The ability to test software is invaluable in all areas of computer science, but it is often neglected in computer science curricula. Test adequacy criteria (TAC), tools that measure the effectiveness of a test suite, have been used as aids to improve software testing teaching practices, but little is known about how students respond to them. Studies have examined the cognitive processes of students programming and professional developers writing tests, but none have investigated how student testers test with TAC. If we are to improve how they are used in the classroom, we must start by understanding the different ways that they affect students’ thought processes as they write tests. In this thesis, we take a grounded theory approach to reveal the underlying cognitive processes that students utilize as they test under no feedback, condition coverage, and mutation analysis. We recorded 12 students as they thought aloud while creating test suites under these feedback mechanisms, and then we analyzed these recordings to identify the thought processes they used. We present our findings in the form of the phenomena we identified, which can be further investigated to shed more light on how different TAC affect students as they write tests. i

Neste trabalho, apresentamos alguns métodos para avaliação das previsões do Valor em Risco (VaR). Estes métodos testam um tipo de eficiência, denominada cobertura condicional correta. O poder empírico e a probabilidade do erro de tipo I são comparados através de simulações de Monte Carlo. Além disso, avaliamos um novo método de previsão do VaR, o qual é aplicado nos retornos diários do Ibovespa. Os resultados obtidos mostram que a nova classe de testes, baseados em uma regressão Weibull discreta, em muitos casos, tem poder empírico maior comparando com outros métodos apresentados neste trabalho. / In this paper, we present some procedures for assessing forecasts for the Value at Risk (VaR). These procedures test a type of efficiency, referred as correct conditional coverage. The empirical power and type I error probability are compared through a Monte Carlo simulation. The results show that a new class of tests based on a discrete Weibull regression in most cases has greater power empirical to other methods available in this paper.