PHARM/PCS
  • Home
  • About
    • Ronald J. Tallarida
    • Rodney B. Murray
  • Manual
  • Software
  • Procedures
  • Customers
    • Foreign Customers
  • Contact
Procedures›Procedure 43

PHARM/PCS V5 · Procedure 43 Pairwise and Nonparametric

Proportions: Confidence Limits

Put exact confidence limits on a single proportion, such as a response rate, that stay within 0 and 1 even for small samples.

Try PHARM/PCS V5 → Browse all 48 procedures →

What it computes

Computes confidence limits for a proportion p = m/N using the exact formula (Equations 43.2-43.5) with continuity correction.

Inputs and outputs

You provide

  • Number responding m
  • Total subjects N
  • A confidence level

PHARM/PCS reports

  • The proportion p = m/N
  • Its exact, asymmetric confidence limits

When to use it

When you have a single proportion (response rate, incidence, hit rate) and need a confidence interval around it. It pairs naturally with the chi-square test (Procedure 42) when you move from estimating one proportion to comparing categories.

Reading the results

These exact limits are asymmetric, especially for small N or proportions near 0 or 1, where a symmetric interval would spill past the 0-to-1 bounds; the asymmetry is a feature. The narrower the interval you need, the larger the N required, and near the extremes that requirement grows steeply.

Run proportion confidence limits in PHARM/PCS

Browser-based software with 48 pharmacologic calculation procedures. Academic and commercial licenses.

Try PHARM/PCS V5 →
← Procedure 42Chi-Square Test Procedure 44 →Dunnett's Test (Comparison with a Control)

This page describes the procedure. Calculations run inside the licensed PHARM/PCS V5 application. PHARM/PCS is a trademark of MCS Scientific LLC.

Copyright © 2014-2026  MCS Scientific LLC. All Rights Reserved.