Evaluates a probability of outcome superiority curve, with its confidence
interval, at any differences in the predictor.
Usage
# S3 method for class 'posc'
predict(object, delta, level = object$level, ...)
Arguments
- object
A "posc" object.
- delta
Numeric vector of signed differences in the predictor
(person 1 minus person 2), in the units of the predictor. Negative values
give \(1 - P(|\Delta|)\).
- level
Confidence level for the interval. Defaults to the level used
when the curve was created.
- ...
Unused.
Value
A data frame with columns delta, p (probability that person 1
has the higher outcome), lower and upper (confidence limits; NA
when no interval is available).
Examples
fit <- posc_uni(r = 0.5, n = 100, sd_x = 15)
predict(fit, delta = c(5, 10, 15))
#> delta p lower upper
#> 1 5 0.5541221 0.5335775 0.5766406
#> 2 10 0.6072526 0.5669176 0.6504787
#> 3 15 0.6584543 0.5997878 0.7190156
