Correlation

The relationship between two factors such as height and weight, or color and maturity. The correlation coefficient “r” is expressed as a decimal value ranging from -1 to +1. If “r” equals zero there is no correlation and values approaching 1 + or – indicate a near perfect relationship. Cor- relations less than 0.7 are generally not practically significant to draw positive conclusions about the relationships.

The tendency of two variables to vary together.

A statistical expression of the relationship between two sets of scores, important in evaluating a test’s validity.

Degree or extent to which two measures (variables) occur together. Statisticians report a correlation (symbol r) as ranging from +1.0 (absolute direct agreement) to —1.0 (absolute inverse relationship), with .0 being an uncorrelated, fully random relationship.

The extent to which one of a pair of characteristics affects the other in a series of individuals. Such pairs of observations can be plotted as a series of points on a graph. If all the points on the resulting scatter diagram are in a straight line (which is neither horizontal nor vertical), the correlation coefficient of indicates no dependence of the one characteristic on the other of a straight line type. The regression coefficient is the average extent to which a unit increase of one characteristic influences the increase/decrease of the other. Where several factors appear to correlate with the onset of disease the relative importance of each may be calculated by the statistical technique known as multivariate analysis.

In statistics, the degree to which one variable increases or decreases with respect to another variable. A variable can have a positive or negative correlation with another variable. A positive correlation exists when the coefficient of correlation is + 1 or greater; a negative correlation exists when the coefficient is – 1 or less; a correlation is considered to be nonexistent when the value is zero.