scholarly journals ¿Por qué en ocasiones los resultados de las encuestas de opinión se alejan de la realidad? Un análisis con escenarios en Python

Author(s):  
Francisco Benavides Murillo ◽  
Byron Jiménez Oviedo
Keyword(s):  

Antes de la segunda ronda de las elecciones presidenciales costarricenses del año 2018 en Costa Rica, se realizaron varias encuestas de opinión con el fin de evaluar las preferencias del electorado. Las encuestadoras en Costa Rica indicaban, en algunos casos, un empate técnico entre el candidato de Restauración Nacional y el del Partido Acción Ciudadana, y en otros casos una clara preferencia por el primero. El resultado final de las elecciones fue muy diferente al de estos estudios, y el triunfo del candidato del partido Acción Ciudadana fue claro y totalmente alejado de los márgenes de error asociados con la mayoría de los resultados obtenidos por las empresas encuestadoras. A raíz de esto, una pregunta natural surge: ¿por qué fallan las encuestas electorales? Uno de los argumentos utilizados es que las encuestas fueron realizadas con al menos una semana de antelación a las votaciones, y la opinión del electorado cambió en ese tiempo.  En el presente trabajo se plantean situaciones hipotéticas de posibles escenarios en los que el sesgo de las encuestas se puede presentar, así como una metodología para su detección. Se analiza, en particular, la tasa de respuesta de los encuestados y su influencia en las estimaciones de la encuesta. El hecho es que si los que apoyan a un determinado partido tienen una tendencia generalizada de no responder una encuesta, entonces los resultados de ésta serán sesgados, sin importar el momento en que se hagan las entrevistas. En este trabajo, los escenarios sintéticos son simulados computacionalmente utilizando Python por su interés didáctico. También se simulan métodos de ponderación para compensar estas deficiencias. Este artículo pretende abrir un espacio didáctico de discusión de los aspectos técnicos, matemáticos y computacionales de las encuestas de opinión nacionales, así como producir un escenario sintético computacional que pueda derribar algunos mitos populares frecuentes, como el mito de que sólo mil personas encuestadas no son suficientes para obtener una representación significativa de la opinión del electorado.

Author(s):  
O. E. Bradfute

Maize rayado fino virus (MRFV) causes a severe disease of corn (Zea mays) in many locations throughout the neotropics and as far north as southern U.S. MRFV particles detected by direct electron microscopy of negatively stained sap from infected leaves are not necessarily distinguishable from many other small isometric viruses infecting plants (Fig. 1).Immunosorbent trapping of virus particles on antibody-coated grids and the antibody coating or decoration of trapped virus particles, was used to confirm the identification of MRFV. Antiserum to MRFV was supplied by R. Gamez (Centro de Investigacion en Biologia Celular y Molecular, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica).Virus particles, appearing as a continuous lawn, were trapped on grids coated with MRFV antiserum (Fig. 2-4). In contrast, virus particles were infrequently found on grids not exposed to antiserum or grids coated with normal rabbit serum (similar to Fig. 1). In Fig. 3, the appearance of the virus particles (isometric morphology, 30 nm diameter, stain penetration of some particles, and morphological subunits in other particles) is characteristic of negatively stained MRFV particles. Decoration or coating of these particles with MRFV antiserum confirms their identification as MRFV (Fig. 4).


2001 ◽  
Vol 60 (2) ◽  
pp. 89-98 ◽  
Author(s):  
Alain Clémence ◽  
Thierry Devos ◽  
Willem Doise

Social representations of human rights violations were investigated in a questionnaire study conducted in five countries (Costa Rica, France, Italy, Romania, and Switzerland) (N = 1239 young people). We were able to show that respondents organize their understanding of human rights violations in similar ways across nations. At the same time, systematic variations characterized opinions about human rights violations, and the structure of these variations was similar across national contexts. Differences in definitions of human rights violations were identified by a cluster analysis. A broader definition was related to critical attitudes toward governmental and institutional abuses of power, whereas a more restricted definition was rooted in a fatalistic conception of social reality, approval of social regulations, and greater tolerance for institutional infringements of privacy. An atypical definition was anchored either in a strong rejection of social regulations or in a strong condemnation of immoral individual actions linked with a high tolerance for governmental interference. These findings support the idea that contrasting definitions of human rights coexist and that these definitions are underpinned by a set of beliefs regarding the relationships between individuals and institutions.


2010 ◽  
Vol 30 (S 01) ◽  
pp. S28-S31 ◽  
Author(s):  
J. Arroyo ◽  
L. Salazar-Sánchez ◽  
G. Jiménez-Cruz ◽  
P. Chaverri ◽  
E. Arrieta-Bolaños ◽  
...  

SummaryHaemophilia is the most frequent hereditary haemorrhagic illness and it is due to the deficiency of coagulation factors VIII (haemophilia A, HA) or IX (haemophilia B, HB).The prevalence of this disease varies according to the country, those having better survival rates having also higher prevalences. Specifically in Costa Rica, there are around 130 HA and 30 HB families. This study reports the prevalence and a spatial distribution analysis of both types of the disease in this country. The prevalence of haemophilia in this country is 7 cases per 100 000 men, for HA it is 6 cases per 100 000 and for HB it is 1 case per 100 000 male inhabitants. The prevalence of this disease is low when compared with other populations. This low prevalence could be due to the many patients that have died because of infection with human immunodeficiency virus during the 1980s. The prevalence of haemophilia in Costa Rica is almost one half of that present in developed countries. Nevertheless, the ratio between HA and HB follows world tendency: 5 : 1. In this study, nationwide geographical distribution maps were drawn in order to visualize the origin of severe cases and how this influences the pattern of distribution for both types of haemophilia. By means of these maps, it was possible to state that there is no association between the sites of maximum prevalence of mutated alleles and ethnicity. With this study, haemophilia prevalence distribution maps can be used to improve efforts for the establishment of hemophilia clinics or specialized health centers in those areas which hold the highest prevalences in this country. Also, this knowledge can be applied to improve treatment skills and offer the possibility of developing focused genetic counseling for these populations.


2019 ◽  
Author(s):  
Marino Protti ◽  
◽  
Nathan Bangs ◽  
Peter Baumgartner ◽  
Donald Fisher ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document