Influence de l'administration des preuves sur l'activité mentale menant à la formation du verdict au cours d'un procès pénal

Abstract : The role of order effects has been widely shown and discussed in areas such as memory and social impression formation. This work focuses in a first half on order effects influencing the verdict chosen at the end of a criminal trial. Contrary to impression formation but according to trial's characteristics, it has been hypothesised that a recency effect would influence the verdict's choice. Three groups of students (N = 576) received a mock trial resume with a specific order stemming from the combination of three witnesses, one expert and two ocular witnesses. Results show a recency effect, the last testimony provoking significantly more acquittals if discriminating, and more condemnations if incriminating. The second half of this work starts from Gestalt and sociopsychological researches presenting numerous insights into cognitive organization of perceptions and opinions. It has been postulated that a witness probative value will change according to the emitted verdict, an incriminating witness or expert possessing a higher probative value in a condemning verdict than in an innocenting one, on the other hand a discriminating witness or expert having a higher probative value in an acquittal than in a condemnation. Results using a seven points scale measuring witnesses' probative value confirm this hypothesis. Argumentations written by the subjects to explain their verdict and refering to the accusing expert also show a congruency effect as categories of arguments are identical in case of condemnation or acquittal, the only difference between both types of verdicts residing in the frequency of these categories following the judgement, higher use of incriminating arguments in case of guiltiness and higher use of discriminating ones if the accused is found innocent. Résumé : L'intervention des effets sériels a fait l'objet de nombreuses recherches dans le domaine de la mémoire et de la formation d'impression en psychologie sociale. Ce travail s'intéresse dans une première partie aux effets d'ordre pouvant influencer le choix du verdict à la fin d'un procès pénal. Contrairement aux résultats obtenus en matière de formation d'impression, mais conformément aux caractéristiques d'un procès, l'hypothèse de l'intervention d'un effet de récence a été formulée, affirmant que les derniers témoins influencent le plus le choix du verdict. Trois groupes d'étudiants (N = 576) ont lu le résumé d'un procès fictif présentant trois témoignages, deux témoins visuels à décharge et un expert à charge. Chaque groupe recevait un ordre spécifique de présentation des témoins de sorte que l'expert se trouvait en première, deuxième ou troisième position. Les résultats montrent un effet de récence, le dernier témoin provoquant davantage d'acquittements s'il est disculpant et davantage de condamnations s'il est inculpant. La seconde partie de ce travail émane des recherches effectuées dans le domaine de la théorie de la forme et de la psychologie sociale ayant un intérêt marqué pour l'organisation cognitive de nos perceptions et de nos opinions. L'hypothèse que nous avons posée souligne le lien entretenu entre l'évaluation de la force probante d'un témoin et le verdict émis : un témoignage discriminant possédera plus de poids en cas d'acquittement qu'en cas de condamnation, inversement un témoignage incriminant aura plus de poids en cas de condamnation qu'en cas d'acquittement. L'utilisation d'une échelle en sept points mesurant la force probante des deux types de témoins confirme cette hypothèse, l'estimation de la valeur accordée à un même témoin variant selon le type de verdict choisi. Les argumentations de chaque verdict ont également montré que les catégories d'arguments se référant à l'expert étaient identiques en cas de condamnation ou d'acquittement de l'inculpé, par contre les fréquences de ces catégories entretiennent un lien congruent avec le verdict, celles inculpantes étant majoritairement utilisées pour asseoir la culpabilité mais peu représentées en cas d'acquittement, inversement celles disculpantes apparaissant bien plus lorsqu'il s'agit d'innocenter l'inculpé que pour le condamner.

Pouvoir discrétionnaire et redevabilité de la bureaucratie de guichet : les taxateurs d'une caisse de chômage comme acteurs de mise en oeuvre

Résumé en français La thèse de doctorat porte, de manière générale, sur le rôle des acteurs administratifs dans la mise en oeuvre de l'action publique. En particulier, dans le contexte de l'implémentation de la politique suisse d'assurance-chômage, la recherche se focalise sur l'étude de la bureaucratie de guichet, à savoir un type spécifique d'administrations et d'agents publics situés aux premières lignes de l'action étatique et qui sont en contact direct et quotidien avec les usagers. La thèse a pour objectif principal d'obtenir une compréhension détaillée des impacts concrets dont sont porteurs deux types de réformes majeures du secteur public : les réformes de Nouvelle Gestion Publique (NGP) et les nouvelles technologies informatiques. Le questionnement central de la recherche consiste en une étude approfondie des effets de ces réformes sur les bureaucraties de guichet, et en particulier sur deux aspects centraux les concernant : d'une part, le niveau de pouvoir discrétionnaire que les agents publics de base disposent dans l'application de la loi fédérale sur l'assurance-chômage, autrement dit leur marge de manoeuvre ; d'autre part, les manières au travers desquelles ces mêmes agents sont appelés à rendre des comptes quant à leurs actions et leurs décisions, à savoir l'enjeu plus large de la redevabilité publique des acteurs administratifs. Ces enjeux ont été analysés au niveau empirique dans le contexte organisationnel d'une caisse publique cantonale de chômage ayant expérimenté les réformes évoquées ci-dessus. L'organisation choisie a été investiguée au travers d'une étude de cas ethnographique approfondie (observation directe du travail quotidien des agents, entretiens semi-directifs, analyse de documents) pendant une période de six mois environ entre 2008 et 2009. L'analyse empirique fournit quatre résultats : a) de manière générale, les taxateurs de la caisse de chômage disposent d'un faible niveau de pouvoir discrétionnaire ; b) le degré de pouvoir discrétionnaire varie selon le type de tâche ; c) les agents sur le terrain rendent des comptes auprès d'une multiplicité d'acteurs, sur une variété d'aspects de leur travail et au travers de différents mécanismes de contrôle ; d) les outils de NGP et les nouvelles technologies informatiques ont peu d'impact sur l'étendue du pouvoir discrétionnaire des agents mais contribuent à influencer le type de redevabilité publique pratiquée à ce niveau. Summary in English This PhD dissertation deals with the role of public administration in policy implementation. In the Swiss context of unemployment insurance policy, it focuses on street-level bureaucracy, a specific type of public organisations and agents located at the frontline of public action, that is to say low-level civil servants who are in direct, daily and face-to-face contact with citizens. The dissertation aims at a deep understanding of what are the concrete impacts of two main important changes touching public sector organizations : New Public Management reforms (NPM) and Information and Communication technologies (ICT). The main research question consists in assessing the impacts of those reforms on two central issues regarding street-level bureaucrats : on the one hand, the effective degree of discretion frontline agents do have in implementing the federal law on unemployment insurance ; on the other hand, the ways through which these bureaucrats are held accountable about their action and decisions, i.e. accountability regimes at the street-level. These issues have been empirically addressed in the organisational context of a cantonal Unemployment Insurance Funds having experienced the above mentioned reforms. The organisation has been investigated through an in-depth ethnographic case-study (direct observation of daily work, semi-structured interwiews, documentary analysis) in 2008 and 2009 for approximately six months. The empirical analysis indicates that a) in general, street-level agents do exert low degree of policy discretion; b) the level of discretion is variable from one specific task to another ; c) frontline workers are held accountable to many actors, on various aspects of their work and through different mechanisms of control ; d) NPM and ICT instruments have few impact on the issue of policy discretion but more on the type of street-level accountability which is concretely practised at this level.

Intraocular penetration of penciclovir after oral administration of famciclovir: a population pharmacokinetic model.

OBJECTIVES: We developed a population model that describes the ocular penetration and pharmacokinetics of penciclovir in human aqueous humour and plasma after oral administration of famciclovir. METHODS: Fifty-three patients undergoing cataract surgery received a single oral dose of 500 mg of famciclovir prior to surgery. Concentrations of penciclovir in both plasma and aqueous humour were measured by HPLC with fluorescence detection. Concentrations in plasma and aqueous humour were fitted using a two-compartment model (NONMEM software). Inter-individual and intra-individual variabilities were quantified and the influence of demographics and physiopathological and environmental variables on penciclovir pharmacokinetics was explored. RESULTS: Drug concentrations were fitted using a two-compartment, open model with first-order transfer rates between plasma and aqueous humour compartments. Among tested covariates, creatinine clearance, co-intake of angiotensin-converting enzyme inhibitors and body weight significantly influenced penciclovir pharmacokinetics. Plasma clearance was 22.8 ± 9.1 L/h and clearance from the aqueous humour was 8.2 × 10(-5) L/h. AUCs were 25.4 ± 10.2 and 6.6 ± 1.8 μg · h/mL in plasma and aqueous humour, respectively, yielding a penetration ratio of 0.28 ± 0.06. Simulated concentrations in the aqueous humour after administration of 500 mg of famciclovir three times daily were in the range of values required for 50% growth inhibition of non-resistant strains of the herpes zoster virus family. CONCLUSIONS: Plasma and aqueous penciclovir concentrations showed significant variability that could only be partially explained by renal function, body weight and comedication. Concentrations in the aqueous humour were much lower than in plasma, suggesting that factors in the blood-aqueous humour barrier might prevent its ocular penetration or that redistribution occurs in other ocular compartments.

Effects of prolonged administration of the angiotensin converting enzyme inhibitor CGS 16617 in normotensive volunteers.

A new, orally active angiotensin converting enzyme (ACE) inhibitor, CGS 16617, has been evaluated in normotensive subjects during acute and prolonged administration. Single ascending doses of CGS 16617 20 to 100 mg were given to 9 normotensive volunteers at one week intervals and the changes in blood pressure, plasma ACE and renin activity were examined up to 72 h after drug intake. Also, CGS 16617 50 mg/day or placebo were given for 30 days to 8 and 6 normotensive subjects, respectively, maintained on an unrestricted salt diet. Blood pressure was measured daily in the office and ambulatory blood pressure profiles were also obtained before, during and after therapy, using the Remler M 2000 blood pressure recording system. CGS 16617 was an effective and long lasting ACE inhibitor. It did not induce a consistent change in blood pressure, but, the individual responses were very variable and several subjects experienced a clear decrease in the average of the blood pressures recorded during the daytime.

Home care: a safe and attractive alternative to inpatient administration of intensive chemotherapies.

OBJECTIVE: The objective of this study was to evaluate feasibility, safety, perception, and costs of home care for the administration of intensive chemotherapies. METHODS: Patients receiving sequential chemotherapy in an inpatient setting, living within 30 km of the hospital, and having a relative to care for them were offered home care treatment. Chemotherapy was administered by a portable, programmable pump via an implantable catheter. The main endpoints were safety, patient's quality of life [Functional Living Index-Cancer (FLIC)], satisfaction of patients and relatives, and costs. RESULTS: Two hundred days of home care were analysed, representing a total of 46 treatment cycles of intensive chemotherapy in 17 patients. Two cycles were complicated by technical problems that required hospitalisation for a total of 5 days. Three major medical complications (heart failure, angina pectoris, and major allergic reaction) could be managed at home. Grades 1 and 2 nausea and vomiting occurring in 36% of patients could be treated at home. FLIC scores remained constant throughout the study. All patients rated home care as very satisfactory or satisfactory. Patient benefits of home care included increased comfort and freedom. Relatives acknowledged better tolerance and less asthenia of the patient. Home care resulted in a 53% cost benefit compared to hospital treatment (420 ± 120/day vs. 896 ± 165/day). CONCLUSION: Administration of intensive chemotherapy regimens at home was feasible and safe. Quality of life was not affected; satisfaction of patients and relatives was very high. A psychosocial benefit was observed for patients and relatives. Furthermore, a cost-benefit of home care compared to hospital treatment was demonstrated.