What is a statistics calculator?

A statistics calculator computes summary statistics — mean, median, mode, variance, standard deviation, range, quartiles — from a list of numbers. It also produces a histogram so you can see the distribution at a glance.

This is the workhorse for data analysis: descriptive statistics that summarize "what is the typical value, how spread out is it, are there outliers?"

Common statistics defined

Mean (average)

mean = sum of values / count of values

The center of mass. Sensitive to outliers — one ₹10 crore salary in a list of ₹5 lakh salaries skews the mean.

Median

The middle value when sorted. If even count, average of two middles. Robust to outliers — a CEO's salary doesn't move the median.

Mode

The most frequently occurring value. Useful for categorical data ("most common shoe size sold"). Can be multiple (bimodal) or absent.

Range

range = max − min

Simplest spread metric.

Variance

variance = sum((value - mean)²) / N

Mean of squared deviations from the mean. In units squared (so ₹² for incomes — awkward).

Standard deviation (σ)

σ = sqrt(variance)

The "typical distance from the mean", in the original units. Most useful spread metric.

Quartiles (Q1, Q2, Q3)

• Q1 = 25th percentile (lowest 25% are below this)
• Q2 = median (50th)
• Q3 = 75th percentile
• IQR = Q3 − Q1 (interquartile range — robust spread measure)

Worked example

Dataset: monthly salaries (₹ thousands):

45, 55, 60, 60, 65, 70, 75, 80, 90, 200

Sorted (already done).

Count (N) = 10
Sum = 800
Mean = 800 / 10 = ₹80,000

Median = (5th + 6th) / 2 = (65 + 70) / 2 = ₹67,500
Mode = 60 (appears twice)
Range = 200 − 45 = 155
Min = 45, Max = 200
Q1 = ₹60,000 (between 2nd and 3rd values)
Q3 = ₹80,000 (between 7th and 8th values)
IQR = 80 − 60 = ₹20,000

Variance (population):

Deviations²: (45-80)², (55-80)², ..., (200-80)²
           = 1225, 625, 400, 400, 225, 100, 25, 0, 100, 14400
Sum = 17500
Variance = 17500 / 10 = 1750
σ = sqrt(1750) ≈ 41.83 (₹41,830)

The mean (₹80k) is misleading — the ₹2,00,000 outlier inflates it. The median (₹67.5k) is more representative of a "typical" person. σ of ~42 says the spread is large, dominated by that one outlier.

Population vs sample statistics

When data is a complete population: divide variance by N.

When data is a sample from a larger population: divide variance by N-1 (Bessel's correction). This gives an unbiased estimate of the population variance.

For our example:

Population variance = 17500 / 10 = 1750
Sample variance = 17500 / 9 ≈ 1944
σ_sample ≈ √1944 ≈ 44.1

Difference is small for large N, big for small N. Use sample (N-1) unless you have the full population — which is rare.

Worked example: skewness and kurtosis

Beyond center and spread, you can describe shape:

• Skewness = asymmetry. Positive skew = long right tail (incomes, wait times). Negative = long left tail.
• Kurtosis = "tailedness". High kurtosis = fat tails (stock returns).

For the salary example: skewness > 0 (long right tail from the ₹2L outlier).

When to use which average

Common mistakes

"Average salary at the company is ₹10 lakh"

If the CEO earns ₹2 cr and 99 employees earn ₹5 lakh, the mean is ₹2.45 lakh. But the median is ₹5 lakh — far more representative. News headlines abuse this routinely.

"Standard deviation says we're inconsistent"

Without context, σ is meaningless. σ of 10 on a base of 100 (10%) is very different from σ of 10 on a base of 10,000 (0.1%). Always quote σ relative to mean (coefficient of variation = σ / mean × 100%).

Comparing variances across different units

"Sales had σ = 50, costs had σ = 30, so sales are more volatile." Wrong if sales are 10x bigger than costs.

Treating averages as predictions

"Average height is 170 cm" doesn't mean every person is 170 cm. Without σ, you have no idea if the population is 165-175 or 140-200.

Components and inputs

Data entry

Comma-separated, space-separated, newline-separated, or pasted from spreadsheet.

45, 55, 60, 60, 65, 70, 75, 80, 90, 200

or

45
55
60
...

Output options

• Summary table: count, sum, mean, median, mode, min, max, range
• Spread: variance, σ (population and sample)
• Quartiles: Q1, Q2, Q3, IQR
• Outliers: > Q3 + 1.5×IQR or < Q1 - 1.5×IQR
• Histogram: visualize distribution

Population vs sample toggle

Switch between N and N-1 divisor for variance.

Box plot interpretation

The classic 5-number summary on a box:

|------ box ------|
|   |    |   |   |
min Q1  med Q3  max

• Box length = IQR
• Median line shows skew (off-center = skewed)
• Whiskers = 1.5×IQR from box, or actual extremes
• Dots outside whiskers = outliers

Common real-world uses

Considerations

• Always look at the data, not just summary stats. Anscombe's quartet shows 4 datasets with identical mean/σ/correlation but totally different shapes.
• Outliers are signal AND noise. A ₹2 cr salary in a list of ₹5 lakh is a real datum about the CEO; whether to include it depends on the question you're asking.
• Bigger N = more reliable estimates. σ of 10 from N=5 is shaky; σ of 10 from N=10,000 is solid.
• σ is in the same units as the data. Variance is in units squared (rarely intuitive).

Limitations

• Doesn't fit distributions (use a fitting tool for that).
• Doesn't do hypothesis testing (t-test, chi-square — see specialized stats tools).
• Doesn't handle missing data (drop NaNs first).
• Categorical data needs frequency counts, not these tools.

Related calculators

• Scientific Calculator — manual computation
• Quadratic Solver — algebraic roots
• Permutation & Combination — nPr, nCr
• Percentage — % changes
• Normal Distribution — z-scores, probabilities

Final note. Statistics summarize data — they don't replace looking at the data. Always check the histogram alongside the mean and σ. For skewed data (incomes, wait times, prices), prefer median over mean. For mission-critical decisions, report percentiles (P50, P95, P99) rather than averages — they say much more about real-world performance.