Categorial factors

Previous  Top  Next

There are factors that can not be counted or otherwise measured by numbers. They resist any meaningful quantification. Examples for this type of factors are styles of music or persons that fill a role within a workgroup. If one person is replaced by another, measuring the difference in the persons contribution to group performance is difficult to put into numbers. In contrast, there is a "distance" between alternatives when we compare e.g. different cars. The factor "fuel consumption" influencing buying decisions can be put into numbers quite easily and the difference between values has a meaning.
Values of categorial factors can be marked as being identical or different, but no gradual congruence can be stated. On the opposite, numeric factors have an obvious order of values. Moreover, the distance between input values serves as a corrollary for differences in action on target values. That is impossible when it comes to categorial factors: Values are identical or different. That leads to another observation: Categorial factors only show goal-promoting or goal-adverse values. "Neutral" values only exist if a factor is irrelevant.
The FACTORFINDER-Software is designed to solve analytical problems with categorial factors (and not numeric factors). Analytical problems with categorial factors allow differenciation into goal-promoting and goal-adverse values. Intermediate levels - as it would be possible with numeric factors - can not be assigned.

For dinner, alternatively rice, potatoes or noodles are served as side dishes. The factor "side dish" is described by categories. Each value is in the context of a specific menue (ergo a variant) advantageous or not.

A categorial attribute of an example that does not have any influence on the target criterion shall be called an attribute or a dimension of that example. It is not a factor. As soon as an attribute promotes or inhibits the realisation of the target criterion, it is a factor.