propinterval(x[,y],cover:fraction,[options]) propinterval(s1,n1[,s2,n2],cover:fraction,[options]) options can include upperb:T or lowerb:T, plus4:T|F. |
propinterval() computes a z-confidence interval for a population proportion or difference of population proportions, depending on whether data for one or two variables are given as arguments. By default, propinterval() computes a two-sided interval, but you may choose one-sided alternatives by using one of lowerb:T or upperb:T. Arguments may be either vectors of 0/1 data, or may be counts of successes and numbers of trials. You specify the coverage rate via cover:value. You may specify the Agresti "add 4" procedure via plus4:T. This will add 2 successes and two failures (split among the two samples for two-sample tests). The output is a vector containing the estimate and interval bounds.