tinterval(x[,y],cover:fraction,[upper:T or lower:T, pool:T|F]) |
tinterval() computes a t-confidence interval for a population mean or difference of population means, depending on whether one or two variables are given as arguments. By default, tinterval() computes a two-sided interval, but you may choose one-sided alternatives by using one of lowerb:T or upperb:T. You specify the coverage rate via cover:value. You may choose pooled or unpooled variances in two-sample tests via pool:T or pool:F. The output is a vector containing the estimate and interval bounds. Because the variance of the data must be known, these intervals are rarely used in practical data analysis.