Behavioral and emotional problems (BEPs) are known to affect children's ability to shape and maintain effective social relationships. BEPs are typically categorized into two main factors: internalizing and externalizing behaviors. Internalizing behaviors represent introverted problems, directed inwardly to the individual. While externalizing behavior patterns represent behaviors that are directed outwardly. Behaviors, emotions and thoughts are experienced by all people but on a continuum rather than in terms of absence versus presence of the behavior. The child behavior checklist (CBCL) is used to measure BEPs. The system of CBCL (parent form) measures also includes a teacher rating form and a youth self-report. Using 62 teachers and 311 students, the present study assessed convergent and discriminant validity using a correlated trait, correlated method minus one [CT-C(M-1)] model. The results showed low to moderate teacher-student agreement on the traits. To extend the theoretical structure of the teacher and self-report forms, the present study assessed the nested structure of the data using a multilevel model. Results revealed the nested structure of the data should not be ignored.

The purpose of this research was to develop and test an instrument to measure or quantify the construct of inner strength. Inner strength was synthesized into a theory based on twelve years of qualitative data from direct participant quotes, defined by Roux et al. (2002) as a central human resource that promotes well-being and healing. The inner strength questionnaire (ISQ) has undergone extensive psychometric evaluation which resulted in several versions. The third version had 37-items and was believed to have four subscales that mirrored the theoretical themes that emerged through the qualitative data. The sample was composed of 281 women with a variety of chronic health conditions including breast cancer, multiple sclerosis, diabetes, and heart disease. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were employed to assess the structure of the ISQ. Internal consistency was used to assess the reliability of each hypothesized subscale and the entire ISQ. Convergent and discriminant validity, were analyzed using the multitrait-multimethod by Campbell and Fiske (1959). For convergent validity, the entire instrument and each of the subscales was correlated with similar theoretical subscales of the Mastery of Stress Instrument (Younger, 1993). For discriminant validity, the Center for Epidemiological Studies in Depression (CESD) was correlated with the entire ISQ. Both the EFA and CFA revealed a four factor model. The four factors were labeled 1) mental, 2) connectedness, 3) knowing and searching, 4) physical. The reliability for all items on the ISQ was 0.91. The reliabilities for each of the subscales were 0.85 for mental, 0.96 for connectedness, 0.85 for knowing and searching, and 0.83 for physical. The multitrait-multimethod matrix revealed correlation coefficients for the relationship between the ISQ and MSI to be 0.55 and a correlation coefficient of -0.19 between the ISQ and the CESD.A four factor model is supported by the statistical data. Through face-to-face interviewing and input from content expert reviewers, ten items were eliminated from the 37-item (version 3) instrument to form a new version of the instrument. The new instrument supports the metasynthesis generated by Roux (2002) except for the fifth theme, entitled "new normal", which was found to be a consequence of having inner strength, not a component of inner strength.

The aim of this research programme was to examine the 3P model of learning (Biggs, 1987a, 1999). The first stage necessarily involved an examination of the Study Process Questionnaire (SPQ) (Biggs, 1987a), an instrument developed to measure the process component of the model. The structure of the SPQ was examined utilising exploratory and confirmatory factor analysis of undergraduate responses (n= 260). The results indicated the higher-order factor structure of deep-achieving and surface-achieving-motive provided the most reliability and a better model fit than either the subscales or scales of the SPQ. The construct validity of the two constructs deep and surface was assessed next using a multitrait-multimethod matrix (MTMM) constructed from the three measures of the self-report questionnaire, interview ratings and written assessments from first-year students (n = 50). The results indicated good convergent validity between the deep scale of the SPQ and the interview ratings on the deep scale, between the deep scale on the SPQ and the written assessment ratings, and between the interview ratings and written assessment ratings. The results indicated good convergent validity between the surface scale on the SPQ and the interview ratings on the surface scale, but not between the surface scale on the SPQ and the written assessment ratings, and between the interview ratings and written assessment ratings. The discriminant validity between deep and surface was good for the SPQ, but not for either the interview or the written assessment. The findings indicate the deep and surface scales of the SPQ adequately measure the underlying deep and surface constructs. The retest reliability of the SPQ was then examined utilising Spearman’s Rho to assess the rank-order correlations with a sample of third-year students (n=87). Over a period of three months there were significant correlations for the surface motive, surface strategy, deep strategy, achieving motive and achieving strategy subscales of the SPQ, suggesting good reliability for these subscales. The results at the scale level of the SPQ result in similar conclusions. There was a moderate significant correlation for the surface, deep and achieving scales of the SPQ, suggesting the scales have good reliability over a period of three months. There was also a moderate significant correlation for the surface-achieving-motive and deep-achieving scales over a period of three months. The stability of SPQ scores was also assessed utilising a series of one-way repeated measures MANOVA’s with a sample of third-year undergraduates (n = 64). The results suggest some change occurs in self-reported use of approaches to learning between the first and third-years of an undergraduate degree programme. The role of the teaching-learning environment was next examined. Utilising a within-subjects design, undergraduate students (n=48) concurrently enrolled in traditional (viz. lecture and tutorial) and non-traditional (viz. workshops and group projects) subjects completed the SPQ to describe their approaches to learning in each subject. A series of 2x2 repeated measures MANOVA’s were undertaken. The results indicated students were likely to change their approach to learning based on their perceptions of the learning environment (traditional or non-traditional subject). However, those students identified as predominantly surface learners significantly increased their deep scale scores in the non-traditional subject when compared to deep learners. The next study examined a range of personality (locus of control, sensing function, thinking function, intelligence) and demographic variables (age, gender, year of study) to assess which were good predictors of deep and surface approaches to learning. A series of regression analyses identified age, sensing function and locus of control as significant predictors of the surface, surface-achieving-motive, and deep approaches to learning. Locus of control was found to be a significant predictor of the deep-achieving approach to learning. The final study examined the 3P model of learning. Based on the results of earlier studies in the research programme the situational component of the presage factors was not included. The model was examined using structural equation modelling (n= 394). Two initial models were tested using both the three (deep, surface, achieving) and two (surface-achieving-motive and deep-achieving) process factor models. The three process factor model provided the better model fit. The results suggest deep and surface approaches to learning do not mediate between personal characteristics and learning outcomes (i.e. GPA). The results of this series of studies suggest the need for further research into the SPQ and the 3P model of learning. The implications of the research programme are also discussed.

There continues to be doubt regarding the validity of laboratory aggression paradigms. This paper provides an investigation of the construct validity of one prominent aggression task, the Taylor Aggression Paradigm (TAP), within a Multitrait Multimethod Matrix (MTMM) methodology. Participants consisted of 151 male undergraduate psychology students with a median age of 19 years old (M=19.45, SD = 2.03). Participants completed self-report and behavioral measures of aggression, impulsivity, and pro-social behavior which were analyzed using a Correlated Trait – Correlated Method Confirmatory Factor Analysis model. Results supported the construct validity of the MTMM model and the TAP. This study provides one of the only a priori tests of construct validity for the TAP and provides a basis for additional validation studies using this methodology.

Sambandsanalys av surveydata kompliceras av förekomsten av slumpfel och metodeffekter i de avgivna svaren. Dessa felkällor kan leda till betydande över- eller underskattning av sambanden mellan undersökta variabler, och riskerar att leda till felaktiga slutsatser. En sedan länge känd metod för att estimera och kontrollera för slumpfel och metodeffekter i enkätdata är den så kallade Multitrait Multimethod-metoden (MTMM). Nackdelen med MTMM-metoden är att den kräver att varje fråga som ska analyseras ställs minst tre gånger i samma enkät, vilket gör att ansatsen i praktiken oftast inte är möjlig att tillämpa. Sedan 2012 finns dock ett verktyg, SQP 2.0, som gör det möjligt att skatta slumpfel och metodeffekter i enkätdata utan att genomföra MTMM-experiment. Den här uppsatsen utvärderar prediktionerna från SQP 2.0 på en enkät om arbetslivsrelaterade frågor besvarad av svenska och brittiska respondenter. Programvarans prediktioner jämförs med resultat från MTMM-experiment, och effekten på sambandsanalysen av att tillämpa SQP-programvarans prediktioner av slumpfel och metodeffekter studeras. Slutsatserna från studien är att SQP 2.0 ger predicerade värden för slumpfel och metod-effekter som ligger nära de MTMM-estimerade i de flesta fall, men att betydande avvikelser också förekommer. Vidare konstateras att korrigering för slumpfel och metodeffekter har en betydande effekt på absolutvärdet av korrelationerna mellan variabler under de studerande förhållandena: korrelationerna i det studerade fallet ökar kraftigt efter justering. Korrelationernas relativa storlek förändras i mindre utsträckning, men för en enkät med mer varierade frågetyper kan också dessa påverkas kraftigt. / Correlation analysis of survey data is complicated by the presence of random errors and method effects in the answers given. These factors can lead to significant over- or underestimation of the correlations between variables. A well-established method for estimating and controlling for random error and method effects in survey data is the Multitrait-Multimethod (MTMM) approach. The disadvantage of this method is that it requires that each survey item to be analysed is asked multiple times in the same survey, making the approach hard to use in many practical situations. Since 2012 there is a tool available, SQP 2.0, to predict random errors and method effects in survey data without performing MTMM experiments. This paper evaluates the use of predictions from the SQP 2.0 software on the correlation analysis of a survey of work related matters answered by Swedish and British respondents. The software's predictions are compared with results from MTMM experiments, and the effect on the correlations of applying the SQP software's predictions of random errors and method effects are studied. The conclusion of the study is that SQP 2.0 gives predicted values for random error and method effects that are close to the MTMM-estimates in most cases, but that considerable deviations also occur. It is further concluded that controlling for random error and method effects has a significant effect on the absolute values of the correlations between variables in the studied cases: the correlations in the study increase substantially after adjustment. The relative sizes of the correlations between variables change to a lesser extent, but a questionnaire with more varied question types may have lead to different results in this respect.

Resumo: Foram estimados parâmetros genéticos para pesos do nascimento à idade adulta de animais da raça Canchim por meio de análises uni, bi e multicaracterística, e por modelos de regressão aleatória. Os dados analisados são provenientes do rebanho da raça Canchim da Embrapa Pecuária Sudeste, localizada no município de São Carlos, São Paulo. Os pesos foram obtidos do nascimento até os nove anos e meio de idade. Nas análises uni, bi e multicaracterística foram utilizados pesos em ao nascimento, desmama, 12, 18, 24 e 30 meses de idade e, na idade adulta. Na análise unicaracterística foram utilizados quatro modelos, em que, diferentes efeitos aleatórios (efeito genético materno e de ambiente permanente materno) foram adicionados em seqüência. Nas análises de regressão aleatória, foram utilizados pesos de fêmeas do nascimento até os nove anos e meio de idade, considerando como funções base polinômios de Legendre e funções "b-splines". A variância residual foi modelada utilizando 1, 4, 11 e 19 classes. Foram utilizados 12 modelos de regressão aleatória sobre polinômios de Legendre de segunda à sétima ordem para modelar a trajetória da variância dos efeitos genético aditivo direto e aditivo materno e de ambiente permanente direto e materno. Vinte modelos de regressão aleatória sobre funções "bsplines" foram considerados, empregando polinômios linear, quadrático e cúbico para cada segmento individual. Polinômios do mesmo grau foram considerados no modelo para todos os efeitos aleatórios. Até sete segmentos foram utilizados para os efeitos genético direto e de ambiente permanente do animal. Para os efeitos genético aditivo materno e de ambiente permanente materno foi utilizado um único segmento com dois nós nos extremos da curva Os efeitos maternos influenciam os pesos do nascimento aos dois anos de idade sendo o peso à desmama... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Genetic parameters were estimated for weights taken from birth to mature age in Canchim (5/8 Charolais + 3/8 Zebu) cattle of breed by one, two and multitrait analyses and by random regression models. The data analyzed were from a herd of Canchim beef cattle belonging to Embrapa's Southeast Cattle Research Center, located in São Carlos county, state of São Paulo. Weights were taken from birth to nine and half years of age. Weights at birth, weaning, 12, 18, 24 and 30 months of age and at mature age were analyzed using one, two and multitrait models. In onetrait analyses, four models were tested, in which different random effects (genetic and environment permanent maternal effects) were added. For random regression models age of cow varied from birth to 3.542 days of age. Legendre polynomials and b-splines functions of age at recording were used as basis functions in random regression models. Residual variances were modeled by a step function with 1, 4, 11 and 19 classes. A total of 12 random regression models using Legendre polynomials as basis functions, from second to seventh order, were used to model direct and maternal genetic effects, animal and maternal permanent environmental effects. A total of twenty analyses, considering linear, quadratic and cubic b-splines functions and up to nine knots, were carried out. Spline functions of the same order were considered for all random effects. Maternal effects influenced weights from birth until two years of age, being weaning weight the most affected by maternal effects. Direct heritabilities obtained by twotrait and multitrait analyses were higher than estimates obtained from onetrait analyses. In order to estimate genetic parameters for weights after selection it is important to consider weights before selection and the multitrait analyses is the most adequate. The model with direct and maternal genetic effects, animal and maternal... (Complete abstract click electronic access below) / Orientador: Maurício Mello de Alencar / Coorientadora: Lúcia Galvão de Albuquerque / Banca: Humberto Tonhati / Banca: Henrique Nunes de Oliveira / Banca: Lenira El Faro Zadra / Banca: Maria Eugênia Zerlotti Mercadante / Doutor

This dissertation studies the impact of the mode of data collection on the quality of answers to survey questions, defined as the product of reliability and validity. Using data from the Netherlands about different topics (media, social and political trust, satisfaction, political orientation, left-right self-placement, attitudes toward immigration), it shows that the quality is similar in a computed assisted face-to-face survey using show cards (the European Social Survey, ESS) and a web survey based on a probability sample (the LISS panel). This is true both at the level of single items and composite scores. It suggests that standardised relationships across variables can be compared across these two modes. On the contrary, telephone interviews lead to some differences in quality. For complex concepts, measurement equivalence also holds, meaning that means and unstandardised relationships can be compared across the faceto- face and web surveys mentioned previously. / Esta tesis estudia el impacto que el método de recolección de datos en encuestas tiene sobre la calidad de las respuestas, definida como el producto de la fiabilidad y la validez. Utilizando datos de Holanda sobre temas diversos (utilización de los medios de comunicación, confianza social y política, satisfacción, orientaciones políticas, autoubicación en la escala izquierda-derecha, actitudes hacia la inmigración), se muestra que la calidad es similar en una encuesta cara-a-cara asistida con ordenador y utilizando tarjetas (la Encuesta Social Europea) y una encuesta online basada en una muestra probabilística (el panel LISS). Esto se cumple tanto para los indicadores simples, como para indicadores complejos. Los resultados sugieren que las relaciones estandardizadas entre variables son comparables entre los dos métodos de recolección. Al contrario, las entrevistas telefónicas producen diferencias de calidad. Para conceptos complejos, la equivalencia de las mediciones también está garantizada: las medias y las relaciones no estandarizadas son comparables en las entrevistas cara-a-cara y online.

Made available in DSpace on 2014-06-11T19:32:15Z (GMT). No. of bitstreams: 0 Previous issue date: 2008-09-26Bitstream added on 2014-06-13T20:03:29Z : No. of bitstreams: 1 baldirey_fs_dr_jabo.pdf: 751142 bytes, checksum: 4a3bdaf84424f12abb37e90cff761fdc (MD5) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Foram estimados parâmetros genéticos para pesos do nascimento à idade adulta de animais da raça Canchim por meio de análises uni, bi e multicaracterística, e por modelos de regressão aleatória. Os dados analisados são provenientes do rebanho da raça Canchim da Embrapa Pecuária Sudeste, localizada no município de São Carlos, São Paulo. Os pesos foram obtidos do nascimento até os nove anos e meio de idade. Nas análises uni, bi e multicaracterística foram utilizados pesos em ao nascimento, desmama, 12, 18, 24 e 30 meses de idade e, na idade adulta. Na análise unicaracterística foram utilizados quatro modelos, em que, diferentes efeitos aleatórios (efeito genético materno e de ambiente permanente materno) foram adicionados em seqüência. Nas análises de regressão aleatória, foram utilizados pesos de fêmeas do nascimento até os nove anos e meio de idade, considerando como funções base polinômios de Legendre e funções “b-splines”. A variância residual foi modelada utilizando 1, 4, 11 e 19 classes. Foram utilizados 12 modelos de regressão aleatória sobre polinômios de Legendre de segunda à sétima ordem para modelar a trajetória da variância dos efeitos genético aditivo direto e aditivo materno e de ambiente permanente direto e materno. Vinte modelos de regressão aleatória sobre funções “bsplines” foram considerados, empregando polinômios linear, quadrático e cúbico para cada segmento individual. Polinômios do mesmo grau foram considerados no modelo para todos os efeitos aleatórios. Até sete segmentos foram utilizados para os efeitos genético direto e de ambiente permanente do animal. Para os efeitos genético aditivo materno e de ambiente permanente materno foi utilizado um único segmento com dois nós nos extremos da curva Os efeitos maternos influenciam os pesos do nascimento aos dois anos de idade sendo o peso à desmama... / Genetic parameters were estimated for weights taken from birth to mature age in Canchim (5/8 Charolais + 3/8 Zebu) cattle of breed by one, two and multitrait analyses and by random regression models. The data analyzed were from a herd of Canchim beef cattle belonging to Embrapa’s Southeast Cattle Research Center, located in São Carlos county, state of São Paulo. Weights were taken from birth to nine and half years of age. Weights at birth, weaning, 12, 18, 24 and 30 months of age and at mature age were analyzed using one, two and multitrait models. In onetrait analyses, four models were tested, in which different random effects (genetic and environment permanent maternal effects) were added. For random regression models age of cow varied from birth to 3.542 days of age. Legendre polynomials and b-splines functions of age at recording were used as basis functions in random regression models. Residual variances were modeled by a step function with 1, 4, 11 and 19 classes. A total of 12 random regression models using Legendre polynomials as basis functions, from second to seventh order, were used to model direct and maternal genetic effects, animal and maternal permanent environmental effects. A total of twenty analyses, considering linear, quadratic and cubic b-splines functions and up to nine knots, were carried out. Spline functions of the same order were considered for all random effects. Maternal effects influenced weights from birth until two years of age, being weaning weight the most affected by maternal effects. Direct heritabilities obtained by twotrait and multitrait analyses were higher than estimates obtained from onetrait analyses. In order to estimate genetic parameters for weights after selection it is important to consider weights before selection and the multitrait analyses is the most adequate. The model with direct and maternal genetic effects, animal and maternal... (Complete abstract click electronic access below)

Les coûts de la reproduction sont un compromis biologique (trade-off ) fondamental en théorie des histoires de vie. Par ce compromis, le succès, pour un organisme, d’un évènement de reproduction réduit sa survie et sa fertilité futures. Pour les écologues, ce trade-off correspond principalement à un compromis physiologique résultant d’un processus d’allocation ayant lieu à chaque instant et au niveau de chaque individu. Au contraire, en démographie évolutive, il est envisagé comme un trade-off génétique découlant du polymorphisme génotypique d’un gène pléiotropique agissant de manière antagoniste sur la reproduction aux jeunes âges et la fitness aux âges élevés. L’étude des mécanismes des coûts de la reproduction, physiologiques et génétiques, de leur possible cohabitation et de leur effets relatifs, croisés et conjoints est le sujet de cette thèse. Un examen attentif de la définition originelle des coûts de la reproduction par Williams (1966), nous permet de construire un modèle théorique des coûts physiologiques intégrant leurs aspects mécaniques et évolutifs. Cette construction nous permet d’induire l’intensité des coûts de la reproduction selon la position d’un organisme sur trois continuums d’histoire de vie: "slow-fast", "income-capital breeders" et "quantity-quality".A partir de la décomposition, par Stearns (1989b), de l’architecture des contraintes d’histoire de vie en trois parties – le niveau génotypique, la structure intermédiaire et le niveau phénotypique – nous étendons notre modèle conceptuel pour y intégrer à la fois des trade-offs physiologiques et génétiques. Cela nous permet d’inférer les effets de l’environnement, de sa variance et de la stochasticité individuelle sur la détectabilité de chaque famille de coûts. La différence entre coûts physiologiques et génétiques se retrouve également dans leur modélisation mathématique. Il est donc nécessaire de développer de nouveaux modèles permettant d’incorporer coûts physiologiques et génétiques. Nous proposons ensuite une méthode vectorielle de construction d’un tel type de modèle, que nous appelons Matrice de Projection de Population Multitrait (MPPM). Ce dernier peut implémenter chaque type de coût en l’intégrant dans la matrice en tant que trait. Nous étendons ensuite aux MPPMs les techniques d’analyse de sensibilité, standards en démographie évolutive, des modèles à un trait aux MPPMs. Surtout, nous décrivons un nouvel outil d’analyse, pertinent en théorie des histoires de vie et en démographie évolutive: la Trait Level Analysis. Elle consiste à comparer des modèles qui partagent les mêmes propriétés asymptotiques. Ceci est rendu possible par le repliement d’une MPPM selon certains traits, une opération qui réduit le nombre de traits du modèle en moyennant ses transitions selon les abondances ergodiques relatives. Ainsi, la Trait Level Analysis permet de mesurer l’importance évolutive des coûts de la reproduction en comparant des modèles implémentant ces coûts, avec des versions ergodiquement équivalentes de ces modèles mais repliées selon les traits supportant les compromis. Nous utilisons des méthodes, classiques et nouvelles, de calculs des moments de la fitness – gradient de sélection, variance du succès reproducteur, variance environnementale – que nous appliquons aux modèles avec coûts et sans coûts afin de mesurer leurs effets démographiques et évolutifs. Nous présentons les effets conjoints des coûts physiologiques et génétiques sur la distribution par âge des taux vitaux d’une population. Nous montrons également comment les coûts physiologiques influencent les deux composants de la sélection efficace, en aplatissant le gradient de sélection d’un côté et en accroissant la taille efficace de la population de l’autre. Enfin, nous démontrons comment l’effet tampon des coûts sur les variances environnementales et démographiques améliore la résilience d’une population soumise aux coûts physiologiques de la reproduction / Costs of reproduction are pervasive in life history theory. Through this constraint, the reproductive effort of an organism at a given time negatively affects its later survival and fertility. For life historians, they correspond mostly to a physiological trade-off that stems from an allocative process, occurring at each time-step, at the level of the individual. For evolutionary demographers, they are essentially about genetic trade-offs, arising from a genetic variance in a pleiotropic gene acting antagonistically on early-age and late-age fitness components. The study, from an evolutionary demographic standpoint, of these mechanisms and of the relative, cross and joint effects of physiological and genetic costs, is the aim of this thesis. The close examination of Williams (1966)’s original definition of the physiological costs of reproduction led us to produce a theoretical design of their apparatus that accounts for both their mechanistic and evolutionary mechanisms. This design allowed us to make predictions with regards to the strength of costs of reproduction for various positions of organisms on three life-history spectra: slow-fast, income-capital breeders and quality-quantity. From Stearns (1989b)’s tryptic architecture of life history trade-offs –that divides their structure into the genotypic level, the intermediate structure and the phenotypic level – we devised a general framework, which models the possible cohabitation of both physiological and genetic costs. From this, we inferred differing detectability patterns of both types of costs according to the environmental conditions, their variance and individual stochasticity. We could also establish that both costs buffer environmental variations, but with varying time windows of effect. Their dissimilarity emerges also from the differences between mathematical projection models specific to each cost. A new family of evolutionary models is therefore required to implement both physiological and genetic trade-offs. We then describe the vector-based construction method for such a model which we call Multitrait Population Projection Matrix (MPPM) and which allows incorporating both types of costs by embedding them as traits into the matrix. We extend the classical sensitivity analysis techniques of evolutionary demography to MPPMs. Most importantly, we present a new analysis tool for both life history and evolutionary demography: the Trait Level Analysis. It consists in comparing pairs of models that share the same asymptotic properties. Such ergodic equivalent matrices are produced by folding, an operation that consists in reducing the number of traits of a multi-trait model, by averaging transitions for the traits folded upon, whilst still preserving the asymptotic flows. The Trait Level Analysis therefore allows, for example, to measure the evolutionary importance of costs of reproduction by comparing models incorporating them with folded versions of these models from which the costs are absent. Using classical and new methods to compute fitness moments – selection gradient, variance in reproductive success, environmental variance - in models with and without the costs, we can show their effects on various demographic and evolutionary measures. We reveal, in this way, the combined effects of genetic and physiological costs on the vital rates of an age-structured population. We also demonstrate how physiological costs affect both components of effective selection, as they flatten the slope of selection gradients and increase the effective size of a population. Finally, we show how their buffering of environmental and demographic variance confer greater resilience to populations experiencing physiological costs of reproduction

La France présente une grande diversité de conditions pédoclimatiques et par conséquent des systèmes d'élevage très variés. Pourtant, les schémas de sélection actuels en bovins laitiers considèrent que les meilleurs reproducteurs sont les mêmes quel que soit le type de système d'élevage, c'est-à-dire qu'il n'existe pas d'interactions Génotype*Milieu. L'objectif de cette thèse est de mesurer les interactions Génotype*Milieu en France sur les caractères laitiers et fonctionnels dans les 3 races bovines laitières principales (Normande, Montbéliarde, Holstein).Un point crucial dans ce type d'étude est la façon de définir le milieu. L'innovation de cette thèse est l'utilisation des profils Troupeau-Jour de Contrôle. Ce sont des coproduits du modèle génétique basé sur les contrôles élémentaires. Ils reflètent la production permise par la conduite de troupeau au cours du temps. Ils présentent l'avantage d'être disponibles à partir des bases de données nationales et d'être uniquement le reflet de la conduite, contrairement à d'autres définitions qui mêlent effet génétique et effet d'environnement au sein de la définition du milieu ou qui se focalisent sur certains points précis de la conduite sans prendre en compte son effet global. La description des profils Troupeau-Jour de Contrôle de plus de 15000 élevages normands, montbéliards et holsteins par des méthodes de lissage de séries temporelles, d'analyses factorielles et de classification a permis de créer 2 définitions du milieu en vue de l'étude des interactions Génotype*Milieu : des milieux définis comme des groupes d'élevages aux conduites distinctes ou un milieu défini comme un continuum à travers une ou des variables synthétiques.L'importance des interactions Génotype*Milieu a été estimées à partir de 2 types de modèles : un modèle multicaractères qui valorise la définition du milieu sous forme de groupes d'élevages et un modèle de norme de réaction qui valorise, quant à lui, le milieu défini comme un continuum. Les avancées méthodologiques proposées dans cette thèse concernent les modèles de normes de réaction. Des approches permettant de prendre en compte plusieurs variables d'environnement au sein d'un même modèle et de les résumer au sein d'une matrice génétique de rang réduit sont mises en avant.Aucun reclassement n'a été mis en évidence : les meilleurs reproducteurs sont les mêmes quel que soit le système d'élevage. Les schémas des sélections actuels sont donc performants. Il existe tout de même une interaction Génotype*Milieu significative sous forme d'effet d'échelle : la variabilité des valeurs génétiques des animaux est plus importante dans les systèmes d'élevage plus intensifs. Cet effet d'échelle ne sera pas pris en compte dans les modèles d'évaluation génétique, en revanche, il est possible d'imaginer un indicateur utilisable sur le terrain pour mesurer les écarts de performances, dus à cet effet, auxquels il faut s'attendre. / Because of the diversity of pedoclimatic conditions in France, dairy farms have very diversified herd management systems. For this reason, some breeders question the efficiency of the existing breeding schemes for their own management system. To overcome these concerns, a genotype by environment interaction study at the French national level has been considered necessary. The aim of this thesis is to assess the presence of genotype by environment interactions in Normande, Montbéliarde and Holstein breeds for production and functional traits.A tricky point in genotype by environment interaction studies is the environmentdefinition. The innovation of this thesis deals with the use of Herd-Test Day profiles.They are co-products of the French test day model. They reflect the production dueto herd management over time. They are available in national databases and only reflect herd management effect contrary to other definitions in which there is a confusion between genetic and environmental effects in the environment definition or which focus on specific features of the herd management without taking into account its global effect. Herd Test Day profiles of more than 15,000 herds have been studied through time series smoothing, factor analysis and clustering methods. It led to 2 definitions of the environment for the genotype by environment interaction study : environments defined as herd groups or one to several environmental gradients.Genotype by environment interactions were assessed with 2 models : the multitrait model and the reaction norm model. The first one uses herd groups as definition of the environment whereas reaction norm model considers the environment as a gradient. Several methodological improvements have been suggested for reaction norm models : taking into account several environmental gradients in a reaction norm model and summarizing them through a reduced rank genetic matrix.No reranking has been shown : the best parents are the same whatever the herd management system. Consequently, current breeding schemes are relevant. However, a scale effect exists : the variability of animal breeding values is higher in intensive herds. Genetic models will not account for this scale effect. However, a tool useful for breeders could indicate the deviation between expected performances and actual performances due to this scale effect.