What is descriptive statistics?

Descriptive statistics are summary measures that describe characteristics of a dataset – central tendency (where data tends to be), spread (how variable data is), and shape (distribution patterns). They’re foundational to any data analysis – whether you’re a student doing homework, a businessperson reviewing sales numbers, a researcher analyzing experiments, or a casual user understanding test scores. The most common measures: mean (arithmetic average), median (middle value), mode (most frequent), range (spread from min to max), variance and standard deviation (how scattered data is around the mean). Each tells a different story – high mean with low standard deviation indicates consistent high values, while same mean with high standard deviation indicates wider variation.

How to use this tool

1. Enter numbers — Comma-separated or space-separated. Decimals supported, negative numbers OK.
2. Read all statistics — Count, sum, mean, median, mode, range, min/max, standard deviation (population and sample), variance.
3. Check sorted view — The dataset shown sorted ascending – useful for visual inspection of distribution.

Statistical formulas

• Mean (μ or x̄): sum / count = average
• Median: middle value when sorted (or average of two middle values if even count)
• Mode: most frequently occurring value(s)
• Range: max – min
• Variance (population): Σ(x – mean)² / N – average squared deviation
• Standard deviation (population): √Variance
• Variance (sample): Σ(x – mean)² / (N – 1) – uses N-1 (Bessel’s correction)
• Standard deviation (sample): √Sample Variance

Population vs Sample:

• Population: data IS the entire group (e.g. all students in a class). Use N in formulas.
• Sample: data is subset of larger group (e.g. survey of 100 from millions). Use N-1 for unbiased estimate.

Examples

Dataset: 23, 45, 12, 67, 34, 89, 23, 56, 78, 45

• Count: 10
• Sum: 472
• Mean: 47.2
• Sorted: 12, 23, 23, 34, 45, 45, 56, 67, 78, 89
• Median: (45 + 45) / 2 = 45 (average of 5th and 6th values)
• Mode: 23 and 45 (both appear 2 times)
• Range: 89 – 12 = 77
• Sample Standard deviation: 24.1

Interpretation:

Average is 47.2 with values spread ±24.1 from the mean. Most values within 47.2 ± 24.1 = 23 to 71. Outliers: 12 and 89. The data spans 12-89, which is broad but not extreme.

Tips & best practices

• Use median (not mean) when data has outliers – means are pulled by extreme values, medians aren’t
• Income data: median tells you ‘typical’ person better than mean (which gets inflated by billionaires)
• Standard deviation is in same units as original data – easier to interpret than variance
• For sample data (subset of population), use N-1 divisor (Bessel’s correction) – gives unbiased estimate
• Coefficient of variation = standard deviation / mean × 100. Helps compare spread of different datasets
• Symmetric distribution: mean = median. Skewed right: mean > median (long tail right, e.g. income). Skewed left: mean < median
• Z-score = (value – mean) / std dev. Tells you how many standard deviations a value is from the mean

Limitations & notes

These are descriptive statistics only – don’t infer causation. Don’t extrapolate beyond your data range. For inferential statistics (testing hypotheses, p-values, confidence intervals), use specialized software (R, Python, SPSS). The mode shown may be misleading for continuous data where exact repetition is rare – histograms or kernel density estimates are better for distribution shape.

Frequently Asked Questions

Mean vs median – which should I use?

Mean for symmetric distributions without outliers. Median for skewed distributions or when outliers matter. Income, real estate prices, and other distributions with long right tails are usually reported by median. Test scores and most natural measurements are often reported by mean.

What does standard deviation tell me?

Average distance from the mean. Low SD: data tightly clustered around mean (consistent). High SD: data spread widely (variable). For normal distributions, ~68% of data is within ±1 SD, ~95% within ±2 SD, ~99.7% within ±3 SD (’68-95-99.7 rule’).

Why is sample SD divided by N-1 not N?

Bessel’s correction. When you compute mean from your sample, the data has been used to estimate the mean – so you’ve lost one ‘degree of freedom’. Dividing by N-1 instead of N gives an unbiased estimate of the population SD. Important for inferential statistics where sample SD is used to estimate population parameters.

What’s the mode of [1,2,3,4,5]?

No mode (all values unique, each appears once). Or all values are modes. Convention varies – some say ‘no mode’, others say ‘all values are modes’. This calculator says ‘None’ when no value repeats.

Why do I get different answers from Excel?

Excel has STDEV.S (sample, divides by N-1, equivalent to our ‘sample SD’) and STDEV.P (population, divides by N, equivalent to our ‘population SD’). Sometimes ‘STDEV’ alone is sample. Check which formula you’re using. Our calculator shows both.

What does it mean if mean = median?

Distribution is symmetric (no skew). Examples: heights, test scores tend to be symmetric. If mean > median, data is right-skewed (long tail of high values, e.g. income, wealth). If mean < median, data is left-skewed (long tail of low values, e.g. retirement ages, lifespans).

How big should my sample size be?

Depends on research goals. For percentages with 5% margin of error and 95% confidence: ~400 samples needed for any-size population. For more precision (1% margin), 10,000 samples. For pilot studies, even 30-50 can show patterns. Statistical power calculations help determine exact requirements.

Related tools

Percentage Calculator · Percentage Change · GCD & LCM