abstract: This dissertation examines a planned missing data design in the context of mediational analysis. The study considered a scenario in which the high cost of an expensive mediator limited sample size, but in which less expensive mediators could be gathered on a larger sample size. Simulated multivariate normal data were generated from a latent variable mediation model with three observed indicator variables, M1, M2, and M3. Planned missingness was implemented on M1 under the missing completely at random mechanism. Five analysis methods were employed: latent variable mediation model with all three mediators as indicators of a latent construct (Method 1), auxiliary variable model with M1 as the mediator and M2 and M3 as auxiliary variables (Method 2), auxiliary variable model with M1 as the mediator and M2 as a single auxiliary variable (Method 3), maximum likelihood estimation including all available data but incorporating only mediator M1 (Method 4), and listwise deletion (Method 5). The main outcome of interest was empirical power to detect the mediated effect. The main effects of mediation effect size, sample size, and missing data rate performed as expected with power increasing for increasing mediation effect sizes, increasing sample sizes, and decreasing missing data rates. Consistent with expectations, power was the greatest for analysis methods that included all three mediators, and power decreased with analysis methods that included less information. Across all design cells relative to the complete data condition, Method 1 with 20% missingness on M1 produced only 2.06% loss in power for the mediated effect; with 50% missingness, 6.02% loss; and 80% missingess, only 11.86% loss. Method 2 exhibited 20.72% power loss at 80% missingness, even though the total amount of data utilized was the same as Method 1. Methods 3 – 5 exhibited greater power loss. Compared to an average power loss of 11.55% across all levels of missingness for Method 1, average power losses for Methods 3, 4, and 5 were 23.87%, 29.35%, and 32.40%, respectively. In conclusion, planned missingness in a multiple mediator design may permit higher quality characterization of the mediator construct at feasible cost. / Dissertation/Thesis / Doctoral Dissertation Psychology 2015

abstract: The goal of diagnostic assessment is to discriminate between groups. In many cases, a binary decision is made conditional on a cut score from a continuous scale. Psychometric methods can improve assessment by modeling a latent variable using item response theory (IRT), and IRT scores can subsequently be used to determine a cut score using receiver operating characteristic (ROC) curves. Psychometric methods provide reliable and interpretable scores, but the prediction of the diagnosis is not the primary product of the measurement process. In contrast, machine learning methods, such as regularization or binary recursive partitioning, can build a model from the assessment items to predict the probability of diagnosis. Machine learning predicts the diagnosis directly, but does not provide an inferential framework to explain why item responses are related to the diagnosis. It remains unclear whether psychometric and machine learning methods have comparable accuracy or if one method is preferable in some situations. In this study, Monte Carlo simulation methods were used to compare psychometric and machine learning methods on diagnostic classification accuracy. Results suggest that classification accuracy of psychometric models depends on the diagnostic-test correlation and prevalence of diagnosis. Also, machine learning methods that reduce prediction error have inflated specificity and very low sensitivity compared to the data-generating model, especially when prevalence is low. Finally, machine learning methods that use ROC curves to determine probability thresholds have comparable classification accuracy to the psychometric models as sample size, number of items, and number of item categories increase. Therefore, results suggest that machine learning models could provide a viable alternative for classification in diagnostic assessments. Strengths and limitations for each of the methods are discussed, and future directions are considered. / Dissertation/Thesis / Doctoral Dissertation Psychology 2018

Este trabalho objetivou compreender os fatores envolvidos na infidelidade em homens e mulheres casados ou coabitando a partir de quatro domínios importantes para sua compreensão: características pessoais, características do companheiro(a), casamento e contexto. Para responder aos objetivos, o tema foi desenvolvido ao longo de cinco estudos. O primeiro artigo tratou de uma revisão sistemática da literatura sobre a infidelidade nos últimos cinco anos no cenário nacional e internacional e buscou identificar a maneira como a temática vem sendo pesquisada. Constatou-se deficiências nas pesquisas, tais como a não explicitação da definição da infidelidade, grande diversidade de instrumentos e baixo rigor metodológico. Desta forma, o segundo estudo teve por objetivo propor uma medida de infidelidade a partir do que homens e mulheres consideram comportamentos de infidelidade emocional, sexual, virtual emocional e virtual sexual. Este artigo demonstrou a não existência de diferenças significativas entre os comportamentos de infidelidade de homens e mulheres e sugeriu tratar a infidelidade como um fenômeno único, pois as tipologias acabam por gerar uma visão reducionista do constructo O estudo subsequente tratou de investigar a vivência da infidelidade em homens e mulheres, descrevendo frequência, tipos de comportamentos e motivos de busca da infidelidade. Ambos os sexos tiveram comportamentos de infidelidade semelhantes e o principal motivador foi a insatisfação com a relação e com o companheiro(a). Em seguida, os dois últimos estudos visaram investigar a diferença entre homens e mulheres que foram infiéis daqueles que se mantiveram fiéis ao parceiro(a) e delinear os preditores da infidelidade. Os achados apontaram para o domínio casamento tendo uma grande importância na diferenciação dos grupos, além de ser um preditor da infidelidade juntamente com as variáveis contextuais. Desta forma, dada a importância do domínio casamento para a compreensão da temática, conclui-se que a infidelidade se trata de um fenômeno relacional, demonstrando a necessidade de clínicos avaliarem a qualidade da relação e o contexto que predispõe ao risco de infidelidade. / This study aimed to understand the factors involved in infidelity in married or cohabiting men and women considering four important domains: personal characteristics, partner characteristics, marriage and context. To answer the objectives of this thesis, the manuscript was developed over five studies. The first article performed a systematic review of the national and international literature about infidelity in the past five years to identify the infidelity research scenario. The results show some shortcomings in the researchs, such as lack of infidelity definition, great diversity of instruments and low methodological accuracy. The second study aimed to propose an infidelity mesure considering wich behaviors men and women define as emotional, sexual, emotional-virtual and sexual-virtual infidelity. This article evidences the absence of significant differences between the behaviors that men and women consider infidelity and suggest treating infidelity as a unique phenomenon, as far as the typology generates a reducionist view of the construct. The subsequent study investigated the experience of infidelity in men and women, describing frequency, types of behaviors and reasons for infidelity The results show similarites between men and women considering the infidelity behaviors and point to the main raison for infidelity was the partner and relationship dissatisfaction for both men and women. Then the last two studies aimed to investigate the differences between men and women who were unfaithful from those who remained faithful to the partner and to delineate the predictors of the infidelity. The findings pointed that the marriage variables have a great importance in differentiate the groups, besides being a predictor of infidelity with contextual variables. Thus, given the importance of the marriage domain for understanding the phenomenon, it is concluded that infidelity is a relational theme, demonstrating the need for clinicians to assess the quality of the relationship and the context that predispose to the risk of infidelity.

Introdução: A escala de Senso de Coerência (SOC), desenvolvida por Antonovsky, vem sendo utilizada no mundo inteiro, inclusive no Brasil. Embora seja crescente o número de publicações brasileiras envolvendo o tema, o processo de adaptação da escala não está claro na literatura. A versão adaptada não foi submetida a uma rigorosa análise fatorial, etapa importante no processo de validação. Objetivo: avaliar o processo de adaptação transcultural e as propriedades psicométricas da versão brasileira da escala de SOC. Dois objetivos teóricos foram propostos: 1) revisar sistematicamente estudos de adaptação transcultural da versão brasileira da escala de Senso de Coerência em português; 2) testar a estrutura dimensional da versão brasileira da escala de SOC com 13 itens. Método: Para a revisão sistemática foram pesquisadas quatro bases de dados eletrônicas para artigos publicados até dezembro de 2016: LILACS, PubMed, Scielo e Scopus com restrição para o Brasil, contendo vocabulário livre e controlado. Um revisor avaliou os critérios de elegibilidade e extraiu os dados. Foram identificadas 60 referências, mas após leitura de títulos e resumos foram selecionados 21 artigos e 2 teses, mas no final 17 publicações foram incluídas. Para a análise da dimensionalidade da escala, foram utilizados quatro bancos de dados de estudos brasileiros independentes usando a escala de SOC com 13. Amostra 1 foi realizado em uma amostra de 1760 mães e 1771 adolescentes em 36 cidades do Sul do Brasil. Amostra 2 compreendeu uma amostra de 1100 adultos de 20 anos de idade ou mais na cidade de São Leopoldo, sul do Brasil. Amostra 3 foi em uma amostra de 720 adultos e idosos em Porto Alegre, Brasil. Amostra 4 compreendeu 664 adolescentes estudantes da cidade de Goiânia, no Centro-Oeste do Brasil. Todos os quatro tinham desenho transversal e foram aprovados por diferentes Comitês de Ética em Pesquisa. Todas as análises psicométricas foram conduzidas com MPlus versão 7.11. Resultados: Resultado 1: Apenas um artigo realizou análise fatorial para escala SOC-29. Verificou-se que até o momento os estudos não consideraram etapas específicas para adaptação para cultura brasileira, apesar de apresentarem etapas metodológicas para validade de conteúdo e confiabilidade, consistência interna, 8 estudos (SOC-13) alfa de Cronbach variaram de 0,61 a 0,81 e 4 estudos (SOC-29) de 0,77 a 0,87 e os resultados foram satisfatórios. Um único estudo mostrou cargas fatoriais baixas para alguns itens para as análises dos componentes principais. Na AFC da amostra 1 os resultados mostraram que o ajuste total do modelo não foi aceitável (RMSEA.....;CFI=0.77; TLI=0.73;e WRMR=3.27) para o modelo de 1 fator e para 3 fatores (RMSEA=0.12; CFI=0.78; TLI=0.72; e WRMR=3.23). Nas amostras 2 e 3, a AFE mostrou dois eigenvalues maiores que 1 (4.39 e 1.67) e (4.72 e 1.60), respectivamente. Os itens SOC1 SOC2 e SOC3 formaram um fator não interpretável. Depois AFC foi realizada com a amostra 4, tirando os itens não interpretáveis, e o ajuste do modelo ficou aceitável. Conclusão: A avaliação das propriedades psicométricas da versão brasileira da escala de SOC mostrou que a escala precisa de ajustes. O modelo de 1 fator com 10 itens mostrou um bom ajuste estatístico, mas precisa ser revisado, por razões teóricas, se a remoção de itens afetou a validade de conteúdo da escala e se é capaz de medir todo o construto. Assim, estudos qualitativos podem ajudar a validade teórica do processo de adaptação transcultural. / Introduction: The Sense of Coherence (SOC) scale, developed by Antonovsky, has been used worldwide, including Brazil. Although the number of Brazilian publications involving the theme is increasing, the adaptation process of the scale is not clear and the adapted version has not been subjected to a rigorous factor analysis, an important step in the validation process. Objective: to evaluate the cross-cultural adaptation process and psychometric properties of the Brazilian version of the SOC scale. Two specific objectives are proposed: 1) systematically review cross-cultural adaptation studies of the Brazilian version of the SOC scale in Portuguese; and 2) to test the dimensional structure of the Brazilian version of the SOC scale with 13 items. Method: For the systematic review, four electronic databases were searched for articles published until December 2016: LILACS, PubMed, Scielo and Scopus with restriction for Brazil, containing free and controlled vocabulary. One reviewer assessed the eligibility criteria and extracted the data. We identified 60 articles, but after reading titles and abstracts, 21 papers and 2 theses were selected and finally 17 publications included. For the dimensional structure of the scale, four datasets of independent Brazilian studies using SOC scale with 13 items were used. Sample 1 was conducted in a sample of 1760 mothers and 1771 adolescents in 36 cities in southern Brazil. Sample 2 comprised 1100 adults aged 20 years or older in the city of São Leopoldo, south Brazil. Sample 3 comprised 720 adults and elders in Porto Alegre, Brazil. Sample 4 comprised 664 adolescent from Goiânia, in Mid-West Brazil. All four samples had cross-sectional design and were approved by the different Ethics Committee. All psychometric analyzes were conducted with MPlus version 7.1. Results: The systematic review showed that included studies have not consider all specific steps for cross-cultural adaptation, despite presenting methodological steps for content validity and reliability. Internal consistency for SOC-13 was reported in 8 studies with a Cronbach's alpha ranging from 0.61 to 0.81 and for SOC-29 in 4 studies ranging from 0.77 to 0.87. A single study showed low factor loads for some items in principal component analysis. In our CFA of sample 1, results showed an unacceptable fit ( CFI = 0.77; TLI = 0.73; and WRMR = 3.27) for 1 factor model and for 3 factors (RMSEA = 0.12; CFI = 0.78; TLI = 0.72; and WRMR = 3.23). The, we run EFA on samples 2 and 3 and they showed, respectively, two eigenvalues greater than 1 (4.39 and 1.67) and (4.72 and 1.60). Items SOC1, SOC2 and SOC3 formed an uninterpretable factor. After, we run another CFA using sample 4, removing the uninterpretable items, and model fit was acceptable. Conclusion: The evaluation of the psychometric properties of the Brazilian version of the SOC scale showed that the scale needs adjustments. The 1 factor model with 10 items showed a good statistical fit but needs to be revised on theoretical grounds if item removal affected the content validity of the scale and whether the new version is capable of measuring the entire construct. Thus, qualitative studies may help in theoretical validation of the cross-cultural adaptation.

A intolerância à incerteza (IU) é considerada um traço transdiagnóstico referente a crenças negativas sobre a incerteza e suas consequências. Ultimamente, a IU vem sendo associada com diversos transtornos relacionados à ansiedade. A presente dissertação teve o objetivo de realizar a adaptação transcultural e investigar propriedades psicométricas da Intolerance of Uncertainty Scale (IUS-12, Versão Abreviada) para uso no Brasil. O processo de adaptação transcultural foi dividido em sete etapas discutidas no primeiro estudo, onde também são apresentadas evidências de validade de conteúdo da escala. O segundo estudo teve como objetivo acumular evidências de validade de construto através de análise fatorial confirmatória e análise da consistência interna da IUS-12. Além disso, a validade convergente da escala foi avaliada através de suas correlações com instrumentos que medem construtos relacionados. Adicionalmente, são apresentadas normas e diretrizes para interpretação dos escores da escala. Os resultados da Análise Fatorial Confirmatória indicam que a IUS-12 possui estrutura fatorial semelhante a escala original contendo dois fatores: Intolerância à Incerteza Inibitória e Intolerância à Incerteza Prospectiva. Discute-se os demais resultados relativos às propriedades psicométricas da escala no segundo estudo e considera-se a IUS-12 como uma medida válida para a avaliação da intolerância à incerteza na população brasileira. / Intolerance of Uncertainty (IU) is considered a transdiagnostic factor reflecting negative beliefs about uncertainty and its consequences. IU has been associated with several anxiety-related disorders. The aim of this dissertation was to perform a cross-cultural adaptation and to investigate psychometric properties of the Intolerance of Uncertainty Scale (IUS-12, Short Version) for use in Brazil. The Cross-cultural adaptation process was divided into seven steps which are discussed in the first study, along with evidence of content validity of the scale. The second study aimed to accumulate evidence of construct validity throughout Confirmatory Factor Analysis and internal consistency analysis. In addition, a convergent validity was assessed through its correlations with measures of related constructs. Furthermore, norms and guidelines for scores interpretation are presented. The results of the Confirmatory Factorial Analysis indicate that the IUS-12 has a factorial structure similar to the original scale comprising two factors: Inhibitory IU and Prospective IU. Further results regarding psychometric properties of the scale are discussed in the second study. The IUS-12 appears to be a valid measure for assessment of intolerance to uncertainty in the Brazilian population.

O bullying é um problema comum entre jovens em idade escolar, nos diferentes países, culturas e níveis socioeconômicos. Trata-se de um comportamento agressivo, ofensivo, repetitivo e frequente, perpetrado por uma pessoa contra outra ou por um grupo contra outros, com a intenção de ferir e humilhar, estabelecendo-se uma relação desigual de poder. O envolvimento com bullying está associado a pior ajustamento psicossocial, problemas de aprendizagem, evasão escolar ou de trocas frequentes de escolas, entre outros. Embora o comportamento de bullying seja multicausal, ainda são escassos estudos que avaliem a relação entre habilidades sociais de adolescentes e bullying. Um dos aspectos que dificultam a realização de pesquisas nessa área está relacionado à falta de instrumentos validados. Os objetivos do presente estudo são verificar a validade de constructo do Questionário de Bullying de Olweus (QBO) versão agressor e versão vítima e verificar a associação entre habilidades sociais e bullying em adolescentes. Trata-se de um estudo com alunos de ambos os sexos, oriundos de escolas da rede pública de Porto Alegre, do 5º ao 9º ano do ensino fundamental, com idade entre 10 e 17 anos. Para verificar o envolvimento com bullying, utilizou-se o QBO, com 23 questões para versão vítima e 23 para versão agressor com quatro opções de resposta (1=nenhuma vez a 4=várias vezes por semana). A validade de constructo foi verificada com a Teoria de Resposta ao Item (TRI), utilizando-se o modelo de resposta gradual e o crédito parcial generalizado. As habilidades sociais foram avaliadas com o Inventário de Habilidades Sociais para Adolescentes (IHSA), que é validado no Brasil. Para verificar a associação entre bullying e habilidades sociais, o critério de idade foi de 12 a 17 anos, conforme a recomendação do IHSA. O estudo foi aprovado pelo Comitê de Ética em Pesquisa da Universidade Federal do Rio Grande do Sul e da Secretaria Municipal de Saúde de Porto Alegre (CAAE 19651113.5.0000.5338). Os instrumentos, ambos de autorrelato, foram respondidos pelos adolescentes no horário de aula e após a autorização dos pais. Os resultados estão apresentados em dois artigos. No primeiro, foi incluído um total de 703 adolescentes, sendo 380(54%) meninas, com média de idade de 13(DP=1,58) anos. Após a análise para a construção dos escores finais do QBO pelo modelo da TRI, observou-se que a probabilidade de um adolescente responder à opção 3 (uma vez por semana) é zero para ambas as versões. De acordo com as curvas característica do item (CCI), optou-se pela unificação das alternativas 3 e 4 para mensuração mais fidedigna a realidade do comportamento de bullying. Os itens com maior discriminação para classificar como vítima foram, respectivamente, 20 (Disseram coisas maldosas sobre mim ou sobre a minha família); 15 (Fui perseguido[a] dentro ou fora da escola) e 3 (Me ameaçaram). Na versão agressor, os itens com maior discriminação foram, respectivamente, 22 (Forcei a agredir outro[a] colega); 15 (Persegui dentro ou fora da escola) e 3 (Fiz ameaças). No segundo artigo, foram incluídos 467 alunos, sendo 245(52,5%) do sexo feminino, com média de idade de 13,3(DP=1,18) anos. Considerando-se a interação habilidades sociais e sexo, as meninas apresentaram associação significativa com menor frequência do autocontrole (p=0,010) e da civilidade (p=0,031) e maior dificuldade das habilidades de autocontrole (p=0,033) e desenvoltura social (p=0,009). Em relação aos tipos de envolvimento com bullying, 59(12,6%) dos adolescentes classificaram-se como vítima, 60(12,8%) como agressores e 175(37,5%) como agressores vítimas. Observou-se associação significativa entre as meninas vítimas de bullying e maior dificuldade na habilidade de empatia comparada aos meninos (p=0,012) e aos demais tipos de bullying (p=0,022). Também foram as vítimas, independentemente do sexo, que apresentaram maior dificuldade em termos de autocontrole, assertividade, abordagem afetiva e o total das habilidades sociais em comparação aos não envolvidos (p<0,05). Os resultados sugerem que a utilização da TRI permite a construção de uma medida de avaliação mais objetiva e precisa do comportamento de bullying. Por meio da validação de constructo do QBO e da associação com habilidades sociais, o estudo demonstrou que existe um importante déficit de determinadas habilidades sociais nos diferentes tipos de envolvimentos com bullying, principalmente entre as vítimas, quando comparadas aos não envolvidos. Portanto, intervenções que incluam técnicas para melhorar as habilidades sociais podem desempenhar um relevante papel preventivo no envolvimento com bullying no ambiente escolar. / Bullying is a common problem among school-age children and adolescents across different countries, cultures and socioeconomic levels. Bullying consists of the repeated infliction of aggressive and offensive behaviors by one person against another or one group against others, with the intent of hurting or humiliating the victim, resulting in an imbalance of power. Involvement in bullying is associated with issues such as poor social adjustment, learning difficulties, truancy and frequent changes of schools. Although bullying is known to be a multicausal phenomenon, few studies have evaluated the relationship between social skills and bullying in adolescents. A major challenge to research in the area is the lack of validated instruments which can be used to assess either of these variables. The goals of the present study were therefore to evaluate the construct validity of the Olweus Bully/Victim Questionnaire (OBVQ) and investigate the association between social skills and bullying in adolescents. This study involved participants of both genders aged between 10 and 17 years recruited from grades five through nine of public schools in the city of Porto Alegre. Involvement in bullying was evaluated using the OBVQ, which consists of 23 victim questions and 23 bully questions, scored on a scale of 1 (never) to 4 (several times a week). Construct validity was investigated using item response theory (ITR), by means of graduated response and generalized partial credit models. Social skills were evaluated using the Adolescent Social Skills Inventory (ASSI), which has been validated for use in Brazilian adolescents. The association between bullying and social skills was investigated in a sample of 12- to 17-year olds, which corresponds to the age group for which the ASSI was validated. This study was approved by the Research Ethics Committee of the Federal University of Rio Grande do Sul and the Municipal Health Service of Porto Alegre (CAAE 19651113.5.0000.5338). The OBVQ and ASSI, both of which are self-report instruments, were administered to adolescents during school hours following parental consent. The results of this procedure are presented in two articles. The first involved a sample of 703 adolescents, of whom 380 (45%) were girls, with a mean age of 12 (SD=1.58) years. The analysis of the IRT model for OBVQ scores revealed that the probability of an adolescent responding to any of the items with option 3 (once a week) was zero in both versions of the questionnaire. Therefore, based on item characteristic curves (ICC), alternatives 3 and 4 were combined to ensure a more reliable measure of bullying behaviors. The most discriminating items in the victim questionnaire were items 20 (One or more classmates said bad things about me or my family); 15 (I was persecuted inside or outside the school) and 3 (I was threatened). The most discriminating items in the bully version were items 22 (Forced someone to hit a classmate); 15 (Persecuted a classmate inside or outside the school) and 3 (Made threats). The second article involved 467 students, of whom 245(52.5%) were female. The mean age of the sample was 13.3 (SD=1.18) years. An interaction between social skills and gender was identified, whereby girls were found to engage less frequently in the behaviors described in the selfcontrol (p=0.010) and civility scales (p=0.031), and found it more difficult to perform the behaviors listed in the self-control (p=0.033) and social ease scales (p=0.009). An analysis of student involvement in bullying revealed that 59 (12.6%) participants were considered victims, 60(12.8%) were classified as bullies and 175(37.5%) were categorized as bullyvictims. Female victims of bullying had more difficulty expressing empathy than boys (p=0.012) and participants with different types of involvement in bullying (p=0.022). Bullying victims of both genders also had more difficulty engaging in the behaviors listed in the self-control, assertiveness, and affective approach subscales, and obtained higher total scores on the difficulty engaging in ASSI behaviors than adolescents not involved in bullying (p<0.05). The results suggested that IRT can be used to develop a more objective and precise measure of bullying. By construct validating the OBVQ and verifying its association with social skills, the present study showed that the latter are significantly impaired in adolescents involved in bullying, especially the victims, as compared to those not involved. Therefore, interventions involving the improvement of social skills may play an important role in preventing bullying in schools.

abstract: The use of exams for classification purposes has become prevalent across many fields including professional assessment for employment screening and standards based testing in educational settings. Classification exams assign individuals to performance groups based on the comparison of their observed test scores to a pre-selected criterion (e.g. masters vs. nonmasters in dichotomous classification scenarios). The successful use of exams for classification purposes assumes at least minimal levels of accuracy of these classifications. Classification accuracy is an index that reflects the rate of correct classification of individuals into the same category which contains their true ability score. Traditional methods estimate classification accuracy via methods which assume that true scores follow a four-parameter beta-binomial distribution. Recent research suggests that Item Response Theory may be a preferable alternative framework for estimating examinees' true scores and may return more accurate classifications based on these scores. Researchers hypothesized that test length, the location of the cut score, the distribution of items, and the distribution of examinee ability would impact the recovery of accurate estimates of classification accuracy. The current simulation study manipulated these factors to assess their potential influence on classification accuracy. Observed classification as masters vs. nonmasters, true classification accuracy, estimated classification accuracy, BIAS, and RMSE were analyzed. In addition, Analysis of Variance tests were conducted to determine whether an interrelationship existed between levels of the four manipulated factors. Results showed small values of estimated classification accuracy and increased BIAS in accuracy estimates with few items, mismatched distributions of item difficulty and examinee ability, and extreme cut scores. A significant four-way interaction between manipulated variables was observed. In additional to interpretations of these findings and explanation of potential causes for the recovered values, recommendations that inform practice and avenues of future research are provided. / Dissertation/Thesis / M.A. Educational Psychology 2013

abstract: ABSTRACT The phenomenon of cyberbullying has captured the attention of educators and researchers alike as it has been associated with multiple aversive outcomes including suicide. Young people today have easy access to computer mediated communication (CMC) and frequently use it to harass one another -- a practice that many researchers have equated to cyberbullying. However, there is great disagreement among researchers whether intentional harmful actions carried out by way of CMC constitute cyberbullying, and some authors have argued that "cyber-aggression" is a more accurate term to describe this phenomenon. Disagreement in terms of cyberbullying's definition and methodological inconsistencies including choice of questionnaire items has resulted in highly variable results across cyberbullying studies. Researchers are in agreement however, that cyber and traditional forms of aggression are closely related phenomena, and have suggested that they may be extensions of one another. This research developed a comprehensive set of items to span cyber-aggression's content domain in order to 1) fully address all types of cyber-aggression, and 2) assess the interrelated nature of cyber and traditional aggression. These items were administered to 553 middle school students located in a central Illinois school district. Results from confirmatory factor analyses suggested that cyber-aggression is best conceptualized as integrated with traditional aggression, and that cyber and traditional aggression share two dimensions: direct-verbal and relational aggression. Additionally, results indicated that all forms of aggression are a function of general aggressive tendencies. This research identified two synthesized models combining cyber and traditional aggression into a shared framework that demonstrated excellent fit to the item data. / Dissertation/Thesis / Ph.D. Educational Psychology 2013

abstract: Fibromyalgia (FM) is a chronic pain condition characterized by debilitating fatigue. This study examined the dynamic relation between interpersonal enjoyment and fatigue in 102 partnered and 74 unpartnered women with FM. Participants provided three daily ratings for 21 days. They rated their fatigue in late morning and at the end of the day. Both partnered and unpartnered participants reported their interpersonal enjoyment in the combined familial, friendship, and work domains (COMBINED domain) in the afternoon. Additionally, partnered participants reported their interpersonal enjoyment in the spousal domain. The study was guided by three hypotheses at the within-person level, based on daily diaries: (1) elevated late morning fatigue would predict diminished afternoon interpersonal enjoyment; (2) diminished interpersonal enjoyment would predict elevated end-of-day fatigue; (3) interpersonal enjoyment would mediate the late morning to end-of-day fatigue relationship. In cross-level models, the study explored whether individual differences (between-person) in late morning fatigue and afternoon interpersonal enjoyment would moderate within-person relations from late morning fatigue to afternoon interpersonal enjoyment, and from afternoon interpersonal enjoyment to end-of-day fatigue. Furthermore, it explored whether the hypothesized relationships at the within-person level would also emerge at the between-person level (between-person mediation models). Multilevel structural equation modeling and multilevel modeling were employed for model testing, separately for partnered and unpartnered participants. Within-person mediation models supported that on high fatigue mornings, afternoon interpersonal enjoyment was dampened in the spousal and combined domains in partnered and unpartnered samples. Moreover, low afternoon interpersonal enjoyment in both the spousal and combined domains predicted elevated end-of-day fatigue. Afternoon interpersonal enjoyment mediated the relationship of late morning to end-of-day fatigue in the combined domain but in not the spousal domain. Cross-level moderation analyses showed that individual differences in afternoon spousal enjoyment moderated the day-to-day relation between afternoon spousal enjoyment and end-of-day fatigue. Finally, the mediational chain was not observed at the between-person level. These findings suggest that preserving interpersonal enjoyment in non-spousal relations limits within-day increases in FM fatigue. They highlight the importance of examining domain-specificity in interpersonal enjoyment when studying fatigue, and suggest that targeting enjoyment in social relations may improve the efficacy of existing treatments. / Dissertation/Thesis / Ph.D. Psychology 2013

abstract: In order to analyze data from an instrument administered at multiple time points it is a common practice to form composites of the items at each wave and to fit a longitudinal model to the composites. The advantage of using composites of items is that smaller sample sizes are required in contrast to second order models that include the measurement and the structural relationships among the variables. However, the use of composites assumes that longitudinal measurement invariance holds; that is, it is assumed that that the relationships among the items and the latent variables remain constant over time. Previous studies conducted on latent growth models (LGM) have shown that when longitudinal metric invariance is violated, the parameter estimates are biased and that mistaken conclusions about growth can be made. The purpose of the current study was to examine the impact of non-invariant loadings and non-invariant intercepts on two longitudinal models: the LGM and the autoregressive quasi-simplex model (AR quasi-simplex). A second purpose was to determine if there are conditions in which researchers can reach adequate conclusions about stability and growth even in the presence of violations of invariance. A Monte Carlo simulation study was conducted to achieve the purposes. The method consisted of generating items under a linear curve of factors model (COFM) or under the AR quasi-simplex. Composites of the items were formed at each time point and analyzed with a linear LGM or an AR quasi-simplex model. The results showed that AR quasi-simplex model yielded biased path coefficients only in the conditions with large violations of invariance. The fit of the AR quasi-simplex was not affected by violations of invariance. In general, the growth parameter estimates of the LGM were biased under violations of invariance. Further, in the presence of non-invariant loadings the rejection rates of the hypothesis of linear growth increased as the proportion of non-invariant items and as the magnitude of violations of invariance increased. A discussion of the results and limitations of the study are provided as well as general recommendations. / Dissertation/Thesis / Ph.D. Psychology 2013