MEASURES OF ASSOCIATION

Statistics that are used to quantify the strength and nature of the relationship between two variables in a cross-tabulation are called measures of association . Many different measures of association exist because "association" can be defined in many different ways. The measures differ in how they can be interpreted and in how they define perfect and intermediate levels of association. They also differ in the level of measurement required for the variables. For example, if two variables are measured on an ordinal scale, it makes sense to talk about their values increasing or decreasing together. Such a statement would be meaningless for variables measured on a nominal scale.

No single measure of association is best for all situations. To choose the best one for a particular situation, you must consider the type of data and the way you want to define association. If a certain measure has a low value for a table, this does not necessarily mean that the two variables are unrelated. It can also mean that they are not related in the way that the measure can detect. But you should not calculate a lot of measures and then report only the largest. Select the appropriate measures in advance. If you look at enough different measures, you increase your chance of finding significant associations in the sample that do not exist in the population.