Correlation Calculators
This page presents tools that summarise numerical or categorical relationships, providing measures that indicate direction and strength across different data types.
-
Cramer's V Calculator
A measure based on a contingency table shows how strongly two categorical variables are associated.
Example use: Checking whether snack choices differ between groups at a small gathering.
Inputs: contingency table data
Outputs: grand total, degrees of freedom, chi-square value, cramer's v
Visual: grouped bars showing how frequently each category appears across the columns
-
Kendall's Tau Calculator
A rank-based measure compares the ordering of two variables to show how consistently they increase or decrease together.
Example use: Comparing the order in which two people rate the brightness of several lamps.
Inputs: data 1, data 2
Outputs: sample size, kendall's tau, concordant pairs, discordant pairs, total possible pairs
Visual: a scatter of paired observations with a straight trend line showing the overall direction
-
Partial Correlation Calculator
A relationship between two variables is assessed while holding a third variable constant to show how much of their link remains after adjustment.
Example use: Checking whether the time spent reading and the time spent writing are related once daily free time is accounted for.
Inputs: variable x, variable y, control variable z
Outputs: pearson correlation between x and y, pearson correlation between x and z, pearson correlation between y and z, partial correlation, interpretation
Visual: a scatter display comparing the three variables and showing how their relationships differ
-
Pearson Correlation Calculator
A linear relationship between two continuous variables shows how strongly they rise or fall together around a straight-line trend.
Example use: Comparing the number of minutes spent walking with the number of minutes spent outdoors on different days.
Inputs: dataset 1 variable, dataset 2 variable
Outputs: sample size, mean of dataset 1, mean of dataset 2, standard deviation of dataset 1, standard deviation of dataset 2, covariance, pearson correlation, coefficient of determination, regression line
Visual: a scatter of data points with a straight line showing the long-term trend
-
Point-Biserial Correlation Calculator
A relationship between a binary variable and a continuous variable shows how the average value differs between the two groups.
Example use: Comparing the reading times of people who choose tea versus those who choose water during a break.
Inputs: binary variable, continuous variable, standard deviation type
Outputs: point-biserial correlation, mean of group 1, mean of group 0, total mean, standard deviation, group 1 size, group 0 size, total sample size
Visual: a scatter display showing each group's values along a binary axis
-
Polychoric Correlation Calculator
A relationship between two ordinal variables estimates the strength of an underlying continuous association that the ordered categories reflect.
Example use: Comparing how two people rank the comfort of several chairs using ordered categories.
Inputs: ordinal dataset x, ordinal dataset y
Outputs: sample size, categories in x, categories in y, estimated polychoric correlation, pearson approximation
Visual: a scatter of observations placed according to their ordinal positions on both axes
-
Spearman Rank Correlation Calculator
A rank-based measure compares how consistently two variables move together once their values are replaced by their ordered positions.
Example use: Checking whether the order in which someone listens to songs matches the order in which they rate their enjoyment.
Inputs: data 1, data 2
Outputs: sample size, sum of squared rank differences, spearman rank correlation, coefficient of determination, mean of data 1, mean of data 2
Visual: a scatter of ranked pairs with a straight trend line showing the overall direction
Correlation Calculators FAQs
Ranked data often aligns with Spearman or Kendall measures, both reflecting directional consistency rather than strict linear movement.
Pearson summarises linear relationships between numerical values, whereas Spearman evaluates ranked positions to capture monotonic patterns.
Partial correlation isolates the direct link between two variables by removing variation explained by additional influencing variables.