PHARM/PCS V5 · Procedure 2 Dosage and Statistics
Mean, Standard Deviation and Confidence Limits
Summarize a set of replicate measurements: the mean, its scatter, and how tightly the population mean is pinned down, at a 90%, 95%, or 99% confidence level.
What it computes
For a sample of n observations, computes the arithmetic mean, sample standard deviation, standard error of the mean, and confidence limits for the population mean.
Inputs and outputs
You provide
- Your observations, one number per line or separated by commas or spaces
- A confidence level: 90%, 95%, or 99%
PHARM/PCS reports
- Mean and sample standard deviation
- Standard error of the mean
- The t value used, with n − 1 degrees of freedom
- Confidence limits for the population mean
When to use it
Any time you have several replicate measurements of one quantity (assay replicates, animals in a group, repeated readings) and want a compact, honest summary of the result and its precision. It's the descriptive foundation the comparative tests build on: the t-tests (Procedures 39 and 40) and the ANOVA family starting at Procedure 33.
Reading the results
SD describes the scatter in your data and doesn't shrink as you add observations; the standard error and confidence limits describe how tightly the mean itself is pinned down, and they do tighten as n grows. Report SD when characterizing variability, the confidence limits (or SE) when characterizing the precision of the estimate, and resist quoting SE simply because it looks smaller.
What changed in V5
The 1986 program required you to read a t value from a printed table. V5 computes the inverse Student's t directly, to about six decimals, so any degrees of freedom and any confidence level are available. For the book example the rebuild returns t = 2.3646 where the printed table gives 2.365.
This page describes the procedure. Calculations run inside the licensed PHARM/PCS V5 application. PHARM/PCS is a trademark of MCS Scientific LLC.