school students as well as an online replication (Chapter 4). In parallel, I conducted an extensive meta-analysis of women’s drinking behaviors and other risks across the menstrual cycle to understand the endocrine influences at play (Chapter 5). Finally, I analyzed the life-history theory of risky drinking at a population level through a hierarchical modeling of binge drinking frequency and intensity from 1997 to 2006 across the United States (Chapter 6). Taken together, those findings can help to build the foundations for an inter-disciplinary approach to binge drinking and create specific micro-targeted prevention programs.

Bakgrund: Tidigare forskning har visat att det finns en signifikant könsskillnad gällande finansiellt risktagande, där kvinnor generellt sett har visat sig vara mindre risktagande än män. Dessutom visar teorier att en investerares risktagande bland annat varierar beroende på dennes tidigare erfarenheter samt en mängd olika karaktärsdrag. Samtidigt har tidigare studier funnit samband mellan olika variabler, så som finansiell förmåga samt övertro på sin egen förmåga, och nivå av risktagande. Idag äger kvinnor bara en tredjedel av hushållens totala aktier i Sverige men enligt Nordnets försäljningssiffror från januari 2017 har antalet kvinnliga aktieägare ökat i en snabbare takt än antalet manliga aktieägare under samma period, vilket vittnar om ett ökat kvinnligt intresse för investeringar. Syfte: Syftet med uppsatsen är att undersöka huruvida det finns något samband mellan faktorerna bakgrund, finansiell förmåga, övertro samt investeringsvana och kvinnors riskprofiler vid investeringsbeslut av finansiell karaktär. Vidare ämnar studien analysera den erhållna empirin i relation till tidigare studier inom området, för att öka förståelsen för vad som styr kvinnligt risktagande. Genomförande: Studien genomfördes via en kvantitativ forskningsmetod, där empirisk data samlades in genom en enkätundersökning som besvarades av totalt 487 kvinnor. Insamlad data analyserades via en multipel regressionsanalys i SPSS, varpå utfallet jämfördes med tidigare forskning. Slutsats: Studiens resultat visade att civilstånd, övertro, investeringsvana och finansiell förmåga har en signifikant inverkan på vilken riskprofil en kvinnlig investerare har. De tre förstnämnda förhåller sig positivt till risktagande, där en ökning i variablerna leder till ett större risktagande. Utfallet visade däremot att en ökning i finansiell förmåga leder till ett lägre risktagande, vilket gick emot tidigare forskning. Studien har således genererat både empiriskt stöd inom området och nytt bidrag kring vad som styr kvinnligt risktagande. / Background: Previous studies, within the field of behavioural finance and women ́s risk- taking, have all recognized the gender difference when evaluating risk in financial decision- making. In general, women investors tend to be more risk-averse than men, and gender differences seem to be influenced by many aspects and investor-characteristics. Earlier studies have validated the correlation between risk-taking and financial literacy and over- confidence. According to statistical data from Nordnet (2017), the number of women stock- market participants have grown in a faster pace compared to male investors, under the same period. This states that interest for investing have become a popular theme among women. Purpose: The purpose of this study is to investigate whether the factors background, financial literacy, over-confidence and investing experience can explain the risk profile of a woman financial investor. Based on the empirical results, the authors intended to analyze the output in relation to reference studies, in order to deepen the understanding and knowledge of which factors influence women ́s risk-taking in financial decision-making. Completion: The study was conducted by a quantitative method, where the empirical data was collected through a survey with a total of 487 respondents. The data was then analysed in the statistical program SPSS, using a multiple regression analysis, upon which the results were compared to previous studies. Conclusion: The results of the study disclosed that the variables of civil status, over- confidence, investing experience and financial literacy all validated a significant correlation with the risk profile of a female investor. Based on the statistical outcome, civil status, over- confidence and investing experience, demonstrated a positive correlation with the women ́s risk profile. Contrariwise, the output of financial literacy revealed a negative correlation, in which a high financial literacy determines a lower risk-taking. The empirical results can support earlier reference studies, in addition to a contribution of what influence women ́s risk profile in financial decision-making.

The study considered the prevalence of risky sexual behaviour in child headed families that usually affects the adolescents in South Africa. This study aimed at exploring the prevalence of risky sexual behaviour in child headed families, determining the causal factors of sexual promiscuous behaviour and effects of promiscuity. The area of investigation was Qora Village which is in a small town called Idutywa, Eastern Cape. This is an area that is characterised by poverty, illiteracy and unemployment. The sample was comprised of 50 participants and the study was conducted in two phases using mixed methods approach. Social exchange theory underpinned the study. Findings highlighted that these child headed adolescents had a role of caregivers and due to poverty they practiced risky sexual behaviour to bridge the economic gap. Furthermore, teenagers living in child headed families were likely to engage in risky sexual behaviour with people who are older than them; others are forced and coerced; while others perceive such act as a better option to bridge the economic gap that existed within their family backgrounds

O objetivo desta pesquisa foi investigar a relação entre comportamentos sexuais de risco e religiosidade entre jovens brasileiros de nível socioeconômico baixo. A pesquisa foi dividida em dois estudos: transversal (em São Paulo, Porto Alegre, Recife e Campo Grande) e longitudinal (em Porto Alegre). Participaram 4078 jovens com idades entre 14 e 24 anos, 46,5% masculino e 53,5% feminino, vivendo em situação de pobreza. Foi utilizado um instrumento auto-aplicável e confidencial de 109 questões. Os dados foram coletados em escolas e ONGs de forma coletiva. Foram desenvolvidas duas escalas, uma medindo religiosidade e outra comportamento de risco sexual (um indice composto por idade da primeira relação sexual, uso de camisinha, uso de métodos contraceptivos, abuso sexual e gravidez). A média de idade dos jovens participantes foi de 16,14 anos (SD=1,83) e não foi encontrada diferença entre moças e rapazes. Os resultados indicam que 97,2% são heterossexuais e 46,8% já tiveram a primeira relação sexual. A média de idade para o debute sexual foi de 14,24 anos (SD=0,60). A média do escore de comportamento de risco sexual foi de 0,60 (SD 0,60) com diferença significativa entre moças e rapazes (t=8,99; p<0,001), com média mais elevada para as moças (0,68 e 0,51). Em termos de religiosidade, a maioria se auto-declarou como católicos (40,8%), seguidos por aqueles que não têm uma religião (24,5%) e protestantes (20,5%). Houve uma diferença significativa para o nível de religiosidade (t=11,47; p<0,001) entre moças e rapazes, a média do índice de religiosidade para eles foi de -0,21 (SD=1,03) e para elas foi de 0,19 (SD=0,94). A amostra foi dividida em três grupos de religiosidade: alta, média e baixa religiosidade. O grupo com alta religiosidade teve o maior indice de comportamento sexual de risco (F=7,82; p<0,001). De seis indicadores de risco sexual, somente a relação sexual foi associada ao nível de religiosidade (participantes mais religiosos tenderam a atrasar a idade da primeira relação sexual). Dados longitudinais revelaram que a experiência sexual tende a diminuir o nível de religiosidade. Os resultados sugerem que, embora a religiosidade seja um fator de proteção atrasando a primeira relação sexual, esta não se mantem como um fator de proteção significativo após o debute sexual. / The aim of this research was to investigate the relationship between sexual risk taking behavior and religiosity among poor youth in Brazil. The research was divided into two studies: cross-sectional (in the cities of São Paulo, Porto Alegre, Recife and Campo Grande) and longitudinal (in Porto Alegre). The research was conducted with 4078 young people, ages 14-24, 46.5 % males and 53.5% females, living in poor social conditions. Data were collected using a confidential self-administered questionnaire with 109 questions. Data collection was conducted in groups at schools and non-governmental organizations (NGOs). Two scales were developed from the data, the first measuring religiosity and the second measuring sexual risk taking behavior (a composite of age of first sexual intercourse, condom use, contraception method, sexual abuse and pregnancy). The mean age was 16.14 years old (SD 1.83) and there is no age difference between males and females. Results indicate that 97.2% were heterosexual and 46,8% have had their first sexual intercourse. The mean age for the sexual debut was 14.24 years old (SD 1.93). The mean score of sexual risk taking behavior was 0.60 (SD 0.60) for the overall sample. A significant difference was found between males and females (t=8.99; p<0,001), showing that the girls have higher scores than the boys (0.68 and 0.51). Concerning religiosity, the majority was catholic (40.8%), followed by people who do not have a religion (24.5%) and protestants (20.5%). There was a significant gender difference for religiosity (t=11.47; p<0.001), the mean for males was -0.21 (SD 1.03) and for females was 0.19 (SD 0.94). The sample was divided into three groups of religiosity: high, medium or low religiosity. The high religiosity group has the highest level of sexual risk taking behavior (F=7.82; p<0.001). Of the six risk-taking indicators, only sexual debut was associated with religiosity (respondents who were more religious tended to delay sexual intercourse). Longitudinal data showed that experiencing sexual intercourse decrease religiosity levels. The findings suggest that, although religiosity plays a protective role delaying sexual debut, this is not a predictive variable for sexual risk taking behavior among poor young people in Brazil, once they are not abstinent.

Made available in DSpace on 2015-07-22T20:50:44Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-02-24 / Este estudo teve por objetivo explorar como a resiliência e alguns de seus aspectos se relacionam aos diferentes padrões de uso de drogas por adolescentes. Foi composta uma amostra representativa de estudantes de escolas particulares de São Paulo (Brasil), selecionada por estratos e conglomerados. Os dados foram coletados através de questionário de auto-preenchimento aplicado em sala de aula. Para mensurar o uso no mês de álcool, binge de álcool, tabaco e outras drogas (maconha, inalantes, ansiolíticos, estimulantes, cocaína, ecstasy e crack) foi utilizado um questionário proposto pela OMS. Para avaliar resiliência, foi aplicada a Escala de Resiliência. O principal fator da escala, encontrado na validação brasileira, foi nomeado “resolução de ações e valores”. As classes econômicas foram mensuradas pela escala da ABEP. Os dados foram analisados por estatística descritiva, bivariada (GLM), análise de cluster e regressão logística. O nível de significância adotado foi de 5%. Os 2691 estudantes que participaram da amostra apresentaram média de 16 anos (IC=15,9-16,1) e 95,5% eram de classes mais favorecidas (A e B). A droga mais consumida foi o álcool (50,2%), seguida do tabaco (14,1%) e outras drogas (11,6%). O consumo binge de álcool foi relatado por 31,8%. A análise de cluster definiu quatro grupos quanto ao uso de drogas no mês: não usou drogas, álcool, álcool/binge/tabaco, álcool/binge/tabaco/outras drogas. A resiliência não se relacionou aos padrões de uso de drogas avaliados nessa população. Apesar disso, o fator de “resolução de ações e valores”, que avaliou auto-estima, determinação, disciplina, bom humor, prontidão para ajuda e adaptabilidade, contribuiu para uma menor chance de uso de álcool, binge, uso de tabaco e de outras drogas. Para esse padrão de uso, os principais aspectos protetores foram a disciplina e a determinação. A disciplina também diminuiu as chances do uso de álcool, associado a episódios de binge e uso de tabaco. Esses resultados ressaltam a relevância de aspectos relacionados a ações e valores, especialmente disciplina e determinação, para prevenção ao uso de drogas na adolescência, cuja prática tem sido um grande desafio. / The aim was assess how resilience and its factors are related to different drug use patterns in adolescents. Private school students from São Paulo (Brazil) were selected by a representative, stratified and cluster sample. It was applied a self-report questionnaire proposed by WHO measuring alcohol, binge drinking, tobacco and other drugs (marijuana, inhalants, tranquilizers, stimulants, cocaine, ecstasy and crack) 30-day use. The resilience was assessed by Resilience Scale, which in Brazilian validation the main factor was named “capacity to act based on ones values”. The economic class was measured by ABEP scale. Weighted data were analyzed by descriptive, bivariate (GLM), cluster analysis and logistic regression statistics. Significance level was set at 5%. The sample was comprised of 2691 students with mean age 16 (CI=15.9-16.1), 52% were female and 95.5% were from upper economic classes. The most prevalent drug consumption was alcohol (50.2%), followed by tobacco (14.1%) and other drugs (11.6%). Binge drinking was related by 31.8%.Cluster analysis identified four groups according to 30-day drug use: do not use drugs, alcohol, alcohol/ binge drinking/ tobacco, alcohol/ binge drinking/ tobacco/ other drugs. The resilience was not associated with different drug use patterns in this population. However, the factor “capacity to act based on ones values” that evaluate self-steam, determination, self-discipline, good mood, readiness to help and adaptability, contribute to decrease the odds of alcohol use, binge drinking and other drug use. The main protect factors for this drug use pattern was self-discipline and determination. Self-discipline also decreased the odds of alcohol associated with binge drinking and tobacco use. These results highlight the relevance of aspects related to “capacity to act based on ones values”, especially self-discipline and determination, to drug use prevention in adolescence. / TEDE / BV UNIFESP: Teses e dissertações

O objetivo desta pesquisa foi investigar a relação entre comportamentos sexuais de risco e religiosidade entre jovens brasileiros de nível socioeconômico baixo. A pesquisa foi dividida em dois estudos: transversal (em São Paulo, Porto Alegre, Recife e Campo Grande) e longitudinal (em Porto Alegre). Participaram 4078 jovens com idades entre 14 e 24 anos, 46,5% masculino e 53,5% feminino, vivendo em situação de pobreza. Foi utilizado um instrumento auto-aplicável e confidencial de 109 questões. Os dados foram coletados em escolas e ONGs de forma coletiva. Foram desenvolvidas duas escalas, uma medindo religiosidade e outra comportamento de risco sexual (um indice composto por idade da primeira relação sexual, uso de camisinha, uso de métodos contraceptivos, abuso sexual e gravidez). A média de idade dos jovens participantes foi de 16,14 anos (SD=1,83) e não foi encontrada diferença entre moças e rapazes. Os resultados indicam que 97,2% são heterossexuais e 46,8% já tiveram a primeira relação sexual. A média de idade para o debute sexual foi de 14,24 anos (SD=0,60). A média do escore de comportamento de risco sexual foi de 0,60 (SD 0,60) com diferença significativa entre moças e rapazes (t=8,99; p<0,001), com média mais elevada para as moças (0,68 e 0,51). Em termos de religiosidade, a maioria se auto-declarou como católicos (40,8%), seguidos por aqueles que não têm uma religião (24,5%) e protestantes (20,5%). Houve uma diferença significativa para o nível de religiosidade (t=11,47; p<0,001) entre moças e rapazes, a média do índice de religiosidade para eles foi de -0,21 (SD=1,03) e para elas foi de 0,19 (SD=0,94). A amostra foi dividida em três grupos de religiosidade: alta, média e baixa religiosidade. O grupo com alta religiosidade teve o maior indice de comportamento sexual de risco (F=7,82; p<0,001). De seis indicadores de risco sexual, somente a relação sexual foi associada ao nível de religiosidade (participantes mais religiosos tenderam a atrasar a idade da primeira relação sexual). Dados longitudinais revelaram que a experiência sexual tende a diminuir o nível de religiosidade. Os resultados sugerem que, embora a religiosidade seja um fator de proteção atrasando a primeira relação sexual, esta não se mantem como um fator de proteção significativo após o debute sexual. / The aim of this research was to investigate the relationship between sexual risk taking behavior and religiosity among poor youth in Brazil. The research was divided into two studies: cross-sectional (in the cities of São Paulo, Porto Alegre, Recife and Campo Grande) and longitudinal (in Porto Alegre). The research was conducted with 4078 young people, ages 14-24, 46.5 % males and 53.5% females, living in poor social conditions. Data were collected using a confidential self-administered questionnaire with 109 questions. Data collection was conducted in groups at schools and non-governmental organizations (NGOs). Two scales were developed from the data, the first measuring religiosity and the second measuring sexual risk taking behavior (a composite of age of first sexual intercourse, condom use, contraception method, sexual abuse and pregnancy). The mean age was 16.14 years old (SD 1.83) and there is no age difference between males and females. Results indicate that 97.2% were heterosexual and 46,8% have had their first sexual intercourse. The mean age for the sexual debut was 14.24 years old (SD 1.93). The mean score of sexual risk taking behavior was 0.60 (SD 0.60) for the overall sample. A significant difference was found between males and females (t=8.99; p<0,001), showing that the girls have higher scores than the boys (0.68 and 0.51). Concerning religiosity, the majority was catholic (40.8%), followed by people who do not have a religion (24.5%) and protestants (20.5%). There was a significant gender difference for religiosity (t=11.47; p<0.001), the mean for males was -0.21 (SD 1.03) and for females was 0.19 (SD 0.94). The sample was divided into three groups of religiosity: high, medium or low religiosity. The high religiosity group has the highest level of sexual risk taking behavior (F=7.82; p<0.001). Of the six risk-taking indicators, only sexual debut was associated with religiosity (respondents who were more religious tended to delay sexual intercourse). Longitudinal data showed that experiencing sexual intercourse decrease religiosity levels. The findings suggest that, although religiosity plays a protective role delaying sexual debut, this is not a predictive variable for sexual risk taking behavior among poor young people in Brazil, once they are not abstinent.

O objetivo desta pesquisa foi investigar a relação entre comportamentos sexuais de risco e religiosidade entre jovens brasileiros de nível socioeconômico baixo. A pesquisa foi dividida em dois estudos: transversal (em São Paulo, Porto Alegre, Recife e Campo Grande) e longitudinal (em Porto Alegre). Participaram 4078 jovens com idades entre 14 e 24 anos, 46,5% masculino e 53,5% feminino, vivendo em situação de pobreza. Foi utilizado um instrumento auto-aplicável e confidencial de 109 questões. Os dados foram coletados em escolas e ONGs de forma coletiva. Foram desenvolvidas duas escalas, uma medindo religiosidade e outra comportamento de risco sexual (um indice composto por idade da primeira relação sexual, uso de camisinha, uso de métodos contraceptivos, abuso sexual e gravidez). A média de idade dos jovens participantes foi de 16,14 anos (SD=1,83) e não foi encontrada diferença entre moças e rapazes. Os resultados indicam que 97,2% são heterossexuais e 46,8% já tiveram a primeira relação sexual. A média de idade para o debute sexual foi de 14,24 anos (SD=0,60). A média do escore de comportamento de risco sexual foi de 0,60 (SD 0,60) com diferença significativa entre moças e rapazes (t=8,99; p<0,001), com média mais elevada para as moças (0,68 e 0,51). Em termos de religiosidade, a maioria se auto-declarou como católicos (40,8%), seguidos por aqueles que não têm uma religião (24,5%) e protestantes (20,5%). Houve uma diferença significativa para o nível de religiosidade (t=11,47; p<0,001) entre moças e rapazes, a média do índice de religiosidade para eles foi de -0,21 (SD=1,03) e para elas foi de 0,19 (SD=0,94). A amostra foi dividida em três grupos de religiosidade: alta, média e baixa religiosidade. O grupo com alta religiosidade teve o maior indice de comportamento sexual de risco (F=7,82; p<0,001). De seis indicadores de risco sexual, somente a relação sexual foi associada ao nível de religiosidade (participantes mais religiosos tenderam a atrasar a idade da primeira relação sexual). Dados longitudinais revelaram que a experiência sexual tende a diminuir o nível de religiosidade. Os resultados sugerem que, embora a religiosidade seja um fator de proteção atrasando a primeira relação sexual, esta não se mantem como um fator de proteção significativo após o debute sexual. / The aim of this research was to investigate the relationship between sexual risk taking behavior and religiosity among poor youth in Brazil. The research was divided into two studies: cross-sectional (in the cities of São Paulo, Porto Alegre, Recife and Campo Grande) and longitudinal (in Porto Alegre). The research was conducted with 4078 young people, ages 14-24, 46.5 % males and 53.5% females, living in poor social conditions. Data were collected using a confidential self-administered questionnaire with 109 questions. Data collection was conducted in groups at schools and non-governmental organizations (NGOs). Two scales were developed from the data, the first measuring religiosity and the second measuring sexual risk taking behavior (a composite of age of first sexual intercourse, condom use, contraception method, sexual abuse and pregnancy). The mean age was 16.14 years old (SD 1.83) and there is no age difference between males and females. Results indicate that 97.2% were heterosexual and 46,8% have had their first sexual intercourse. The mean age for the sexual debut was 14.24 years old (SD 1.93). The mean score of sexual risk taking behavior was 0.60 (SD 0.60) for the overall sample. A significant difference was found between males and females (t=8.99; p<0,001), showing that the girls have higher scores than the boys (0.68 and 0.51). Concerning religiosity, the majority was catholic (40.8%), followed by people who do not have a religion (24.5%) and protestants (20.5%). There was a significant gender difference for religiosity (t=11.47; p<0.001), the mean for males was -0.21 (SD 1.03) and for females was 0.19 (SD 0.94). The sample was divided into three groups of religiosity: high, medium or low religiosity. The high religiosity group has the highest level of sexual risk taking behavior (F=7.82; p<0.001). Of the six risk-taking indicators, only sexual debut was associated with religiosity (respondents who were more religious tended to delay sexual intercourse). Longitudinal data showed that experiencing sexual intercourse decrease religiosity levels. The findings suggest that, although religiosity plays a protective role delaying sexual debut, this is not a predictive variable for sexual risk taking behavior among poor young people in Brazil, once they are not abstinent.