A análise de dados de proporções apresenta, em geral, certas dificuldades uma vez que a distribuição subjacente a tais dados pode ser considerada binomial, que não segue as pressuposições básicas para o ajuste de um modelo matemático. Algumas transformações são sugeridas, mas nem sempre bons resultados são obtidos. No enfoque de modelos lineares generalizados, a estatística que mede a qualidade do ajuste do modelo para os dados é chamada deviance. Ocorre que a distribuição da deviance é desconhecida. No entanto, para dados com distribuição binomial, pode-se aproximar a distribuição da deviance por uma distribuição qui-quadrado, mas tal aproximação não é boa para tamanhos pequenos de amostra. Para melhorar essa aproximação, alguns fatores de correção para os dados são sugeridos, mas os resultados obtidos ainda não são bons para pequenas amostras. Assim, o objetivo deste trabalho é propor um novo fator de correção para os dados seguindo uma distribuição binomial, de modo a se obter uma melhora na distribuição da deviance para qualquer tamanho de amostra. Para isto, adiciona-se uma constante à variável resposta e, através do valor esperado da deviance, calcula-se tal constante de modo a reduzir o erro cometido na aproximação. Simulações da distribuição binomial e o cálculo da deviance são feitos e QQ-plots são utilizados para a comparação com a distribuição qui-quadrado / not available

Trata-se da análise de 2222 consultas de enfermagem para o binômio mãe-filho executadas por 14 enfermeiras obstetras, no ano de 1983 em um Hospital Universitário, com o propósito de verificar se os procedimentos normatizados na consulta de enfermagem estão possibilitando o levantamento de informações relativas à clientela atendida e direcionando as intervenções da enfermeira, permitindo efetiva individualização das condutas de enfermagem. Foram levantados 5024 problemas, dos quais 3548 nos recém-nascidos e 1476 nas puérperas. Quanto às condutas de enfermagem executadas, estas foram em número de 4768 e conforme o registro, assim distribuídas: 714 - cuidados executados; 3270 - orientações ministradas; 252 - encaminhamentos para nova consulta com enfermeira; 458 - encaminhamentos para consulta médica; 48 - encaminhamentos para nutricionista, e 26 - encaminhamentos para a assistente social. Não tiveram suas condutas registradas, 1485 problemas. Os tipos de problemas levantados pelas enfermeiras não variaram: a atuação das enfermeiras não se restringe às condutas estabelecidas no guia de orientação e as orientações independentes do registro dos problemas foram em número bastante significativas. A maneira como a consulta de enfermagem se desenvolveu, possibilitou a caracterização da enfermeira na sua função independente, mas utilizando a interdependência de forma coerente, científica e individualizando a clientela. / This paper reviews 2.222 nursing consultations for the mother-child binomial, carried out by 14 obsteinic-nurses in 1983, at a teaching Hospital with the objective to find out if the standardized procedures for the nursing consultations are enabling the survey of information related to the clientele assisted and if the said procedures are directing the nurse\'s interference, to the actual individualization of the nursing approach. 5024 problems have been surveyed, of which 3548 in newborns and 1476 in mothers. As to the nursing approach utilized, there were 4768 and, according to the data, they were distributed as follows: 714 - care services performed; 3270 - instructions given; 252 - references to a new nursing consultation; 458 - references to medical consultation; 48 - references to nutritionists; and 26 - references to social workers. 1485 problems did not have their related nursing approach registered. The types of problems surveyed by the nurses did not show any variation; the nurse\'s activities are not limited to the ones established in the instructions manual and the number of orientations, regardless of the registration of problems, was significant. The way in which the nursing consultation was developed characterizes the nursing approach, as an independent function using, however, the interdependent function in a consistent and scientific manner as well as the individualization of the clientele.

Let p be a prime, p a power of p and 𝔽q the finite field with q elements. Any function φ: 𝔽q → 𝔽q can be unqiuely represented by a polynomial, 𝔽φ of degree < q. If the map x ↦ Fφ(x) induces a permutation on the underlying field we say Fφ is a permutation polynomial. Permutation polynomials have applications in many diverse fields of mathematics. In this dissertation we are generally concerned with the following question: Given a polynomial f, when does the map x ↦ F(x) induce a permutation on 𝔽q. In the second chapter we are concerned the permutation behavior of the polynomial gn,q, a q-ary version of the reversed Dickson polynomial, when the integer n is of the form n = qa - qb - 1. This leads to the third chapter where we consider binomials and trinomials taking special forms. In this case we are able to give explicit conditions that guarantee the given binomial or trinomial is a permutation polynomial. In the fourth chapter we are concerned with permutation polynomials of 𝔽q, where q is even, that can be represented as the sum of a power function and a linearized polynomial. These types of permutation polynomials have applications in cryptography. Lastly, chapter five is concerned with a conjecture on monomial graphs that can be formulated in terms of polynomials over finite fields.

Master of Science / Department of Statistics / Christopher Vahl / Among the most used statistical concepts and techniques, seen even in the most cursory of introductory courses, are the confidence interval, binomial distribution, and sample size estimation. This paper investigates a particular case of generating a confidence interval from a binomial experiment in the case where zero successes are expected. Several current methods of generating a binomial proportion confidence interval are examined by means of large-scale simulations and compared in order to determine an ad-hoc method for generating a confidence interval with coverage as close as possible to nominal while minimizing width. This is then used to construct a formula which allows for the estimation of a sample size necessary to obtain a sufficiently narrow confidence interval (with some predetermined probability of success) using the ad-hoc method given a prior estimate of the probability of success for a single trial. With this formula, binomial experiments could potentially be planned more efficiently, allowing researchers to plan only for the amount of precision they deem necessary, rather than trying to work with methods of producing confidence intervals that result in inefficient or, at worst, meaningless bounds.

Diploma thesis deals with models of asset pricing. We investigated in detail three classical models: binomial, Black-Scholes and Merton model. These models are widely used to date, although they were first published a few decades ago. It is because they are relatively simple and easy-to-use. The models were originally derived for option pricing however they can be used for the wide range of financial instruments. The theoretical part of the thesis includes an introduction to options and models derivation. The practical part consists of the sensitivity analyst and empirical test of the models. S&P 500 index options data were used for this purpose. The result is that Merton model seems to be the most accurate.

A set-theoretical point of view to study algebraic numbers has been introduced. We extend a result of Navarro-Bermudez concerning shift invariant measures in the Cantor space which are topologically equivalent to shift invariant measures which correspond to some algebraic integers. It is known that any transcendental numbers and rational numbers in the unit interval are not binomial. We proved that there are algebraic numbers of degree greater than two so that they are binomial numbers. Algebraic integers of degree 2 are proved not to be binomial numbers. A few compositive relations having to do with algebraic numbers on the unit interval have been studied; for instance, rationally related, integrally related, binomially related, B1-related relations. A formula between binomial numbers and binomial coefficients has been stated. A generalized algebraic equation related to topologically equivalent measures has also been stated.

No description available.

This research project investigates the effectiveness of mentors on rates of self-reported criminal offending for released offenders. I use data from the Serious and Violent Offender Reentry Initiative (SVORI) study (2004-2007), which sought to evaluate factors relating to high-risk offenders outcomes post release in an effort to reduce the societal problem of mass incarceration. Previous research has examined the use of mentors to improve the delinquent and criminogenic behavior of youth, but little is known about the effectiveness of mentors used to aid imprisoned adult males. I utilize negative binomial analysis to compare the number of self-reported criminal activities among released offenders with mentors versus those without mentors, and assess if the values varied between different reported levels of need for mentoring. Results indicate that mentoring did not reduce the rate of post-release offending at a statistically significant level. Reasons for the lack of significant results and policy implications are discussed.

Several large-sample confidence intervals for the ratio of independent binomial proportions are compared in terms of exact coverage probability and width. A non-iterative approximate Bayesian interval is derived and its frequency properties are superior to all of the non-iterative confidence intervals considered. The approximate Bayesian interval, which is very easy to compute, has performance characteristics that are very similar to the computationally intensive score method. Two sample size determination formulas are presented, one for desired absolute precision and the other for desired relative precision.

Having the ability to predict enrollment is an important task for any school's recruiting team. The purpose of this study was to identify significant factors that can be used to predict the spatial distribution of enrollments. As a case study, we used East Tennessee State University (ETSU) pharmacy school, a regional pharmacy school located in the Appalachian Mountains. Through the application of a negative binomial regression model, we found that the most important indicators of enrollment volume for the ETSU pharmacy school were Euclidean distance, probability (based on competing pharmacy schools' prestige, driving distance between schools and home and tuition costs), and the natural barrier of the Appalachian Mountains. Using these factors, together with other control variables, we successfully predicted the spatial distribution of enrollments for ETSU pharmacy school. Interestingly, gender also surfaced as a variable for predicting the pharmacy school's enrollment. We found female students are more sensitive to the geographic proximity of home to school.