Made available in DSpace on 2013-08-07T18:42:16Z (GMT). No. of bitstreams: 1 000418921-Texto+Completo-0.pdf: 1274519 bytes, checksum: 8981beef3de1308b9b81b49d455f1623 (MD5) Previous issue date: 2009 / An efficient distribution of tasks in computational grids has major challenges that have to be overcomed by a scheduling algorithm. These challenges are related to the characteristics of the environment, such, scalability, heterogeneity, dynamicity, among others. Most of the existing scheduling algorithms use information obtained from the environment or the replication of tasks in order to achieve a satisfactory performance in the execution of applications. Algorithms that rely on resource information to accomplish their task are restricted to environments in which such information is available and reliable all across the grid. On the other hand, algorithms that use replication could be used in any environment but they introduce a significant overhead to the scheduling. However, although reliable information is usually not available across the Grid because of the dynamicity of the resources, it’s possible to have updated resource information inside a local site. This motivates the development of algorithms that use a hybrid technique. In this case the information is used to perform the scheduling when available - for example in the local site - and only when it’s not available the scheduler uses replication, thus reducing its overhead. This work proposes such an hybrid algorithm called TMRorR - Task More Reliable or Replicate for the Ourgrid middleware. The algorithm was simulated with GridSim and implemented in OurGrid to perform the experiments. Obtained results showed a better performance of TMRorR in comparison to the Ourgrid scheduler in several scenarios and a reduction in the overhead of creating replica. / A distribuição de tarefas de forma eficiente em grades computacionais possui grandes desafios que devem ser considerados por um algoritmo de escalonamento. Esse desafios estão relacionados com as características das grades, como a escalabilidade, heterogeneidade, dinamicidade, entre outros. Dentre as diversas propostas de algoritmos de escalonamento existentes, grande parte utilizam da informação obtida do ambiente ou da replicação de tarefas como forma de atingir um desempenho satisfatório na execução das aplicações dentro da grade. Os algoritmos de escalonamento que dependem somente das informações para realizar o escalonamento estão restritos a ambientes em que as mesmas estejam disponíveis e sejam confiáveis. Já os algoritmos que utilizam somente de replicação não dependem de nenhuma informação. A existência de ambientes em que a presença de informação não é totalmente confiável ou não atende a todos os recursos da grade, motiva o surgimento de algoritmos que utilizam uma técnica híbrida. Neste caso a informação, quando disponível, é utilizada para realizar o escalonamento, porém, quando não, utiliza-se da replicação de tarefas. Este trabalho realiza um estudo acerca do escalonamento de tarefas realizado pelo WQR do OurGrid e apresenta um novo algoritmo para escalonamento de aplicações Bag-of-Tasks aqui denominado TMRorR (Task for More Reliable or Replicate).Este algoritmo utiliza uma técnica híbrida e, considera sempre que um recurso que possui informação é sempre mais confiável do um que não possui e, também, faz uma distinção entre recursos pertencentes ao domímio local e recursos pertencentes a outros domínios, tendo regras de escalonamento com algumas diferenças para ambos os casos. O algoritmo foi simulado utilizando o GridSim e implementado no OurGrid para realizar os experimentos. Comparando os resultados obtidos, eles mostraram um desepenho melhor do TMRorR em relação WQR na maioria dos casos, como também, uma redução no desperdício de ciclos de CPU com a realização de um controle maior na criação de réplicas.

Rorschach records of children tend to contain a high percentage of percepts which pertain to animals and animal details. Since teachers of primary grades deal with many animal pictures and stories in their daily work routine, it may be possible that the records of these persons may show a higher percentage of animal responses than might ordinarily be expected of intelligent adults in the normal population.

No description available.

No description available.

High stakes testing in reading demands that educators are providing the appropriate instruction to ensure student’s success on state and national assessments. Causal comparative research was conducted to examine the research questions. 6 reading assessments were used for the study: (a) the 2007 administration of the fourth and eighth grade NAEP, (b) the 2006-2007 administration of the fourth and eighth grade MCT, and (c) the 2007-2008 administration of the fourth and eighth grade MCT2. Data were drawn from the Mississippi Department of Education’s website and from the National Assessment of Education Progress’ website. District level data were available for both Mississippi Curriculum Tests and state level data were available for the NAEP. Results revealed that there were statistically significant differences between achievement levels for fourth and eighth grade students on the MCT and MCT2. Comparison of the means for the two reading tests at the fourth and eighth grade level indicated that students scored statistically significant lower on the MCT2 than they did on the MCT. Results also revealed that there were higher percentages of fourth grade students scoring minimal on NAEP than on the MCT and MCT2. In the basic category, there were higher percentages of students scoring basic on the MCT2 and higher percentages of students scoring proficient and advanced on the MCT. Relying on the targeted percentage of students at the achievement levels of proficient and advanced, the percentages of students on the MCT and MCT2 exceeded NAEP levels. At the eighth grade level, there were higher percentages of students scoring minimal and basic on NAEP and higher percentages of students scoring proficient and advanced on the MCT. Relying on the targeted percentage of students at the achievement levels of proficient and advanced, the percentages of students on the MCT and MCT2 exceeded NAEP levels. Based on the above findings, the MCT2 is more aligned to NAEP, given that there were statistically significant differences between the MCT and MCT2. Further studies are needed in other subject areas to ensure state tests alignment with NAEP

While many clinical practice guidelines that use the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach have ample published evidence to support their recommendations, certain situations (e.g. rare, rapidly emerging, or understudied diseases) may be limiting. As a result, it is common for guideline developers to seek out indirect evidence from other areas to fill in these evidence gaps. The GRADE evidence-to-decision (EtD) framework, which offers a structured and transparent development process for guidelines, includes additional research evidence domains (e.g. feasibility, acceptability, equity) for panels to consider in their decision-making process. This confluence of both considerations of indirect information and increasing literature domain sources to draw from when making decisions weighs on the simplicity of literature presentation. Herein, firstly, we described an example of specific decision-rules for including indirect evidence and the implications of the rules for presenting results to decision-makers. Secondly, we provided a comprehensive overview of how guideline developers currently report economic information across GRADE evidence-to-decision frameworks. Lastly, we ranked the most important study characteristics suggested in the literature by economists to consider as decision-rules when assessing indirectness (transferability) of economic evaluations chosen as research evidence in a GRADE guideline. We conclude that developers, with the help of their panels, should work to establish and report clear decision-rules and the rationale for indirect evidence that they select for their clinical practice guidelines. This has the potential to simplify the presentation of indirect evidence for panels and developers, as well as, to reduce decision-making confusion, time demands and guideline funder costs. / Thesis / Candidate in Philosophy / While many clinical practice guidelines have ample published evidence to support their recommendations, certain situations (e.g. rare, rapidly emerging, or understudied diseases) may be limiting the available literature. As a result, it is common for guideline developers to seek out indirect evidence from other, but related areas, to fill in these evidence gaps. Selection of available indirect evidence may be better than basing decisions on no evidence, in particular in situations of clinical equipoise. However, including all potentially relevant indirect evidence may represent an overuse of evidence. Indirect evidence refers to information sources with related populations, interventions, outcomes or comparisons, which could reasonably be extrapolated; but, are not entirely specific to the research topic at hand. This confluence of both indirect information considerations and desire by many for increasing literature sources to draw from weighs on the simplicity of an overall summary of literature presented during a guideline recommendation decision-making process. Herein, firstly, we described an example of explicit decision-rules for including indirect evidence that were specific and the implications of the rules for presenting results to decision-makers. Secondly, we provided a comprehensive overview of how guideline developers currently report economic information across guideline frameworks, in particular with respect to indirectness. Lastly, we described the most important study characteristics suggested by economists to consider as decision-rules when assessing economic evaluations for use as research evidence in a guideline. This work presents important concepts for guideline developers to consider when choosing indirect evidence sources in their clinical practice guidelines. Our findings have the potential to simplify the presentation of indirect evidence for guideline panels and developers, as well as, to reduce decision-making confusion, time demands and guideline funder costs.

The purpose of this project is to describe the necessary elements and process involved in the development of a Language Arts unit for Grade Eight students in Catholic Schools entitled "Journeys". The project focuses on the importance of Catholic education and curriculum, as well as on the developmental characteristics and educational needs of the students to which the unit is geared. The unit itself is based on the expectations as outlined in The Ontario Curriculum and on the principles of Curriculum development and Catholic Curriculum design and implementation. The completed "Journeys" unit, ready for teacher and student use, is the result of the information obtained and is included at the end of the project. / Thesis / Master of Arts in Teaching (MAT)

No description available.

No description available.

The problem set forth in this investigation is to determine which of the two methods is more effective in teaching reading and spelling in the first grade. Recent investigations would lead one to believe that a phonetic approach to reading and spelling is more effective than the non-phonetic approach.