Im Rahmen der Item-Response-Theorie evaluiert die kumulative Dissertationsschrift verschiedene Methoden und Modelle zur Identifikation von Kontexteffekten in Large-Scale Assessments. Solche Effekte können etwa in quantitativen empirischen Schulleistungsstudien auftreten und zu verzerrten Item- und Personenparametern führen. Um in Einzelfällen abschätzen zu können, ob Kontexteffekte auftreten und dadurch die Gefahr verzerrter Parameter gegeben ist (und falls ja, in welcher Weise), müssen IRT-Modelle entwickelt werden, die zusätzlich zu Item- und Personeneffekten Kontexteffekte parametrisieren. Solch eine Parametrisierung ist im Rahmen Generalisierter Allgemeiner Linearer Modelle möglich. In der Dissertation werden Positionseffekte als ein Beispiel für Kontexteffekte untersucht, und es werden die statistischen Eigenschaften dieses Messmodells im Rahmen einer Simulationsstudie evaluiert. Hier zeigt sich vor allem die Bedeutung des Testdesigns: Um unverfälschte Parameter zu gewinnen, ist nicht nur ein adäquates Messmodell, sondern ebenso ein adäquates, also ausbalanciertes Testdesign notwendig. Der dritte Beitrag der Dissertation befasst sich mit dem Problem fehlender Werte auf Hintergrundvariablen in Large-Scale Assessments. Als Kontexteffekt wird in diesem Beispiel derjenige Effekt verstanden, der die Wahrscheinlichkeit eines fehlenden Wertes auf einer bestimmten Variablen systematisch beeinflusst. Dabei wurde das Prinzip der multiplen Imputation auf das Problem fehlender Werte auf Hintergrundvariablen übertragen. Anders als bisher praktizierte Ansätze (Dummy-Codierung fehlender Werte) konnten so in einer Simulationsstudie für fast alle Simulationsbedingungen unverfälschte Parameter auf der Personenseite gefunden werden. / The present doctoral thesis evaluates various methods and models of the item response theory to parametrize context effects in large-scale assessments. Such effects may occur in quantitative educational assessments and may cause biased item and person parameter estimates. To decide whether context effects occur in individual cases and lead to biased parameters, specific IRT models have to be developed which parametrize context effects additionally to item and person effects. The present doctoral thesis consists of three single contributions. In the first contribution, a model for the estimation of context effects in an IRT framework is introduced. Item position effects are examined as an example of context effects in the framework of generalized linear mixed models. Using simulation studies, the statistical properties of the model are investigated, which emphasizes the relevance of an appropriate test design. A balanced incomplete test design is necessary not only to obtain valid item parameters in the Rasch model, but to guarantee for unbiased estimation of position effects in more complex IRT models. The third contribution deals with the problem of missing background data in large-scale assessments. The effect which predicts the probability of a missing value on a certain variable, is considered as a context effect. Statistical methods of multiple imputation were brought up to the problem of missing background data in large-scale assessments. In contrast to other approaches used so far in practice (dummy coding of missing values) unbiased population and subpopulation estimates were received in a simulation study for most conditions.

Differenzielles Itemfunktionieren bedeutet für die Trendschätzung durch Linking in querschnittlich angelegten Large-Scale Assessments eine Voraussetzungsverletzung. Solche Voraussetzungsverletzungen können sich negativ auf die Eigenschaften von Trendschätzern auswirken, woraus sich Einschränkungen für die Interpretierbarkeit der Trendschätzung ergeben können. Die vorliegende Arbeit umfasst, eingebettet in einen Rahmungstext, drei Einzelbeiträge, die sich mit der Prüfung der Auswirkung differenziellen Itemfunktionierens unterschiedlicher Provenienz auseinandersetzen. Im ersten Einzelbeitrag wird die Interaktion von Linkingdesigns und Linkingmethoden mit zwischen Ländern und über die Zeit unsystematisch unterschiedlich funktionierenden Items untersucht. Dabei zeigte sich, dass die Wahl des Designs von großer Bedeutung sein kann, während der Performanzunterschied zwischen gängigen Linkingmethoden marginal war. Darüber hinaus führte der häufig praktizierte Ausschluss von differenziell funktionierenden Items zu einem Effizienzverlust. Im zweiten Einzelbeitrag wird die Unsicherheit der Trendschätzung, die entsteht, wenn Items zwischen Ländern und über die Zeit unsystematisch unterschiedlich funktionieren, quantifiziert und in die Berechnung der zugehörigen Standardfehler integriert. Im dritten Einzelbeitrag wird betrachtet, wie differenziellem Itemfunktionieren begegnet werden kann, das durch fehlende Werte und wechselnde Ausfallmechanismen zustande kommt. Wurden die fehlenden Werte inadäquat behandelt, verloren die Trendschätzer ihre Erwartungstreue und Konsistenz sowie an Effizienz. In der Summe wird in der vorliegenden Arbeit identifiziert und hervorgehoben, dass es in den untersuchten Bedingungen je nach Art des differenziellen Itemfunktionierens effektive Möglichkeiten des Umgangs mit diesem gibt, die potenziellen Einschränkungen bei der validen Interpretation der Trendschätzung zumindest teilweise entgegenwirken können. / Differential item functioning signifies a violation of the prerequisites required for trend estimation, which involves the linking of cross-sectional large-scale assessments. Such violations can negatively affect the properties of the trend estimators. Hence, the interpretability of trend estimates will be limited under such circumstances. Embedded within an overarching framework, three individual contributions that examine and deal with the effects of differential item functioning from different origins are presented in the current dissertation. The first article examines the interactions of linking designs and linking methods with items that show unsystematic and differential functioning between countries and across time. It showed that the choice of the design can be of great importance, whereas the difference in performance between common linking methods was marginal. In addition, the exclusion of differentially functioning items, an approach that is frequently used in practice, led to a loss of efficiency. In the second contribution, the uncertainty for the trend estimation resulting from items that show unsystematic and differential functioning between countries and across time is quantified and incorporated into the calculation of the trends' standard errors. The third article focuses on differential item functioning that is induced by missing values and nonresponse mechanisms that change over time. When the missing values were treated inappropriately, the trend estimators lost their unbiasedness, their consistency, and their efficiency. In sum, this dissertation identifies and emphasizes the ideas that, depending on the type of differential item functioning, there are effective ways to deal with it under the investigated conditions, and these can at least partially counteract potential limitations so that the trend estimates can still be interpreted validly.

Die vorliegende Arbeit untersucht die Testteilnahmemotivation von Schülerinnen und Schülern in großangelegten Schulleistungsstudien. Es wurde ein theoretisches Erwartung-Wert-Anstrengung-Modell der Testteilnahmemotivation herausgearbeitet, das in drei empirischen Studien überprüft wurde. Dabei wurde das komplexe Beziehungsgefüge zwischen Erfolgserwartung, dem wahrgenommene Wert des Tests, Anstrengungsbereitschaft und Testleistung untersucht. Datengrundlage der Studie I bildete die erste PISA-Erhebung aus dem Jahr 2000, in der die Testteilnahmemotivation durch Fragen zur Anstrengungsbereitschaft und zum wahrgenommenen Wert des Tests erhoben wurde. In Studie II und III gaben die Jugendlichen, die an der Ländervergleichsstudie im Jahr 2012 teilnahmen, Einschätzungen zu ihrer Erfolgserwartung, dem wahrgenommenen Wert des Tests und ihrer Anstrengungsbereitschaft ab. Die Ergebnisse zeigen, dass Testteilnahmemotivation zur Erklärung individueller Unterschiede in der Testleistung beiträgt (Studie I), auch wenn diverse Hintergrundinformationen der Teilnehmenden berücksichtigt werden (Studie III). Die theoretisch angenommenen Beziehungen im Erwartung-Wert-Anstrengung-Modell wurden fast vollständig bestätigt: Vor allem der wahrgenommene Wert, aber auch die Erwartungen sagten die berichtete Anstrengungsbereitschaft der Teilnehmenden vorher; die Erfolgserwartung und die Anstrengungsbereitschaft wiesen einen Zusammenhang mit der Testleistung auf (Studie II). Im Verlauf eines Leistungstests berichteten die Teilnehmenden im Durchschnitt eine Abnahme in der Anstrengung und dem Wert sowie einen stabilen Verlauf ihrer Erfolgserwartung. Zur Erklärung der Testleistung trug neben der anfänglichen Erfolgserwartung und Anstrengungsbereitschaft auch die Veränderung in der Erfolgserwartung bei (Studie III). Für eine hohe Testleistung ist es wichtig, dass die Teilnehmenden den Test motiviert beginnen und während des Tests selbstsicher bezüglich ihrer Erfolgserwartung bleiben. / The thesis investigates effects of test-taking motivation on test performance in low-stakes assessments. An expectancy-value-effort model of test-taking motivation was developed and tested in three empirical studies. The studies investigated the complex relationship between expectancy for success, perceived value of the test, test-taking effort, and test performance. The database of study I is the first PISA study. Test-taking motivation was assessed with questions about effort and the perceived value of the test. Study II and III are premised on the national assessment study in the year 2012. The students reported their expectancy for success, their perceived value of the test, and their test-taking effort. Overall, the results showed that test-taking motivation explained test performance (study I) although controlling for various students’ background characteristics (e.g., socio-economic background, study III). We found support for nearly all of the theoretically assumed relationships in the expectancy-value-effort model: Expectancy for success and perceived value of the test explained test-taking effort; expectancy for success and test-taking effort had the most pronounced effects on test performance (study II). The students reported, on average, a stable course of expectancy for success over the testing session; perceived importance of the test and test-taking effort slightly decreased within the testing session. The initial expectancy for success and the initial test-taking effort as well as change in expectancy for success explained students’ test performance. Above all, it is crucial that students begin the test with a high level of test-taking motivation and remain confident about a successful test completion to the end of the testing session.

In this thesis, we develop block-decomposition (BD) methods and variants of accelerated *9gradient methods for large-scale conic programming and convex optimization, respectively. The BD methods, discussed in the first two parts of this thesis, are inexact versions of proximal-point methods applied to two-block-structured inclusion problems. The adaptive accelerated methods, presented in the last part of this thesis, can be viewed as new variants of Nesterov's optimal method. In an effort to improve their practical performance, these methods incorporate important speed-up refinements motivated by theoretical iteration-complexity bounds and our observations from extensive numerical experiments. We provide several benchmarks on various important problem classes to demonstrate the efficiency of the proposed methods compared to the most competitive ones proposed earlier in the literature. In the first part of this thesis, we consider exact BD first-order methods for solving conic semidefinite programming (SDP) problems and the more general problem that minimizes the sum of a convex differentiable function with Lipschitz continuous gradient, and two other proper closed convex (possibly, nonsmooth) functions. More specifically, these problems are reformulated as two-block monotone inclusion problems and exact BD methods, namely the ones that solve both proximal subproblems exactly, are used to solve them. In addition to being able to solve standard form conic SDP problems, the latter approach is also able to directly solve specially structured non-standard form conic programming problems without the need to add additional variables and/or constraints to bring them into standard form. Several ingredients are introduced to speed-up the BD methods in their pure form such as: adaptive (aggressive) choices of stepsizes for performing the extragradient step; and dynamic updates of scaled inner products to balance the blocks. Finally, computational results on several classes of SDPs are presented showing that the exact BD methods outperform the three most competitive codes for solving large-scale conic semidefinite programming. In the second part of this thesis, we present an inexact BD first-order method for solving standard form conic SDP problems which avoids computations of exact projections onto the manifold defined by the affine constraints and, as a result, is able to handle extra large-scale SDP instances. In this BD method, while the proximal subproblem corresponding to the first block is solved exactly, the one corresponding to the second block is solved inexactly in order to avoid finding the exact solution of a linear system corresponding to the manifolds consisting of both the primal and dual affine feasibility constraints. Our implementation uses the conjugate gradient method applied to a reduced positive definite dual linear system to obtain inexact solutions of the latter augmented primal-dual linear system. In addition, the inexact BD method incorporates a new dynamic scaling scheme that uses two scaling factors to balance three inclusions comprising the optimality conditions of the conic SDP. Finally, we present computational results showing the efficiency of our method for solving various extra large SDP instances, several of which cannot be solved by other existing methods, including some with at least two million constraints and/or fifty million non-zero coefficients in the affine constraints. In the last part of this thesis, we consider an adaptive accelerated gradient method for a general class of convex optimization problems. More specifically, we present a new accelerated variant of Nesterov's optimal method in which certain acceleration parameters are adaptively (and aggressively) chosen so as to: preserve the theoretical iteration-complexity of the original method; and substantially improve its practical performance in comparison to the other existing variants. Computational results are presented to demonstrate that the proposed adaptive accelerated method performs quite well compared to other variants proposed earlier in the literature.

Memory design is a crucial component of VLSI system design from area, power and performance perspectives. To meet the increasingly challenging system specifications, architecture, circuit and device level innovations are required for existing memory technologies. Emerging memory solutions are widely explored to cater to strict budgets. This thesis presents design methodologies for custom memory design with the objective of power-performance benefits across specific applications. Taking example of STTRAM (spin transfer torque random access memory) as an emerging memory candidate, the design space is explored to find optimal energy design solution. A thorough thermal reliability study is performed to estimate detection reliability challenges and circuit solutions are proposed to ensure reliable operation. Adoption of the application-specific optimal energy solution is shown to yield considerable energy benefits in a read-heavy application called MBC (memory based computing). Circuit level customizations are studied for the volatile SRAM (static random access memory) memory, which will provide improved energy-delay product (EDP) for the same MBC application. Memory design has to be aware of upcoming challenges from not only the application nature but also from the packaging front. Taking 3D die-folding as an example, SRAM performance shift under die-folding is illustrated. Overall the thesis demonstrates how knowledge of the system and packaging can help in achieving power efficient and high performance memory design.

With data volumes increasing at a high rate and the emergence of highly scalable infrastructures (cloud computing, petascale computing), distributed management of data becomes a crucial issue that faces many challenges. This thesis brings several contributions in order to address such challenges. First, it proposes a set of principles for designing highly scalable distributed storage systems that are optimized for heavy data access concurrency. In particular, it highlights the potentially large benefits of using versioning in this context. Second, based on these principles, it introduces a series of distributed data and metadata management algorithms that enable a high throughput under concurrency. Third, it shows how to efficiently implement these algorithms in practice, dealing with key issues such as high-performance parallel transfers, efficient maintainance of distributed data structures, fault tolerance, etc. These results are used to build BlobSeer, an experimental prototype that is used to demonstrate both the theoretical benefits of the approach in synthetic benchmarks, as well as the practical benefits in real-life, applicative scenarios: as a storage backend for MapReduce applications, as a storage backend for deployment and snapshotting of virtual machine images in clouds, as a quality-of-service enabled data storage service for cloud applications. Extensive experimentations on the Grid'5000 testbed show that BlobSeer remains scalable and sustains a high throughput even under heavy access concurrency, outperforming by a large margin several state-of-art approaches.

This thesis consists of four self-contained essays in the sociology of educational stratification. Study I draws on newly collected survey data to assess the biases that arise in estimating socioeconomic differences in achievement when relying on parent and student reported data on social background. The main finding is that student reports on parental occupation overcome both the problem of misreporting that plagues other data collected from children, and the equally damaging problem of selective nonresponse among parents. Conditional estimates of ethnic disparities are relatively unaffected by these issues. Study II deals with student survey reports on the number of books in the home. A prominent string of authors has favoured this variable as a social background proxy over parental occupation or education based on its strong associations with educational outcomes. The paper applies various methods to large-scale student assessment data to show that these associations rest not on higher reliability as commonly assumed, but rather on two types of endogeneity. Low achievers accumulate less books and are also prone to underestimate their number. Study III uses survey and register data to study immigrant parents' education and its associations with children's achievement in recent Swedish cohorts. Two aspects of parental education are distinguished: the absolute years of schooling and a relative place in the source country's educational distribution. Parents' absolute education turns out to predict children's test scores and grades, whereas relative education is a better predictor of their educational aspirations. The result is of some consequence for studies seeking to assess ethnic disparities net of observed parental characteristics. Study IV extends the positional approach of Study III to understand immigrants' self-perceived social status and income satisfaction in European countries. Those higher educated by origin country than host country standards make more dismal assessments of their current situation than do other immigrants in otherwise similar circumstances. This is attributed to a social contrast mechanism and argued to be of relevance in understanding longer-term patterns of social and economic integration, including educational decisions made by the second generation. / Avhandlingen består av fyra fristående studier som alla berör utbildning och social stratifiering. Studie I undersöker med nyinsamlade enkätdata hur sociala skillnader i skolprestation riskerar att felskattas med bakgrundsuppgifter inhämtade från föräldrar respektive elever. Den viktigaste slutsatsen är att elevuppgifter om föräldrars yrke undviker mycket av den felrapportering som behäftar andra elevsvar, liksom det utbredda problemet med selektivt bortfall bland föräldrar. Villkorliga skattningar av etniska skillnader är relativt opåverkade av dessa metodproblem. Studie II granskar elevers uppgifter om antalet böcker i hemmet. En betydande litteratur har förespråkat denna variabel som ett mått på klasstillhörighet framför föräldrars yrke eller utbildning på grundval av starka samband med elevers studieresultat. Uppsatsen tillämpar en rad metoder på data från en internationell kunskapsutvärdering och finner att sambandens styrka inte vilar på högre tillförlitlighet som tidigare förmodats, utan på endogenitetsproblem av två slag. Lågpresterande elever ackumulerar färre böcker och är dessutom benägna att underskatta deras antal. Studie III använder enkät- och registerdata för att belysa utlandsfödda föräldrars utbildning och dess samband med prestationer bland svenska skolbarn. Två aspekter av utbildningsbakgrund särskiljs: föräldrars utbildningsår samt deras relativa placering i ursprungslandets fördelning. Absolut utbildning visar sig predicera elevers testresultat och betyg, medan relativ utbildning är en bättre prediktor för barns aspirationer. Resultatet är av betydelse för studier av etniska skillnader där statistisk kontroll görs för observerbara föräldraegenskaper. Studie IV tillämpar den positionella ansatsen från Studie III för att förstå utlandsföddas självupplevda status och inkomsttillfredsställelse i europeiska länder. Migranter som är mer högutbildade med ursprungslandets mått mätt än värdlandets tenderar att ha en mer negativ bild av sin nuvarande situation än andra i objektivt liknande omständigheter. Detta kan förstås i termer av sociala referensramar och framhålls som relevant i tolkningen av långsiktiga sociala och ekonomiska integrationsmönster, inklusive de utbildningsval som efterföljande generationer gör. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Manuscript. Paper 3: Manuscript. Paper 4: Manuscript.</p>

International Telemetering Conference Proceedings / October 30-November 02, 1995 / Riviera Hotel, Las Vegas, Nevada / The Central Control Unit (CCU) for a decentralised on-board base-band telemetry system is designed for use in launch vehicle missions of the Indian Space Research Organisation (ISRO). This new design is a highly improved and miniaturised version of an earlier design. The major design highlights are as follows: usage of CMOS Field Programmable Gate Array (FPGA) devices in place of LS TTL devices, high level user programmability of TM format using EEPROMs, usage of high density memory for on-board data storage and delayed data transmission, HMC based pre-modulation filter and final output driver etc. The entire system is realised on a single 6 layer MLB and is packaged on a stackable modular frame. This design has resulted in a 1:4 reduction in weight, 1:4 reduction in volume, 1:5 reduction in power consumption and 1:3 reduction in height in addition to drastic reduction of part diversity and solder joints and thus greatly increased reliability. This paper discusses the design approach, implementation details, tools used, simulations carried out and the results of detailed qualification tests done on the realised qualification model.

International Telemetering Conference Proceedings / October 25-28, 1993 / Riviera Hotel and Convention Center, Las Vegas, Nevada / The Hubble Space Telescope (HST) Observatory Management System (HSTOMS), located at the Goddard Space Flight Center (GSFC), provides telemetry, command, analysis and mission planning functions in support of the HST spacecraft. The Telemetry and Command System (TAC) is an aging system that performs National Aeronautics and Space Administration (NASA) Communications (Nascom) block and telemetry processing functions. Future maintainability is of concern because of the criticality of this system element. HSTOMS has embarked on replacing the TAC by using functional elements developed by the Microelectronics Systems Branch of the GSFC. This project, known as the Transportable TAC (TTAC) because of its inherent flexibility, is addressing challenges that have resulted from applying recent technological advances into an existing operational environment. Besides presenting a brief overview of the original TAC and the new TTAC, this paper also describes the challenges faced and the approach to overcoming them.

Le travail présenté ici s'intéresse à la résolution numérique de problèmes de commande optimale stochastique de grande taille. Nous considérons un système dynamique, sur un horizon de temps discret et fini, pouvant être influencé par des bruits exogènes et par des actions prises par le décideur. L'objectif est de contrôler ce système de sorte à minimiser une certaine fonction objectif, qui dépend de l'évolution du système sur tout l'horizon. Nous supposons qu'à chaque instant des observations sont faites sur le système, et éventuellement gardées en mémoire. Il est généralement profitable, pour le décideur, de prendre en compte ces observations dans le choix des actions futures. Ainsi sommes-nous à la recherche de stratégies, ou encore de lois de commandes, plutôt que de simples décisions. Il s'agit de fonctions qui à tout instant et à toute observation possible du système associent une décision à prendre. Ce manuscrit présente trois contributions. La première concerne la convergence de méthodes numériques basées sur des scénarios. Nous comparons l'utilisation de méthodes basées sur les arbres de scénarios aux méthodes particulaires. Les premières ont été largement étudiées au sein de la communauté "Programmation Stochastique". Des développements récents, tant théoriques que numériques, montrent que cette méthodologie est mal adaptée aux problèmes à plusieurs pas de temps. Nous expliquons ici en détails d'où provient ce défaut et montrons qu'il ne peut être attribué à l'usage de scénarios en tant que tel, mais plutôt à la structure d'arbre. En effet, nous montrons sur des exemples numériques comment les méthodes particulaires, plus récemment développées et utilisant également des scénarios, ont un meilleur comportement même avec un grand nombre de pas de temps. La deuxième contribution part du constat que, même à l'aide des méthodes particulaires, nous faisons toujours face à ce qui est couramment appelé, en commande optimale, la malédiction de la dimension. Lorsque la taille de l'état servant à résumer le système est de trop grande taille, on ne sait pas trouver directement, de manière satisfaisante, des stratégies optimales. Pour une classe de systèmes, dits décomposables, nous adaptons des résultats bien connus dans le cadre déterministe, portant sur la décomposition de grands systèmes, au cas stochastique. L'application n'est pas directe et nécessite notamment l'usage d'outils statistiques sophistiqués afin de pouvoir utiliser la variable duale qui, dans le cas qui nous intéresse, est un processus stochastique. Nous proposons un algorithme original appelé Dual Approximate Dynamic Programming (DADP) et étudions sa convergence. Nous appliquons de plus cet algorithme à un problème réaliste de gestion de production électrique sur un horizon pluri-annuel. La troisième contribution de la thèse s'intéresse à une propriété structurelle des problèmes de commande optimale stochastique : la question de la consistance dynamique d'une suite de problèmes de décision au cours du temps. Notre but est d'établir un lien entre la notion de consistance dynamique, que nous définissons de manière informelle dans le dernier chapitre, et le concept de variable d'état, qui est central dans le contexte de la commande optimale. Le travail présenté est original au sens suivant. Nous montrons que, pour une large classe de modèles d'optimisation stochastique n'étant pas a priori consistants dynamiquement, on peut retrouver la consistance dynamique quitte à étendre la structure d'état du système / This work is intended at providing resolution methods for Stochastic Optimal Control (SOC) problems. We consider a dynamical system on a discrete and finite horizon, which is influenced by exogenous noises and actions of a decision maker. The aim is to minimize a given function of the behaviour of the system over the whole time horizon. We suppose that, at every instant, the decision maker is able to make observations on the system and even to keep some in memory. Since it is generally profitable to take these observations into account in order to draw further actions, we aim at designing decision rules rather than simple decisions. Such rules map to every instant and every possible observation of the system a decision to make. The present manuscript presents three main contributions. The first is concerned with the study of scenario-based solving methods for SOC problems. We compare the use of the so-called scenario trees technique to the particle method. The first one has been widely studied among the Stochastic Programming community and has been somehow popular in applications, until recent developments showed numerically as well as theoretically that this methodology behaved poorly when the number of time steps of the problem grows. We here explain this fact in details and show that this negative feature is not to be attributed to the scenario setting, but rather to the use of a tree structure. Indeed, we show on numerical examples how the particle method, which is a newly developed variational technique also based on scenarios, behaves in a better way even when dealing with a large number of time steps. The second contribution starts from the observation that, even with particle methods, we are still facing some kind of curse of dimensionality. In other words, decision rules intrisically suffer from the dimension of their domain, that is observations (or state in the Dynamic Programming framework). For a certain class of systems, namely decomposable systems, we adapt results concerning the decomposition of large-scale systems which are well known in the deterministic case to the SOC case. The application is not straightforward and requires some statistical analysis for the dual variable, which is in our context a stochastic process. We propose an original algorithm called Dual Approximate Dynamic Programming (DADP) and study its convergence. We also apply DADP to a real-life power management problem. The third contribution is concerned with a rather structural property for SOC problems: the question of dynamic consistency for a sequence of decision making problems over time. Our aim is to establish a link between the notion of time consistency, that we loosely define in the last chapter, and the central concept of state structure within optimal control. This contribution is original in the following sense. Many works in the literature aim at finding optimization models which somehow preserve the "natural" time consistency property for the sequence of decision making problems. On the contrary, we show for a broad class of SOC problems which are not a priori time-consistent that it is possible to regain this property by simply extending the state structure of the model