Hemos medido una variable binomial (i.e., 0-1) en dos
grupos.
Describiendo los datos
Así como medimos el promedio de una variable cuantitativa en un grupo
( n ), podemos medir la proporción ( p ).
p = positivos/n ( p=
Pr.positive )
Y asumimos la muestra describe a una población.
μ = np
Y también existe una varianza para esa proporción
σ2 = np(1-p)
Y un error estándar
se.binomial

Inferencia utilizando el Intervalo de Wald modificado (“Wilson´s
Score”, Agresti 2007) para interalos de confianza binomiales.

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