Comment évaluer le rôle que jouent les spin-offs académiques - ces entreprises créées pour valoriser les résultats de la recherche publique ? Pour répondre à cette question, cette thèse envisage les spin-offs dans leurs interactions avec les acteurs hétérogènes qui composent les réseaux techno-économiques de l'innovation. Elle décrit l'entrepreneuriat académique comme un processus d'exploration et de consolidation de nouvelles technologies, marchés et relations. De plus, la thèse examine ce que produisent les collaborations inter-organisationnelles qui se nouent au cours de ce processus. Elle montre notamment la diversité des effets générés, en distinguant la contribution de différents types de partenariats (d'exploration, d'exploitation et ambidextres) à différentes dimensions de l'innovation. Enfin, la thèse étudie la manière dont la valeur de ces associations exploratoires est construite dans la pratique d'une action collective qui engage les spin-offs et leurs partenaires et qui mobilise des dispositifs de " valuation " particuliers.

Understanding the wide range of outcomes achieved by firms trying to implement TQM and similar process improvement initiatives presents a challenge to management science and organization theory: a few firms reap sustained benefits from their programs, but most efforts fail and are abandoned. A defining feature of such techniques is the reliance on the front-line workforce to do the work of improvement, thus creating the possibility of agency problems; different incentives facing managers and workers. Specifically, successfully improving productivity can lead to lay-offs. The literature provides two opposing theories of how agency interacts with the ability of quality-oriented improvement techniques to dramaticlly increase productivity. The 'Drive Out Fear' school argues that firms must commit to job security, while the 'Drive In Fear' school emphasizes the positive role that insecurity plays in motivating change. In this study a contract theoretic model is developed to analyze the role of agency in process improvement. The main insight of the study is that there are two types of job security, internal and external, that have opposite impacts on the firm's abilty to implement improvement initiatives. The distinction is useful in explaining the results of different case studies and can reconcile the two change theories. / National Science Foundation, grant SBR-9422228, the Ford Motor Company and the Harley-Davidson Motor Company

In this thesis, I combined field, greenhouse and common-garden experiments to examine the ecological and evolutionary consequences of plant-herbivore interactions and the genetic architecture of fitness-related traits in the insect-pollinated, self-incompatible, perennial herb Arabidopsis lyrata. More specifically, I examined (1) whether damage to leaves and inflorescences affects plant fitness non-additively, (2) whether trichome production is associated with a cost in terms of reduced tolerance to leaf and inflorescence damage, (3) whether young plant resistance to a specialist insect herbivore varies among populations, and (4) whether the evolution of flowering time, floral display and rosette size is constrained by lack of genetic variation or by genetic correlations among traits. A two-year field experiment in a Swedish population showed that damage to rosette leaves and to inflorescences can affect both current and future plant performance of A. lyrata, and that effects on some fitness components are non-additive. A two-year field experiment in another Swedish population indicated that trichome-producing plants are not less tolerant than glabrous plants to leaf and inflorescence damage. In a greenhouse experiment, acceptability of young plants (5-6 weeks old) to ovipositing females and damage received by Plutella xylostella larvae varied considerably among twelve A. lyrata populations. Both oviposition and leaf damage were positively correlated with rosette size, but trichome density in the trichome-producing morph was apparently too low at this developmental stage to influence resistance to P. xylostella. In a common-garden experiment, flowering time, floral display and rosette size varied among four Scandinavian A. lyrata populations, and displayed significant additive genetic variation in some populations. Yet, strong genetic correlations between flowering start and number of flowers, and between petal length and petal width suggest that these traits may not evolve independently. Taken together, the results indicate the need to consider possible long-term and non-additive effects of herbivore damage to different plant parts, that there is no trade-off between trichome production and tolerance to herbivory, that the importance of morphological defenses against herbivory may change through plant ontogeny, and that considerable genetic variation for traits such as flowering time and floral display can be maintained in natural plant populations.

With the increased use of Electronic Control Units (ECUs) in the automotive industry, the system creation and integration becomes progressively more complex. In order to manage this issue, the AUTomotive Open System Architecture (AUTOSAR) was initiated by several of the larger automotive companies and suppliers. Their goal was to create an architecture which will increase the reusability of software, integration of solutions from multiple suppliers and improve scalability. This thesis is made in collaboration with the company CrossControl and covers questions which are specific to their interests, features the AUTOSAR standard has to offer, evaluation of the use of development tools from different vendors, how migration is supported and how the standard affects the required hardware. Among our conclusions is that the AUTOSAR goals of decoupling software application components from the hardware is met to a high degree. Secondly even though file formats are standardized it is not a seamless integration between development tools. Finally, the choice of hardware is not only affected by properties of the standard, but the current versions of tools also limit the choices.

Borderline Personality Disorder (BPD) has been previously conceptualized as an extreme variant of normal personality traits, captured by continuous indices. A previous study successfully developed and validated a self-report BPD measure, the Minnesota Borderline Personality Disorder Scale (MBPD). I conducted two studies aimed at providing further validation for this measure. Results from Study 1 (clinical sample of substance users) indicated that MBPD exhibited strong positive correlations with measures of convergent validity (self-report and diagnostic measures). Additionally, the MBPD showed similar correlations with external correlates as those of the convergent validity measures, in addition to incremental utility in predicting these external correlates above and beyond negative affect. Third, a Receiver Operating Characteristic (ROC) curve analysis indicated that diagnostic accuracy of the MBPD was excellent for differentiation between BPD and non-BPD individuals. Likewise, Study 2 (non-clinical sample of undergraduate students followed over 6 months) showed strong correlations with an index of convergent validity (self-report measure), similar correlations with external correlates as that of the convergent validity index, and incremental predictive utility. Finally, in this study, the MBPD exhibited high rank-order stability, but significant mean-level and individual-level change over time. These data suggests that these scales are measuring the same latent construct of BPD, providing further evidence for the construct validity of the MBPD.

This thesis investigates the impact of avian malaria (Haemosporidia) parasites on the outcome of interspecific competition between two closely related bird species, pied (Ficedula hypoleuca) and collared (F. albicollis) flycatchers. I further investigated how variation in timing of breeding, life history strategies and immune competence genes (MHC genes) modulate the fitness effects of malaria parasites in one of the two species i.e. collared flycatchers. Collared flycatchers colonized the Baltic island Öland in the late 1950-ties and has since then been expanding their breeding range while competitively excluding pied flycatchers from the favourable habitats (deciduous forests). I investigated the underlying mechanisms behind this exclusion by combining detailed long-term breeding data with modern molecular genetic techniques identifying both the presence/absence and lineage specificity of haemosporidian blood parasites. I found that the rapid decline of pied flycatchers can be explained by the combined effects of competition over nestling sites, hybridization and haemosporidian infections. Haemosporidian infections have a negative impact on survival of pied flycatcher females but no detectable effect on collared flycatchers’ longevity or reproductive success. This may be due to the fact that collared flycatchers carry (and are potentially exposed to) a higher diversity of parasites than pied flycatchers, which in turn may select for a higher diversity of MHC genes and hence a better overall protection from the negative impact of parasites. Indeed, functional MHC diversity correlates negatively with malaria prevalence among collared flycatchers from Gotland. Moreover, I found that both, malaria infection intensity and immunoglobulin level influences how infected collared flycatchers respond to increased nestling food-demands. The latter results mean that there is variation in allocation strategies (i.e. in resource allocation between reproductive effort and immune competence) within the collared flycatcher population. Hence, this population has the ability to respond to novel selection pressures in terms of optimal allocation of resources into immune functions. In summary, my results show that local parasites may facilitate the expansion of a new colonizer. This is important in the context of global climate change that will probably increase the colonization rate of southern species and lead to novel host-parasite interactions.

Electricity generation is a vital element of economic growth, and it is necessary to encourage a growth model that does not endanger the capacity of a country to generate electricity. Generating electricity entails costs. This cost is not only economical but can also be, for example, environmental. This implies that there are different trade-offs associated with choices about how to generate electricity, such as technologies, fuels, impact on the environment, construction costs, budget constraints and so on. The Federal Government owns Mexico’s electricity sector. As such not only does it write the rules of the electricity sector but it also executes these rules. The government has stated a series of guiding principles regulating the electricity sector. These guiding principles reflect the priorities that should be taken into account when designing electricity portfolios. My thesis uses financial tools to offer a new approach to the problem of developing electricity portfolios. I assume that the electricity generation mix can be seen as a portfolio of assets. Using portfolio management techniques, I demonstrate scenarios for efficient portfolios given key assumptions about generation choices and prevailing costs. I also illustrate the implications of prioritizing one guiding principle over the other in terms of portfolio cost. Finally, my use of a portfolio modeling approach highlights the complexities inherent in public policy making given the technical and cost-driven nature of the electric power businesses and value chains. My work provides a possible method for more productive evaluation of various approaches in light of mixed priorities and the broad diversity of stakeholders in Mexico. / text

Social insect colonies are complex systems with emergent properties that arise from the cooperation and interaction amongst individuals within colonies. By dividing reproduction and physical labor amongst them, individuals contribute to the growth and ecological success of their colonies, a success that is greater than individuals could achieve on their own. A key characteristic of social insects is a division of labor amongst workers that is determined primarily either by age, morphology, or dominance. Social insects are considered one of the most ecologically successful groups of organisms on earth. Colony life cycles include: 1) growth, in which workers are produced, 2) reproduction, in which queens and males with reproductive capabilities are produced, and 3) senescence. In life history theory, phenotypic plasticity (i.e. a change in phenotype in response to an environmental change), allows organisms to adjust and optimize fitness in response the change in environments. Central to life history theory is the idea that traits have costs and benefits. Using an energetics framework that considers the costs and benefits of traits contributes to our understanding as to why organisms exhibit the sets of traits that they have within their ecological environments. Using the annual bumble bee Bombus impatiens, my dissertation investigates the effects of resource availability on worker production and on the relative allocation of energy towards growth and reproduction within colonies. Bumble bees have a morphological division of labor and concomitantly, they show large intra-colony size variation amongst workers. Because body size is an important life history trait, I also examined the costs and benefits of producing various sized workers. Lastly, I examined the association among worker body size, metabolic rate (a measure of maintenance costs), and lifespan.

This research examines the exit of a subsidiary from its corporate parent through spin-off, the actions taken by the firm management post spin-off, and the performance implications of those actions, all from the spin-off's perspective. While spin-off announcements are generally met with a positive stock market reaction, what occurs post spin-off remains largely unexamined, with performance predictions regarding spin-off firms often being equivocal. This raises questions as to what generates positive performance for spin-off firms, with agency, transaction cost, and upper echelons theories offering differing, and sometimes conflicting, predictions. By integrating these theoretical perspectives, a model of managerial action and its performance implications is presented. The model examines how the formation of new top management, the establishment of managerial monitoring and incentives, and the severance effects from leaving the corporate structure affect strategic, financial, and institutional actions, and how these actions affect performance. The theory and hypotheses developed in this research are empirically tested on a sample of 176 corporate spin-offs completed by publicly traded firms between 1986 and 1997. Results for the action-based models indicate that background of the CEO or the TMT, as well as CEO options, had no effect on actions. CEO and TMT ownership had opposite effects on financial actions, with TMT ownership increasing the likelihood of strategic actions and CEO ownership increasing the likelihood of institutional actions. Ownership by the parent firm and monitoring by officers of the parent serving as board members had no effect on the likelihood of actions, although having a chairman of the board from the parent decreased the likelihood of strategic actions. Finally, severance effects had limited influence on the actions taken post spin-off. Results for the performance-based models indicate that strategic actions were negatively related to ROA, while financial and institutional actions are positively related to ROA and institutional actions are positively related to market performance. In general, inaction was related to lower Tobin's q, with the signs of the coefficients for the other performance models negative, but not significant. Finally, the spin-off firm's relationship with its corporate parent had limited effect on the link between actions and performance.

Field-programmable gate arrays (FPGAs) have enjoyed increasing use due to their low non-recurring engineering (NRE) costs and their straightforward implementation process. However, it is recognized that they have higher per unit costs, poorer performance and increased power consumption compared to custom alternatives, such as application specific integrated circuits (ASICs). This thesis investigates the extent of this gap and it examines the trade-offs that can be made to narrow it. The gap between 90 nm FPGAs and ASICs was measured for many benchmark circuits. For circuits that only make use of general-purpose combinational logic and flipflops, the FPGA-based implementation requires 35 times more area on average than an equivalent ASIC. Modern FPGAs also contain "hard" specific-purpose circuits such as multipliers and memories and these blocks are found to narrow the average gap to 18 for our benchmarks or, potentially, as low as 4.7 when the hard blocks are heavily used. The FPGA was found to be on average between 3.4 and 4.6 times slower than an ASIC and this gap was not influenced significantly by hard memories and multipliers. The dynamic power consumption is approximately 14 times greater on average on the FPGA than on the ASIC but hard blocks showed promise for reducing this gap. This is one of the most comprehensive analyses of the gap performed to date. The thesis then focuses on exploring the area and delay trade-offs possible through architecture, circuit structure and transistor sizing. These trade-offs can be used to selectively narrow the FPGA to ASIC gap but past explorations have been limited in their scope as transistor sizing was typically performed manually. To address this issue, an automated transistor sizing tool for FPGAs was developed. For a range of FPGA architectures, this tool can produce designs optimized for various design objectives and the quality of these designs is comparable to past manual designs. With this tool, the trade-off possibilities of varying both architecture and transistor-sizing were explored and it was found that there is a wide range of useful trade-offs between area and delay. This range of 2.1 X in delay and 2.0 X in area is larger than was observed in past pure architecture studies. It was found that lookup table (LUT) size was the most useful architectural parameter for enabling these trade-offs.