A correlation coefficient is used in statistics to describe a pattern or relationship between two variables. A negative correlation describes the extent to which two variables move in opposite directions. For example, for two variables, X and Y, an increase in X is associated with a decrease in Y. A negative correlation coefficient is also referred to as an inverse correlation. Correlation relationships are graphed in scatterplots.

Negative Versus Positive Correlation

A negative correlation demonstrates a connection between two variables in the same way as a positive correlation coefficient, and the relative strengths are the same. In other words, a correlation coefficient of 0.85 shows the same strength as a correlation coefficient of -0.85.

Correlation coefficients are always values between -1 and 1, where -1 shows a perfect, linear negative correlation, and 1 shows a perfect, linear positive correlation. The list below shows what different correlation coefficient values indicate:

Exactly 1. A perfect negative (downward sloping) linear relationship

0.70. A strong negative (downward sloping) linear relationship

0.50. A moderate negative (downhill sloping) relationship

0.30. A weak negative (downhill sloping) linear relationship

0. No linear relationship

+0.30. A weak positive (upward sloping) linear relationship

+0.50. A moderate positive (upward sloping) linear relationship

+0.70. A strong positive (upward sloping) linear relationship

Exactly +1. A perfect positive (upward sloping) linear relationship

Another way of thinking about the numeric value of a correlation coefficient is as a percentage. A 20% move higher for variable X would equate to a 20% move lower for variable Y.

Extreme Correlation Coefficients

A correlation coefficient of zero, or close to zero, shows no meaningful relationship between variables. In reality, these numbers are rarely seen, as perfectly linear relationships are rare.

An example of a strong negative correlation would be -.97 whereby the variables would move in opposite directions in a nearly identical move. As the numbers approach 1 or -1, the values demonstrate the strength of a relationship; for example, 0.92 or -0.97 would show, respectively, a strong positive and negative correlation.

Examples of Positive and Negative Correlation Coefficients

For example, as the temperature increases outside, the amount of snowfall decreases; this shows a negative correlation and would, by extension, have a negative correlation coefficient.

A positive correlation coefficient would be the relationship between temperature and ice cream sales; as temperature increases, so too do ice cream sales. This relationship would have a positive correlation coefficient. A relationship with a correlation coefficient of zero, or very close to zero, might be temperature and fast food sales (assuming there's zero correlation for illustrative purposes) because temperature typically has no bearing on whether people consume fast food.

Bottom Line

A negative correlation can indicate a strong relationship or a weak relationship. Many people think that a correlation of –1 indicates no relationship. But the opposite is true. A correlation of -1 indicates a near perfect relationship along a straight line, which is the strongest relationship possible. The minus sign simply indicates that the line slopes downwards, and it is a negative relationship.