Confidence intervals are an important tool to indicate uncertainty of estimates and to give an idea of probable values of an estimate if a different sample from the population was drawn or a different sample of measures was used. Standard symmetric confidence intervals for proportion estimates based on a normal approximation can yield bounds outside the [0,1] scale and poor coverage, because such approximations are generally inappropriate. Many alternative intervals have been proposed to address these issues. This paper discusses a selection of intervals based on scale transformations and continuity corrections and adapts these for use in complex samples. The study expands on the work of Brown, Cai, and DasGupta (2001) and Korn and Graubard (1998) using educational survey designs and complex sample data. Results based on a National Assessment of Educational Progress (NAEP) data resampling study showed that the theoretically appealing Wilson interval yields appropriate coverage with short intervals in most situations.