Correlation Coefficient on a Calculator: Find Pearson’s r

This calculator finds Pearson’s correlation coefficient, r, from paired x and y values. Enter the same number of observations in each list to get a value from −1 to 1, showing the direction and strength of their linear relationship.

Formula

Pearson’s correlation coefficient is calculated as:

r = [nΣ(xy) − (Σx)(Σy)] / √{[nΣ(x²) − (Σx)²][nΣ(y²) − (Σy)²]}

• x and y are corresponding observations in the two data sets; they must be paired in the same order.
• n is the number of pairs, a count with no units.
• Σx and Σy are the sums of the x and y values.
• Σxy is the sum of each x value multiplied by its paired y value.
• Σ(x²) and Σ(y²) are the sums of the squared values.

The coefficient has no units. This calculator uses lists, so enter numeric data separated by commas. The two lists need equal lengths. It calculates the standard sample Pearson correlation; the formula does not require converting measurements between units, provided each variable uses a consistent unit throughout its list.

Worked Example

Using the default lists, the pairs are (1, 2), (2, 3), (3, 5), (4, 4), and (5, 6). Thus, n = 5, Σx = 15, Σy = 20, Σxy = 69, Σ(x²) = 55, and Σ(y²) = 90.

Substitute these totals into the formula:

r = [5(69) − (15)(20)] / √{[5(55) − 15²][5(90) − 20²]}

r = 45 / √(50 × 50) = 45 / 50 = 0.900

The calculator rounds to three decimal places, so the result is 0.900.

When to Use It and How to Read the Result

Use Pearson’s r when you want to summarize how closely two quantitative variables follow a straight-line pattern. For example, each x value might be a study time and its paired y value an exam score. The coefficient describes association, not cause and effect.

A result near +1 indicates a strong positive linear association: larger x values tend to occur with larger y values. A result near −1 indicates a strong negative linear association: larger x values tend to occur with smaller y values. A result near 0 means little or no linear association in the data. It does not rule out a curved relationship.

There is no universal cutoff for calling a correlation weak, moderate, or strong. Interpret its size in context, and inspect a scatterplot when possible. An unusual observation can substantially change r, especially in a small data set.

Step-by-Step

1. Prepare paired observations. Choose two quantitative variables measured for the same cases. Each x value must correspond to the y value in the same position. Remove a pair only when you have a clear reason to exclude that observation; do not delete just one value and shift the remaining pairings.
2. Enter the x list. In the first list, replace the defaults with your x observations, separated by commas. Do not type labels, units, or extra punctuation as list entries.
3. Enter the y list. Replace the second list with the corresponding y observations, in exactly the same order. Check that both lists contain the same number of values.
4. Read the result. The output labeled Pearson correlation coefficient is r. A positive or negative sign gives the direction; the distance from zero indicates the strength of the linear association.
5. Check the pattern. If you can, plot the paired values. Look for curved patterns, clusters, or influential outliers that a single correlation number can hide.
6. Report it responsibly. State the coefficient and, when useful, the sample size and context—for example, “The five pairs had a positive correlation, r = 0.900.” Avoid saying one variable caused the other based on correlation alone.

Common Mistakes

• Pairing values incorrectly: The first x must go with the first y, the second x with the second y, and so on. Sorting one list independently changes the data relationship.
• Using different list lengths: Every observation needs a partner. Find and correct missing or extra entries before calculating.
• Mixing records or units: Keep measurements consistent within each variable. Converting every x value from centimeters to meters does not change r, but mixing centimeters and meters in the same list can distort it.
• Assuming any relationship is linear: Pearson’s r measures linear association. A strong curved pattern can still produce a coefficient close to zero.
• Overstating the meaning: Correlation does not prove that changes in one variable cause changes in the other. Other variables or coincidence may explain an association.
• Rounding input values too early: Enter the original precision where possible. Round the displayed result for reporting, not each observation before calculating.

Tips

• Keep the original data in a separate table, with clear column headings, so you can check pairings and correct typos.
• Use a scatterplot alongside r. It can reveal outliers, separate groups, and nonlinear shapes that the coefficient alone cannot show.
• Include the number of pairs when reporting a result. The same r can have very different levels of uncertainty in small and large samples.
• Remember that changing the measurement scale consistently—for example, converting all distances from miles to kilometers—does not change the correlation.

Troubleshooting

• The calculator rejects a list: Use numbers separated by commas. Remove words, units, empty entries, and trailing punctuation, then try again.
• The lists have unequal lengths: Count the observations in both lists and restore the missing paired value or remove the incomplete pair from both lists.
• The result is undefined or unavailable: Check that each list has at least two usable observations and that neither list is constant. If all x values or all y values are identical, the denominator in the formula is zero and r cannot be calculated.
• The answer is unexpectedly near zero: Verify the pair order and look at a scatterplot. The relationship may be curved, or positive and negative patterns in different groups may cancel out.
• The sign is unexpected: Check for a data-entry error or reversed pairing. Swapping the entire x and y lists should leave r unchanged; reversing the order of just one list changes the association.

Frequently Asked Questions

What is the range of a correlation coefficient?
For Pearson’s r, the range is −1 to +1, inclusive. Values outside that range usually indicate an input or calculation problem.

Can I calculate correlation with only two pairs?
Two nonconstant pairs produce r = +1 or −1, but that result is not enough to establish a reliable pattern. More observations generally make the analysis more informative.

Does a correlation of zero mean the variables are unrelated?
No. It means there is no linear association captured by Pearson’s r. A curved or more complex relationship may still exist.

Does correlation show that one variable causes the other?
No. Correlation describes an association, not causation. A third factor, selection effects, or coincidence may account for the pattern.

Should I use Pearson’s r for ranked or categorical data?
Not usually. Pearson’s r is intended for paired quantitative measurements and linear relationships. For ranks or categories, choose a measure suited to that data type.

Why does changing units not change r?
Consistently multiplying or adding a constant to every value in one variable does not change Pearson’s correlation. Inconsistent units or transformations applied to only some observations can change it.

Related

• Calculate Accuracy Between Two Numbers: Formula and Examples
• BLS CPI Calculator: Estimate Inflation Between Dates
• BP Mean Calculator: Find Your Average Blood Pressure

list x | X values | 1, 2, 3, 4, 5
list y | Y values | 2, 3, 5, 4, 6
out r | Pearson correlation coefficient | - | (count(x) * sum(x * y) - sum(x) * sum(y)) / sqrt((count(x) * sum(x ^ 2) - sum(x) ^ 2) * (count(x) * sum(y ^ 2) - sum(y) ^ 2)) | 3