Free Correlation Coefficient Calculator

What Is a Correlation Coefficient?

A correlation coefficient is a normalized measure (ranging from –1 to +1) that captures the strength and direction of the relationship between two random variables. A value of ±1 indicates the strongest possible association, while 0 suggests no linear or monotonic link. The free online Correlation Coefficient Calculator lets you compute four of the most popular coefficients: Pearson (Pearson $r$), Spearman rank, Kendall tau (including tau‑b and tau‑c), and Matthews correlation (MCC, also known as Pearson phi). Whether you need to find the correlation coefficient for a homework problem or a research project, this tool provides both the final value and the intermediate calculation steps.

Though correlation coefficients are standardized effect‑size metrics, they are distinct from measures like Cohen's d, which focus on group differences.

How to Use the Calculator

1. Choose the coefficient – Select Pearson, Spearman, Kendall, or Matthews from the dropdown menu.
2. Enter your data – Specify how many sample points you have and fill in the $x$ and $y$ coordinates. For the Matthews coefficient, you will instead enter the four values of a 2×2 confusion matrix.
3. Review the result – As soon as you have at least three data pairs (for Pearson/Spearman/Kendall), the tool displays the coefficient. For Pearson, it also shows the coefficient of determination ($R^2$).
4. Show calculation details – Activate this option to see every intermediate sum, product, and deviation that leads to the final coefficient.
5. Interpret the strength – The calculator applies Evan's (1996) scale based on the absolute value of the coefficient:

Absolute value range	Strength
0.8 – 1.0	Very strong
0.6 – 0.8	Strong
0.4 – 0.6	Moderate
0.2 – 0.4	Weak
0.0 – 0.2	Very weak

Pearson Correlation (Pearson $r$)

The Pearson correlation coefficient evaluates the linear relationship between two continuous variables. It is defined as the covariance of $X$ and $Y$ divided by the product of their standard deviations:

$$r = \frac{\sum_{i=1}^{n} (x_i - \bar{x})(y_i - \bar{y})} {\sqrt{\sum_{i=1}^{n} (x_i - \bar{x})^2 \;\sum_{i=1}^{n} (y_i - \bar{y})^2}}$$

Where $\bar{x}$ and $\bar{y}$ are the sample means of $X$ and $Y$, respectively. A low Pearson $r$ does not mean the variables are unrelated—it only indicates that there is no linear pattern. A strong non‑linear (e.g., quadratic or exponential) relationship may still exist. In simple linear regression $Y = aX + b$, the square of the Pearson coefficient equals the coefficient of determination $R^2$, which tells you the proportion of variance in $Y$ explained by $X$.

Spearman Rank Correlation

Spearman's correlation is essentially the Pearson coefficient applied to the ranks of the data. To obtain ranks, sort each sample separately from the smallest to the largest observation, giving rank 1 to the minimum value. Ties are resolved by assigning the average of the ranks that the tied values would have received if they were all distinct.

$$r_s = \frac{\sum_{i=1}^{n} (R_i - \bar{R})(S_i - \bar{S})} {\sqrt{\sum_{i=1}^{n} (R_i - \bar{R})^2 \sum_{i=1}^{n} (S_i - \bar{S})^2}}$$

where $R_i$ and $S_i$ are the ranks of $x_i$ and $y_i$, respectively. Because it works on ranks, the Spearman coefficient detects any monotonic relationship (linear or not), making it more flexible than Pearson.

Kendall Rank Correlation (Tau)

Kendall's $\tau$ measures association by comparing every possible pair of observations. For two pairs $(x_i, y_i)$ and $(x_j, y_j)$:

• Concordant if $x_i < x_j$ and $y_i < y_j$ (or both inequalities are reversed).
• Discordant if $x_i < x_j$ but $y_i > y_j$ (or vice versa).

When there are no ties, the basic coefficient (tau‑a) is:

$$\tau_a = \frac{C - D}{C + D}$$

with $C$ = number of concordant pairs and $D$ = number of discordant pairs.

If ties exist, two adjusted variants are available:

• Tau‑b: suitable when both variables have the same number of possible categories (e.g., a square contingency table). It corrects for ties in $X$ and $Y$:

$$\tau_b = \frac{C - D}{\sqrt{(C + D + T_x)(C + D + T_y)}}$$

where $T_x$ is the number of pairs tied only on $X$ and $T_y$ the number tied only on $Y$.

• Tau‑c (Stuart–Kendall): used when the contingency table is rectangular (different numbers of rows $r$ and columns $c$):

$$\tau_c = \frac{C - D}{\frac{1}{2} n^2 \frac{m-1}{m}}, \quad m = \min(r,c)$$

Matthews Correlation (MCC / Pearson Phi)

The Matthews correlation is designed for binary classification tasks. It is common in machine learning and biomedical studies. Data are arranged in a 2×2 confusion matrix:

	Predicted Positive	Predicted Negative
Actual Positive	TP	FN
Actual Negative	FP	TN

The MCC formula is:

$$MCC = \frac{TP \times TN - FP \times FN} {\sqrt{(TP + FP)(TP + FN)(TN + FP)(TN + FN)}}$$

Interpretation differs slightly from the other coefficients:

• +1 → perfect prediction,
• 0 → no useful information,
• –1 → complete disagreement between prediction and actual outcome.

Manual Example: Computing Pearson $r$ by Hand

Suppose you have two small samples:

$X = \{2,\; 7.5,\; 3,\; 34.2,\; 26\}$
$Y = \{21,\; 12.5,\; 3,\; 11,\; 17\}$

1. Compute the means:
   $\bar{x} = (2 + 7.5 + 3 + 34.2 + 26)/5 = 14.54$
   $\bar{y} = (21 + 12.5 + 3 + 11 + 17)/5 = 12.90$

2. Fill in the deviation table:

$x$	$y$	$x - \bar{x}$	$y - \bar{y}$	$(x-\bar{x})^2$	$(y-\bar{y})^2$	$(x-\bar{x})(y-\bar{y})$
2	21	–12.54	8.10	157.252	65.610	–101.574
7.5	12.5	–7.04	–0.40	49.562	0.160	2.816
3	3	–11.54	–9.90	133.172	98.010	114.246
34.2	11	19.66	–1.90	386.516	3.610	–37.354
26	17	11.46	4.10	131.332	16.810	46.986
Sum				857.832	184.200	25.120

3. Plug into the Pearson formula:

$$r = \frac{25.120}{\sqrt{857.832 \times 184.200}} \approx 0.063$$

The result ($r \approx 0.063$) falls within the expected –1 to +1 range and, according to Evan's scale, indicates a very weak linear relationship. You can verify each step by entering the same data into the correlation coefficient calculator and enabling the calculation details.