1. Other things being equal, the greater the level of confidence desired, the wider will be the confidence interval.
2. Repeated samples of the same size from the same population will always produce 99% confidence intervals of the same width on the binomial parameter p.
3. If the confidence interval does not contain some hypothesized value p0 of the binomial parameter, the hypothesis can be rejected.