Search

The quarantine is an intervention procedure to control the spread of infectious diseases among farms. However, the dilemma persists of which and how many farms need to be quarantined to achieve effective control, or at least, to make the losses minimum. In the present paper we propose an SEIR modified stochastic model to decide with high objectivity and rapidity the best quarantine strategy. Three quarantine strategies with different percentage of quarantine farms are compared by means of Monte Carlo simulation. We find that the minimum final size of an epidemic is obtained if a small fraction of farms with a high number of edges is quarantined. In addition, the proposed model was applied to a set of real data for avian influenza reported in the northwest of Tehuacán Valley, Puebla, México in 1995. [ABSTRACT FROM AUTHOR]

En el mejoramiento genético de plantas y animales es común realizar la evaluación de tratamientos en diversas condiciones ambientales, surgiendo la necesidad de proyectar investigaciones en una serie de experimentos similares, tales como en los diseños de Griffing. En tal caso, en la mayoría de las ocasiones, el análisis combinado de las series de experimentos, incluyendo los de cruzas dialélicas, se han efectuado con un modelo lineal de efectos fijos, a pesar de la naturaleza aleatoria de la aptitud combinatoria general y específica en los experimentos de cruzas dialélicas. Por ello, en el presente trabajo, para los diseños I y III de Griffing establecidos en experimentos de bloques completos al azar, se derivan los mejores predictores lineales e insesgados combinados de los efectos de aptitud combinatoria general, con base en la metodología de estimación de efectos aleatorios en un modelo de efectos mixtos. Adicionalmente, se presenta un algoritmo computacional, en SAS/IML, para el análisis combinado de experimentos. [ABSTRACT FROM AUTHOR]

This article compares the educational level of full-time faculty members at Universidad de Colima with their teaching and academic performance, based on an analysis of variance and canonical discrimination in the 2001-2004 period. During each one of the four years, in terms of teaching performance, no significant difference was found in educational level: professors with doctoral, master's and undergraduate degrees were statistically equal in teaching performance. Educational level was found to have a direct effect, however, on academic performance: the more education the teacher had, the better his academic performance. The predominant factors were teaching duties and research activities, thus minimizing the influence of tutoring and collegiate action. [ABSTRACT FROM AUTHOR]