R vs SAS Confidence Intervals for a Proportion

Introduction

See the summary page for general introductory information on confidence intervals for proportions, including the principles underlying the most common methods.

For more technical derivation of methods for a single proportion, see the corresponding SAS page.

The tables below provide an overview of findings from R & SAS, for calculation of CIs for a Single Sample Proportion.

General Comparison Table For Single Sample Proportions

See the corresponding SAS page and R page for results showing a single set of data which has been run through both SAS and R.

Analysis of One Sample Proportion Supported in R Supported in SAS Results Match
Normal approximation (Wald Method) Yes {cardx} / {ratesci} Yes (default) Yes
Wilson (Score) method Yes {cardx} / {ratesci} Yes Yes
Skewness-Corrected Asymptotic Score (SCAS) method Yes {ratesci} No NA
Agresti-Coull Yes {cardx} / {ratesci} Yes Yes
Jeffreys Bayesian ‘equal-tailed’ Yes {cardx} / {ratesci} Yes Yes
mid-P Yes {ratesci} Yes Yes
Wilson Stratified score Yes {cardx} / {ratesci} No NA
‘exact’ and continuity adjusted methods:
Clopper-Pearson ‘Exact’ Yes {cardx} / {ratesci} Yes (default) Yes
Blaker ‘exact’ Yes {ratesci} Yes Yes
Normal approximation (Wald Method) with continuity adjustment Yes {cardx} / {ratesci} Yes Yes
Wilson (Score) method with continuity adjustment Yes {cardx} / {ratesci} Yes Yes
SCAS method with continuity adjustment Yes {ratesci} No NA

Prerequisites: R Packages

See the R page for more detail.

# Example R packages required
library(cardx)
library(ratesci)