Statistics Fundamentals
Mean, Median, and Standard Deviation: Reading Study Results
Mean, median, and standard deviation preserve different information and hide different details. Their usefulness depends on the distribution behind the summary.
Educational content only. Not medical advice.
The mean and median summarize different centers
The arithmetic mean adds observations and divides by their count, so every value influences it. The median is the middle ordered observation or midpoint between the two middle observations. In a symmetric distribution they may be similar. In a skewed distribution or one containing extreme values, the mean can move substantially while the median remains closer to the typical rank position. Neither summary is automatically correct without the distribution and scientific question.
Standard deviation describes spread around the mean
Standard deviation uses the deviations of observations from their mean to summarize variability in the same unit as the measurement. It is not the average value, not the range, and not the uncertainty of the mean. Interpreting it through familiar percentage rules assumes an approximately normal distribution; those shortcuts can fail for skewed, bounded, multimodal, or very small samples.
Sample size and distribution shape belong beside the summary
The same mean and standard deviation can arise from datasets with different shapes. A dot plot, histogram, box plot, or individual data points can reveal clusters, outliers, floor effects, and missingness that two summary numbers hide. Reported n should represent the independent experimental units, not automatically the number of repeated technical measurements. Pseudoreplication can make a dataset appear larger and more precise than its design supports.
Match the summary to the estimand and analysis
Ask what population or sample the summary describes, when measurements were taken, whether change scores or final values are shown, and how missing observations were handled. For paired or repeated measures, variability of within-subject change differs from cross-sectional variability. Descriptive statistics organize the observed data; they do not establish a causal effect or make a result applicable to an individual reader.
Evidence limits
- Summary statistics can conceal distribution shape, outliers, clusters, and missingness.
- Standard-deviation shortcuts based on normality are not universal.
- Descriptive results alone do not establish causality, generalizability, or individual response.
Sources and further reading
These sources ground the definitions and evidence boundaries on this page. A citation is a route for verification, not an endorsement of a product or personal use.
National Institute of Standards and Technology
NIST/SEMATECH e-Handbook: Measures of Location and Scale
Official reference for descriptive measures, spread, and distribution-aware analysis.
Open sourceClinicalTrials.gov, U.S. National Library of Medicine
How to Read Study Results
Official guide to analysis populations, baseline data, outcome tables, and participant flow.
Open sourceCommon questions
When is the median more informative than the mean?
Often when data are strongly skewed or contain extreme values that pull the mean away from the central rank.
Is standard deviation the uncertainty of the mean?
No. It summarizes spread among observations; uncertainty in an estimated mean is a different quantity.
Can two datasets share the same mean and standard deviation?
Yes. Their shapes, clusters, outliers, and scientific interpretation can still differ substantially.
Continue with context
Build a more rigorous reading workflow
Explore the complete PeptideSchool research workspace, including source trails, study organization, and the Premium Calculator. The tools are educational and are not a substitute for professional judgment.