How to Create a Contingency Table

Contingency tables help analyze data.
••• Hemera Technologies/ Images

When evaluating the relationship between two or more different items or variables from an experiment, use a contingency table. This table allows for an at-a-glance analysis of observations between the variables. The most common type of contingency table is commonly referred to as the 2x2 or 2 row and 2 column contingency table, but can have as many rows and columns as needed for variables to be evaluated.

    Start with two outcomes. In this example, we will use pass versus fail. These are the columns in the table.

    Define the group variables. In our example, these will be the classes. They will be the rows for the table.

    Class 1 A B Class 2 C D

    Input the numbers. Instead of A, B, C and D in our example, we'll use some fictional numbers of students who passed and failed tests. In a real contingency table, numbers will be used, not variables.

    Class 1 13 7 Class 2 19 1

    Tally both ends. This is called "two-tallied."

    Class 1 13 7 20 Class 2 19 1 20 Total 32 8 40

    Compute the P-value. The formula is A/(A+B) - C/(C+D).


    • If the P-value is very small, it is statistically significant and not randomly occurring.

Related Articles

How to Calculate Odds Ratio on a Contingency Table
How to Interpret an Independent T Test in SPSS
How to Report Z-Score Results
How to Calculate the Mean in a Probability Distribution
How to Find Quadratic Equations From a Table
Advantages & Disadvantages of Using Math Tables
How to Find the Midpoint of the Interval
How to Calculate Logit
How to Interpret a Student's T-Test Results
How to Calculate Conditional Probabilities
How to Calculate Regression Coefficient
How to Calculate Correlation Coefficients With an Equation
How to Know if Something Is Significant Using SPSS
How to Do Function Tables in 6th Grade Math
How to Calculate Unexplained Variance
How to Solve a Matrix
How to Calculate Variance From a Ti84
Similarities of Univariate & Multivariate Statistical...
How to Approximate the Mean of Group Data