What's Nuisance Variables ? (Applied Econometrics)

   NUISANCE VRIABLES

Generally in the experimental design we tend to consider only about the dependent and independent variables. We generally tend to neglect the concept of the nuisance variables. The most crucial part of the experimental design is the nuisance variable.

Say that, generally it is seen that the dependent variable readily depends upon on an indefinable property of the subjects, say age of any individual. So, this property would become the nuisance variable. We can set an assumption by designing the experiment that whether how much the nuisance variables affects the dependent variable.

We can also call that in the theory of the stochastic processes, say in the probability theory of the statistics, a nuisance variable is a random variable that is a fundamental to the probabilistic model, but that is of no particular interest in itself or is no longer of the interest.

One of its usage come from the Chapman Kolmogorov equation. Say, a model for a stochastic process which may be defined conceptually using the intermediate variables that are not observed in the practice. If there is the problem to derive the theoretical properties, such as mean, variance and co variances of the quantities which would be observed, then those intermediate variables are nuisance variables.

Basically, in the block experiments, where the terms in the particular model representing the block means, those are often called as "factors", are of no interest. Several approaches to the analysis of such experiments particularly where the  experimental design is the subject to the randomization, we must treat these factors as the random variables.

Furthermore, the nuisance variable is used simply to designate those variables that are marginalized over when finding a marginal distribution. Sometimes, this term is also used in the Bayesian analysis as an alternative to the nuisance parameter, provided that the Bayesian statistics allows parameters to be treated as having the probability distributions.

Hope this adds some knowledge in the experimental design of your research. 

Thank You

Aditya