This research empirically investigates Locke's (1997) model of work motivation by means of quantitative research. The OCQ consisting of three tiered questionnaires was constructed based on Locke's model. OCQ-Tierl deals with core components of Locke's model. OCQ-Tier2 determines which factors caused the incidence of dissatisfaction in OCQ-Tierl. OCQ-Tier3 enables the identification of corrective actions. The OCQ was administered to financial services employees. The results were analysed and Locke's model was tested by means of structural equation modelling using the AMOS graphics programme. The results indicated that the model, suggesting causal links between components within OCQ-Tierl, could not be confirmed. A better fit was found at OCQ-Tier2 and OCQ-Tier3. In testing the causal links across the three tiers per component, the models did not fit the data for "personal actualisation" and "goal achievement". Moderate confirmation of the models was found in the case of "goal setting" and "goal behaviour" across the three tiers after some adaptations were made to the models on the basis of "modification indices", suggested by AMOS. A reasonably good fit was found for the models across the three tiers for "quality of work life". The level of correlation between factors was high because of this, and in some cases some of the factors were merged. Modification indices in the statistical output suggested that improvement was possible if covariance between error terms in the model was allowed. This suggested possible systematic sources of covariance between items not accounted for by the factors in the models. As confirmed by the Cronbach Alpha coefficients within tiers and across tiers, the general level of internal consistency was very high. Possibly response set and response style were the cause of this. This made the testing of models difficult in the present study. So too was it difficult to draw a conclusion about the internal consistency reliability of the measurement of each component across the three tiers, because the high Cronbach coefficients may to some extent be due to the indiscriminate high correlations between items / D.Litt. et Phil. (Industrial Psychology)

Cognitive behaviour therapy and interpersonal psychotherapy are the leading treatments for eating disorders. Little is understood regarding their mechanisms of action. The research described in this thesis investigated the purported mechanisms of action of two transdiagnostic versions of these treatments; enhanced cognitive behaviour therapy (CBT-E) and interpersonal psychotherapy for eating disorders (IPT-ED). A series of mediation studies were embedded within a randomised controlled trial comparing CBT-E and IPT-ED in a transdiagnostic sample of patients with eating disorders. An analytic strategy using multilevel and structural equation modelling was used to assess for statistical mediation. Three of the key purported mediators of action of CBT-E (regularity of eating, frequency of weighing, and frequency of shape checking) and the key purported mediator of IPT-ED (interpersonal problem severity) were studied. Only regularity of eating demonstrated a strong case for being a mediator of the effect of CBT-E (on frequency of binge eating). The findings were inconclusive regarding other putative mediators of the effect of CBT-E, and for interpersonal problem severity being a mediator of the effect of IPT-ED. Limitations of the research included the non-optimal choice of measures and non-optimal timing of measurements for establishing temporal precedence. Future research should investigate the mediating role of both cognitive (e.g. interpretation of weight) and behavioural processes (e.g. frequency of weighing) in CBT-E, and the role of interpersonal functioning in CBT-E and IPT-ED. Research should use daily, or session by session measurement to better assess the temporal relationship between the purported mediator/s and the outcome/s. Experimental designs comparing dismantled versions of treatment would also help determine the relative effects that different treatment procedures have on treatment outcome/s.

Formações ribeirinhas são áreas de transição entre os sistemas terrestre e aquático, caracterizadas por elevada heterogeneidade ambiental e por um amplo espectro de elementos biológicos e valores socioeconômicos. Contrário às grandes bacias hidrográficas, onde geralmente os pulsos de inundação são sazonais, os rios de pequeno porte apresentam regimes de inundação freqüentemente imprevisíveis, onde os gradientes de perturbação podem ser expressos pela razão direta entre intensidade, duração e freqüência. Considerando a distribuição de grupos ecológicos ao longo de um gradiente de inundação e as relações entre distúrbio e variáveis ambientais, o presente trabalho teve como objetivo avaliar os padrões de riqueza e diversidade de espécies arbóreas em um rio com regime de inundações não-previsível. O estudo foi realizado próximo à foz do rio Botucaraí (30º 01’S e 52º 47’W), em quatro parcelas de 1 ha, situadas em diferentes cotas de elevação. Cada parcela foi dividida em 10 transecções contíguas de 10 × 100 m, paralelas à margen do rio, sendo cada transecção posteriormente subdividida em unidades amostrais de 10 × 10 m. Em cada parcela, todos os indivíduos arbóreos vivos com perímetro à altura do peito (pbh) ≥ 15 cm foram amostrados. Coletas superficiais de solo (0 – 20 cm de profundidade) foram realizadas, bem como um detalhamento topográfico das parcelas. A freqüência de distúrbio em cada unidade amostral foi definida a partir da interpolação de cota topográfica de cada unidade amostral com os pulsos de inundação para o período de 1981–2004. Técnicas multivariadas de ordenação e classificação foram aplicadas, a fim de explorar as relações entre distribuição de espécies arbóreas, grupos ecológicos de regeneração, estratificação e dispersão com as variáveis ambientais. Uma modelagem por equações estruturadas definiu um modelo significativo entre freqüência de distúrbio, biomassa e variáveis ambientais sobre a variação na riqueza de espécies. Foram amostrados 5.779 indivíduos arbóreos de 37 famílias botânicas. A interação entre distúrbio e biomassa revelou um pico de riqueza de espécies com o aumento da biomassa e a diminuição da freqüência de distúrbios. O modelo estrutural final explicou 79% da variação na riqueza de espécies e 67% na variação sobre biomassa. A análise dos grupos ecológicos revelou menor estratificação vertical em direção às áreas mais inundáveis. Estas áreas também apresentaram maior proporção de indivíduos de espécies pioneiras e dispersão autocórica. Variações florísticas ao longo do gradiente sugerem que a seleção de habitat influencia os padrões de distribuição das espécies, favorecendo histórias de vida com características generalistas, em decorrência de um regime de inundações não-previsível. / Riverine systems are transitional areas between land and aquatic ecosystems that present high environmental heterogeneity and including a wide array of socio-economic and biological values. In large hydrographic basins, flood pulses are seasonal and currently predicted which differ substantially from small rivers, where flood pulses are unpredictable and should be referred as a direct interaction of intensity, duration and frequency. By analyzing the distribution of ecological groups across a flood gradient and the relationships among disturbance and environmental variables, the present work addressed to analyze tree species richness and diversity in a small river with unpredictable flood regime. The survey was conducted in the lower sector of rio Botucaraí (30º 01’S e 52º 47’W) where we installed four 1 ha plots across different topographical sites. Each plot comprised ten contiguous 10 × 100 m transects, parallels to the river margin and then, each transect was subdivided in 10 × 10 m sampling units. In each plot we sampled all individual living trees having at least 15 cm of perimeter at the breast height (pbh). We collected samples of the topsoil (0–20 cm depth) and we carried out a detailed topographic survey of each plot. We interpolated the topographical position of each sampling unit and inundation records from 1981–2004 to obtain a disturbance frequency to each position. We applied multivariate ordination and grouping techniques to seek for relationships among environmental variables, tree species distribution and ecological groups of regeneration, stratification and dispersal. A structural equation modelling improved a significant model among flood frequency, biomass and environmental variables over species richness. Field inventory yielded a total of 5,779 trees belonging to 96 species and 36 families. The interaction between disturbance and biomass revealed a species richness peak with higher biomass and lower disturbance frequency. The final structural model explained 79% of variance in species richness and 67% of variance in biomass. The ecological groups’ analyses revealed lower vertical stratification toward frequently inundated areas. Moreover, these sites presented higher proportion of pioneer and autochorous trees. Floristic changes across the gradient suggest that habitat selection influences species distribution patterns, by selecting generalized live history traits due to the unpredictable flooding regime.

Formações ribeirinhas são áreas de transição entre os sistemas terrestre e aquático, caracterizadas por elevada heterogeneidade ambiental e por um amplo espectro de elementos biológicos e valores socioeconômicos. Contrário às grandes bacias hidrográficas, onde geralmente os pulsos de inundação são sazonais, os rios de pequeno porte apresentam regimes de inundação freqüentemente imprevisíveis, onde os gradientes de perturbação podem ser expressos pela razão direta entre intensidade, duração e freqüência. Considerando a distribuição de grupos ecológicos ao longo de um gradiente de inundação e as relações entre distúrbio e variáveis ambientais, o presente trabalho teve como objetivo avaliar os padrões de riqueza e diversidade de espécies arbóreas em um rio com regime de inundações não-previsível. O estudo foi realizado próximo à foz do rio Botucaraí (30º 01’S e 52º 47’W), em quatro parcelas de 1 ha, situadas em diferentes cotas de elevação. Cada parcela foi dividida em 10 transecções contíguas de 10 × 100 m, paralelas à margen do rio, sendo cada transecção posteriormente subdividida em unidades amostrais de 10 × 10 m. Em cada parcela, todos os indivíduos arbóreos vivos com perímetro à altura do peito (pbh) ≥ 15 cm foram amostrados. Coletas superficiais de solo (0 – 20 cm de profundidade) foram realizadas, bem como um detalhamento topográfico das parcelas. A freqüência de distúrbio em cada unidade amostral foi definida a partir da interpolação de cota topográfica de cada unidade amostral com os pulsos de inundação para o período de 1981–2004. Técnicas multivariadas de ordenação e classificação foram aplicadas, a fim de explorar as relações entre distribuição de espécies arbóreas, grupos ecológicos de regeneração, estratificação e dispersão com as variáveis ambientais. Uma modelagem por equações estruturadas definiu um modelo significativo entre freqüência de distúrbio, biomassa e variáveis ambientais sobre a variação na riqueza de espécies. Foram amostrados 5.779 indivíduos arbóreos de 37 famílias botânicas. A interação entre distúrbio e biomassa revelou um pico de riqueza de espécies com o aumento da biomassa e a diminuição da freqüência de distúrbios. O modelo estrutural final explicou 79% da variação na riqueza de espécies e 67% na variação sobre biomassa. A análise dos grupos ecológicos revelou menor estratificação vertical em direção às áreas mais inundáveis. Estas áreas também apresentaram maior proporção de indivíduos de espécies pioneiras e dispersão autocórica. Variações florísticas ao longo do gradiente sugerem que a seleção de habitat influencia os padrões de distribuição das espécies, favorecendo histórias de vida com características generalistas, em decorrência de um regime de inundações não-previsível. / Riverine systems are transitional areas between land and aquatic ecosystems that present high environmental heterogeneity and including a wide array of socio-economic and biological values. In large hydrographic basins, flood pulses are seasonal and currently predicted which differ substantially from small rivers, where flood pulses are unpredictable and should be referred as a direct interaction of intensity, duration and frequency. By analyzing the distribution of ecological groups across a flood gradient and the relationships among disturbance and environmental variables, the present work addressed to analyze tree species richness and diversity in a small river with unpredictable flood regime. The survey was conducted in the lower sector of rio Botucaraí (30º 01’S e 52º 47’W) where we installed four 1 ha plots across different topographical sites. Each plot comprised ten contiguous 10 × 100 m transects, parallels to the river margin and then, each transect was subdivided in 10 × 10 m sampling units. In each plot we sampled all individual living trees having at least 15 cm of perimeter at the breast height (pbh). We collected samples of the topsoil (0–20 cm depth) and we carried out a detailed topographic survey of each plot. We interpolated the topographical position of each sampling unit and inundation records from 1981–2004 to obtain a disturbance frequency to each position. We applied multivariate ordination and grouping techniques to seek for relationships among environmental variables, tree species distribution and ecological groups of regeneration, stratification and dispersal. A structural equation modelling improved a significant model among flood frequency, biomass and environmental variables over species richness. Field inventory yielded a total of 5,779 trees belonging to 96 species and 36 families. The interaction between disturbance and biomass revealed a species richness peak with higher biomass and lower disturbance frequency. The final structural model explained 79% of variance in species richness and 67% of variance in biomass. The ecological groups’ analyses revealed lower vertical stratification toward frequently inundated areas. Moreover, these sites presented higher proportion of pioneer and autochorous trees. Floristic changes across the gradient suggest that habitat selection influences species distribution patterns, by selecting generalized live history traits due to the unpredictable flooding regime.

The internet has changed the way that consumers interact with each other and with companies. Social media appear as a new communication channel for business and enabling closer interaction between enterprises and consumers. The literature explains that the social media has a vital role in the promotion mix and companies can interact with consumers through online platforms (Facebook, Twitter, MySpace and others). Therefore, companies can no ignore the phenomena of social media, because it have become a tool for consumers expose their perception of products and services. This study aimed to evaluate the relationship between the emotional response to advertising on social media (Facebook) and the evaluation of the Coca-Cola brand, under the influence of technology readiness and consumers trust. The methodological procedures used to obtain answers to the objectives that guided this study followed the steps of the survey method, by surveying a cross-sectional, exploratory and descriptive study. To assess the relationship between different variables, we used the technique of structural equation modeling. The population was composed of all Brazilians, social media users, subscribers on Facebook. We obtained a total of 1,196 respondents, and of these, 1,037 are valid questionnaires. The choice of brand Coca-Cola gave up because is considered the most valuable in the world for 13 consecutive years and the most followed brand on Facebook worldwide and the second most followed in Brazil. The confirmation of the properties of unidimensionality, reliability, convergent and discriminant validity of the constructs permitted to propose and validate an integrated model, considered as hybrid. The hybrid model tested 48 variables and 118 estimated parameters. The results show that: (1) the emotional reaction to advertising on social media has a positive and direct influence on the Coca-Cola brand evaluation; (2) consumer trust exercised direct and positive influence on brand evaluation and emotional reaction to advertising on social media; and (3) the technology readiness has not a significant influence on the emotional reaction to advertising on social media. We concluded that it is possible to determine consumer trust as the most important variable to a positive emotional reaction to advertising on social media and a positive brand evaluation. / A internet tem alterado a maneira como os consumidores interagem entre si e com as empresas. As mídias sociais aparecem como um novo canal de comunicação para os negócios permitindo maior proximidade e interação das empresas com os consumidores. A literatura explica que as mídias sociais tem um papel vital dentro do mix de promoção, sendo que as empresas podem interagir com os consumidores por meio de plataformas online (Facebook, Twitter, MySpace e outros). Assim as empresas não podem mais ignorar o fenômeno das mídias sociais, porque elas se tornaram a ferramenta para os consumidores exporem sua percepção sobre produtos e serviços. Este estudo objetivou avaliar a relação entre a reação emocional à propaganda em mídias sociais (Facebook) e a avaliação da marca Coca-Cola, sob a influência da prontidão à tecnologia e da confiança do consumidores. Os procedimentos metodológicos utilizados para obter respostas aos objetivos que orientaram este estudo, seguiram as etapas do método survey, por meio de levantamento de corte transversal, por um estudo de natureza exploratória e descritiva. Para avaliação da relação entre variáveis distintas, utilizou-se a técnica de modelagem de equações estruturais. O universo do estudo foi composto por todos brasileiros, usuários de mídias sociais, inscritos no Facebook. Obteve-se um total de 1.196 respondentes, sendo destes, 1.037 questionários válidos. A escolha da marca Coca-Cola deu-se por esta ter sido a mais valiosa do mundo por 13 anos consecutivos e a marca mais seguida no Facebook em nível mundial e a segunda mais seguida no Brasil. A confirmação das propriedades de unidimensionalidade, confiabilidade, validade convergente e discriminante dos construtos permitiu propor e validar um modelo integrado, considerado híbrido. O modelo híbrido testado, apresentou 48 variáveis e 118 parâmetros estimados. Os resultados evidenciam que: (1) a reação emocional à propaganda em mídias sociais exerceu influência positiva e direta sobre avaliação da marca Coca-Cola; (2) a confiança do consumidor exerceu influência positiva e direta sobre a avaliação da marca e a reação emocional à propaganda em mídias sociais; (3) a prontidão à tecnologia exerceu influência positiva sobre a reação emocional à propaganda em mídias sociais em nível não significante. Conclui-se que é possível determinar a confiança do consumidor como a variável chave para uma reação emocional à propaganda em mídias sociais e avaliação da marca positiva.

Women with histories of interpersonal trauma (physical, sexual, or psychological abuse experienced during childhood, adolescence, and adulthood) are more likely to experience posttraumatic stress symptoms (PTSS) and to develop physical health problems than women without trauma histories. In fact, PTSS and posttraumatic stress disorder (PTSD) have been established in the literature as mediators of the relation between interpersonal trauma and physical health outcomes (e.g., Resnick et al., 1997; Schnurr & Green, 2004). What remains to be determined is a clear understanding of the various mechanisms explaining why individuals with trauma histories, and subsequently PTS symptoms, go on to develop physical health problems. The purpose of this study was to examine the role of health risk behaviours, specifically sexual risk taking and substance use, as possible mechanisms through which interpersonal trauma and PTSS might influence physical health. These relations were examined, through structural equation modelling, in a sample of 475 women currently attending university. Models were tested separately for sexual traumas (childhood sexual abuse and sexual assault experienced during adolescence and adulthood) and nonsexual interpersonal traumas (physical and psychological maltreatment by parents in childhood, witnessing violence between parents, and intimate partner violence in their own relationships). Results indicated that PTSS severity partially mediated pathways from both types of interpersonal trauma, sexual and nonsexual, to adverse health outcomes, contributing to the existing theory that one’s psychological response to a trauma may be more important in determining physical health outcomes than the trauma itself. Furthermore, a significant indirect pathway was found to link nonsexual trauma to risky sexual behaviours through PTSS severity. In addition, PTSS severity fully mediated the relation between nonsexual trauma and substance use behaviours. These latter findings suggest that the likelihood of engaging in substance use and/or risky sexual behaviours may be greater in trauma survivors who are suffering from posttraumatic stress symptoms. Contrary to hypotheses, no significant pathways were found from risky sexual behaviours or substance use to physical health outcomes in the context of trauma variables and PTSS severity. Consequently, these health risk behaviours were not found to operate as mechanisms explaining the link from PTSS severity to physical health outcomes. Limitations and alternative hypotheses are presented. Implications for clinical interventions and recommendations for future research are discussed. / Graduate / 0621 / 0622 / 0384

This research is the outcome of a preeminent interest in the topic of price perception. Pointedly, the perception of prices is part of the purchasing process, the same willingness to pay and the actual purchase behaviour, and is indubitably a perceptual construct. As such, perception is problematic to measure as it does not relate to an observable behaviour. On the other hand, pricing is regarded as an important variable in the marketing mix. This research contributes to theory by augmenting the current knowledge on the perception of prices including the methods used in the measurement of such perception. Moreover, this research addresses a gap in the understanding of how diverse ethnic groups perceive prices. The relationship set in this study between ethnicity and price perception is thought-provoking as it contributes to the current discussion around diversity in the marketplace. For example, the literature shows advances in areas such as multicultural and ethnic marketing and this research makes a significant contribution to these areas from price perception. Accordingly, this study involved a systematic review of the literature and presented a framework that suggested that the formation of price perception is affected by external factors such as culture and ethnicity. Furthermore, a qualitative study examined the formation of price perception around ethnic groups. Next, this research used a quantitative study that sought differences in price perception among ethnic groups. Thus, the quantitative study used a price perception scale (Lichtenstein et al., 1993) and a choice-based conjoint analysis. Also, the study adopted structural equation modelling (SEM) to measure differences among scales and the multinomial logit model to analyse the choice-based conjoint analysis. The findings of both the quantitative and the qualitative studies link to the systematic review and support the framework for the formation and measurement of price perception originally proposed.

The rapid growth of universities in the Least Developing Countries (LDCs) is aimed at enhancing access to tertiary education, which has resulted in a sharp increase in the enrolment rate. However, the quantitative increase has been marred with a correspondingly continuous decline in the quality of education. This is attributed to a wide range of limiting factors mainly classified as institutional problems. Some of these problems include a shortage of resources, limited skills and incompetent human capital, lack of ICT infrastructure, and the ineffective use of existing ICT resources. These problems and others have adversely affected how national education and research network can improve research output and quality of education. In this study, a survey, in the form of an exploratory quantitative research design is used. A descriptive non-experimental quantitative approach was also chosen, and a questionnaire was administered to approximately one hundred and seventy-two (172) participants drawn from twenty-nine (29) Ethiopian Public Universities. The results of the analysis show that the study variables namely NREN service for education, EthERNet, electronic device and research output have a significant and positive impact on the Quality of Education (QE) to differing degrees. Also, the study variables such as NREN service for research, high-performance computing, and remote computing facilities indicated that they had a significant and positive impact on Research Output (RO) to differing degrees. The study explored the impact of EthERNet in improving the quality of education and research output by examining the existing network infrastructure and NREN services. The study employed the use of the Actor-Network Theory (ANT) to assess the existing network infrastructure and NREN services to determine that a reliable network can improve the quality of education and research output. Besides, Structural Equation Model (SEM) was used to identify the positive and negative factors that impact on the roles, relationships, and formation of quality of education and research output. Furthermore, a three-step design science approach was applied to propose and justify the theoretical framework, which is used as a base to develop a service portfolio and roadmap conceived to design the required NREN service for EthERNet. This research contributed to the body of knowledge by finding the missing link between the quality of education and research outputs. From a theoretical perspective, the research contributed a theoretical framework by developing the construct and their measures that can be used in assessing the adoption and usage of technology. Furthermore, the study contributes to the literature by demonstrating an analytical process which could be used as a guide for future NREN service requirement to improve the quality of education and research output with the existing findings being used as a reference point. / School of Computing / Ph. D. (Computer Science)

Measurement problems in terms of different understanding of the meaning, concepts and implications of survey questions as well as tendencies to different response-sets (e.g. acquiescence or social desirability) are reoccurring problems in comparative migration research. Cultural factors, time-varying effects and survey design lead to response biases. It could be shown that depending on the region of origin, migrants tend to extreme or centred responding behaviour. Macroscale changes in attitudes towards disclosure of personal information influence acquiescent biases. Therefore, the question arises whether the presence of measurement equivalence can be assumed, especially within instruments regarding subjective feelings, when comparing migrant and native populations. The aim of the study is to test for intercultural and longitudinal equivalence of the Health Short Form 12 - items scale (SF -12), trying to answer the question whether this instrument is useful to compare health of migrants and native Germans over time. The study contributes to the ongoing discussion on measurement equivalence of the SF -12 by including variables concerning integration and assimilation, employment, family status and varying types of migration and years since migration to the analysis. Using a Multiple Group Structural Equation Modelling approach analyzing longitudinal data of the German Socioeconomic Panel (SOEP), we test the validity and measurement equivalence (configural, metric and scalar invariance) of the latent construct ‘health’ over time and between (non-)migration groups. In addition, different item functioning can be detected for a series of covariates. The study concludes in recommending a reduced version of the SF-12 to improve the avoidance of measurement artefacts when comparing the health of migrants and native Germans.

To obtain valid and reliable research findings it is important to follow the process to validate measuring instruments. This entails determining the psychometric properties of a measure to eliminate or decrease the presence of measurement errors. Measurement errors have a negative impact on the validity of research findings. The aim of this study was to perform a confirmatory factor analytic study on the Aspects of Identity Questionnaire (AIQ-IV) to assess the model fit of the data. Confirmatory factor analysis was employed as part of the process to assess construct validity of a questionnaire to discover the misrepresenting influences of these measurement errors and to provide a foundation for further research. The AIQ-IV was administered to a sample of 157 participants in the South African context including different race, gender, age and occupation groups, drawn by means of convenient sampling. The research results and fit indices indicated that the data reflected a reasonable model fit. Copyright 2010, University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. Please cite as follows: Els, L 2010, A confirmatory factor analytic study of the aspects of identity questionnaire (AIQ-IV), MCom dissertation, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-02092012-122330 / > C12/4/67/gm / Dissertation (MCom)--University of Pretoria, 2010. / Human Resource Management / unrestricted