How To Calculate Covariance Between Two Variables

Use this premium covariance calculator to measure how two variables move together. Enter paired data for X and Y, choose sample or population covariance, and instantly see the result, interpretation, and a scatter chart that helps visualize the relationship.

Sample covariance Population covariance Scatter chart visualization Step-by-step output

Covariance Calculator

Expert Guide: How to Calculate Covariance Between Two Variables

Covariance is one of the foundational ideas in statistics, finance, economics, machine learning, and scientific research. If you want to understand whether two variables tend to move in the same direction or in opposite directions, covariance is usually one of the first measures to examine. At a practical level, covariance helps answer questions like these: do higher advertising costs tend to be associated with higher sales, do hotter temperatures tend to coincide with higher electricity use, or do two investment returns tend to rise and fall together? This guide explains exactly how to calculate covariance between two variables, how to interpret the result, when to use sample versus population covariance, and what common mistakes to avoid.

What covariance means

Covariance measures the joint variability of two variables. Suppose you have variable X and variable Y. If values of X above their average tend to occur with values of Y above their average, the covariance will be positive. If values of X above their average tend to occur with values of Y below their average, the covariance will be negative. If there is no consistent tendency for the variables to move together, the covariance may be near zero.

• Positive covariance: the variables tend to move in the same direction.
• Negative covariance: the variables tend to move in opposite directions.
• Covariance near zero: there is little or no linear co-movement.

It is important to note that covariance is not standardized. That means its magnitude depends on the units of the variables. A covariance of 50 may be large in one context and small in another. This is why analysts often use correlation after computing covariance, especially when comparing relationships across different datasets.

The covariance formulas

There are two common formulas, depending on whether your data represent an entire population or just a sample from a larger population.

Population covariance:

Cov(X, Y) = Σ[(x_i – μ_x)(y_i – μ_y)] / N

Sample covariance:

Cov(X, Y) = Σ[(x_i – x̄)(y_i – ȳ)] / (n – 1)

In these formulas:

• x_i and y_i are individual paired observations.
• μ_x and μ_y are population means.
• x̄ and ȳ are sample means.
• N is the number of population observations.
• n is the number of sample observations.

If you are working with a dataset that is only a subset of a larger real-world process, use the sample covariance formula. If you genuinely have every observation in the full population of interest, use the population formula.

Step-by-step process for calculating covariance

1. List the paired values for X and Y.
2. Compute the mean of X and the mean of Y.
3. Subtract the mean of X from each X value to get deviations.
4. Subtract the mean of Y from each Y value to get deviations.
5. Multiply each X deviation by the matching Y deviation.
6. Add all of those products together.
7. Divide by N for population covariance or by n – 1 for sample covariance.

Worked example

Assume you are studying the relationship between study hours and exam scores for five students. Let X represent study hours and Y represent exam score percentage.

Student	Study Hours (X)	Exam Score (Y)	X – x̄	Y – ȳ	(X – x̄)(Y – ȳ)
1	2	68	-2	-10	20
2	4	74	0	-4	0
3	5	78	1	0	0
4	6	83	2	5	10
5	8	87	4	9	36
Total					66

The mean of X is 5, and the mean of Y is 78. The sum of the products of deviations is 66. Because this is a sample of five students, the sample covariance is:

66 / (5 – 1) = 16.5

The positive value tells us that study hours and exam scores tend to increase together. The more students study, the higher scores tend to be, at least in this small sample.

A positive covariance tells you the direction of the relationship, not the exact strength on a common scale. If you need a standardized metric between -1 and 1, compute correlation.

How to interpret covariance correctly

Interpreting covariance is straightforward in terms of direction but more nuanced in terms of size. Here is the practical interpretation:

• If covariance is positive, X and Y generally move together.
• If covariance is negative, when X goes up, Y tends to go down.
• If covariance is around zero, there may be no clear linear relationship.

However, the magnitude of covariance depends on the units of measurement. If one variable is measured in dollars and another in percentages, the covariance unit becomes dollars-times-percentages. Because of this, a covariance value by itself can be hard to compare across studies, time periods, or industries.

Sample covariance versus population covariance

The distinction matters because the denominator changes. The sample version uses n – 1 to correct bias when estimating the covariance of a larger population from limited observations. The population version uses N because no estimation correction is needed.

Feature	Sample Covariance	Population Covariance
When to use	When your dataset is a subset of a larger population	When your dataset contains the full population of interest
Denominator	n – 1	N
Purpose	Estimate the covariance in the broader population	Describe the exact covariance of the complete set
Typical applications	Surveys, experiments, sample-based financial studies	Full census data, complete production records, fully observed datasets

Real-world uses of covariance

Covariance is widely used because many important decisions depend on whether variables move together.

• Finance: portfolio theory uses covariance between asset returns to understand diversification. If two assets have low or negative covariance, combining them may reduce portfolio risk.
• Economics: analysts examine covariance between inflation and wage growth, interest rates and investment, or GDP and employment metrics.
• Public health: researchers may explore covariance between age and blood pressure, exercise levels and health outcomes, or pollution and respiratory symptoms.
• Business analytics: teams often check covariance between ad spend and sales, website traffic and conversions, or pricing changes and unit demand.
• Machine learning: covariance matrices are central in dimensionality reduction, multivariate modeling, and feature understanding.

Comparison table with practical statistics

Below is a simple illustrative dataset showing how covariance changes across business scenarios. The values are realistic example statistics for explanation purposes.

Scenario	Variable X	Variable Y	Sample Size	Estimated Covariance	Interpretation
Retail marketing	Weekly ad spend	Weekly sales revenue	52 weeks	18,450	Higher ad spend tends to occur with higher sales
Energy demand	Daily temperature	Home heating usage	90 days	-12.8	As temperature rises, heating demand tends to fall
Education study	Hours studied	Test scores	120 students	9.6	Students who study more tend to score higher
Web analytics	Page load time	Conversion rate	30 campaigns	-0.42	Slower pages tend to be associated with lower conversions

Common mistakes when calculating covariance

• Mismatched pairs: each X value must correspond to the correct Y value from the same observation.
• Using the wrong denominator: sample data should typically use n – 1, not n.
• Interpreting magnitude without context: a larger number does not automatically mean a stronger relationship because units matter.
• Ignoring outliers: extreme values can strongly affect covariance.
• Assuming causation: covariance shows co-movement, not proof that one variable causes the other.

Covariance vs correlation

Covariance and correlation are closely related, but they answer slightly different questions. Covariance tells you the direction of joint movement and preserves the original units of the variables. Correlation standardizes covariance by dividing by the product of standard deviations. That gives a scale from -1 to 1, making interpretation and comparison easier.

• Use covariance when you need the raw joint variability or when working with covariance matrices.
• Use correlation when you need a standardized measure of relationship strength.

Why covariance matters in data analysis

Many advanced methods build on covariance. In portfolio optimization, the covariance matrix helps determine how total risk behaves when assets are combined. In principal component analysis, covariance structure reveals which combinations of variables explain the most variation. In multivariate regression and signal processing, covariance plays a direct role in estimation and model structure. Understanding the manual calculation helps you understand what these tools are doing under the hood.

Authoritative references for deeper learning

If you want academically solid definitions and broader context, these resources are excellent starting points:

• U.S. Census Bureau statistical reference material
• University of California, Berkeley statistics notes on covariance and correlation
• NIST Engineering Statistics Handbook

Final takeaway

To calculate covariance between two variables, compute the means of X and Y, find each pair of deviations from the mean, multiply the paired deviations, sum those products, and divide by either N or n – 1 depending on whether you have a population or a sample. A positive result means the variables tend to rise and fall together, a negative result means they tend to move in opposite directions, and a value near zero suggests little linear co-movement. Covariance is simple to calculate, powerful in practice, and essential for understanding how variables behave as a system rather than in isolation.

Use the calculator above whenever you want a quick, accurate way to compute covariance, inspect summary statistics, and visualize the paired relationship in a chart.