A correlation determines what degree a relationship exists between two or more quantifiable variables. Its numerical value, or correlation coefficient, estimates the strength between the two variables. The purpose of conducting correlation studies is to allow us to make predictions about one variable based on what we know about another variable. One of the most common correlations is the relationship between income and education. Time and time again, researchers demonstrate that an individual's annual income increases with education.
Things You'll Need
 Problem statement
 Participants
 Instruments
 Procedure
 Graph
 Pencil
 Pen
Relationship Correlation

Write a problem statement that summarizes the predicted relationship between two or more variables.

Select a large sample of participants or test subjects.

Design a stepbystep procedure that will test the two variables.

Administer selected instruments to collect data from your test subjects.

Graph the independent variable of the y axis, and the dependent variable on the xaxis.

Determine the correlation coefficient of the graph through its slope. Note that a slope of negative one or less has a strong negative correlation, a slope of one or more has a strong positive correlation, and a graph without a slope exhibits no relationship.
Prediction Correlation

Repeat steps 13 from Section 1.

Collect data from test subjects using valid measuring instruments.

Select a predictor, such as accuracy, as a variable in addition to a constant variable.

Calculate each subject's predictor criterion using the formula y=a+bx, in which 'a' is an individual's predictor criterion, 'b' is the constant variable, and 'x' is the correlation coefficient (see Section 1, Step 6).

Graph the predictor criterion by arranging the predictor criterion on the yaxis and the constant variable (i.e. age) on the xaxis. Calculate the slope.
References
 Photo Credit Jupiterimages/Goodshoot/Getty Images