In statistics, a proxy variable is something that in itself does not have great interest, but from which others of great interest can be obtained. For this to be possible, the proxy variable must have a strong correlation, but not necessarily linear or positive, with the inferred value. It has no value if the data does not fit into any relation (the data is represented in a cloud of certainty). Examples
Gross domestic product per capita is often used as a proxy for measuring the standard of living or quality of life.
When processing social data collections, the gender of the respondent is an important variable. As gender commonly dictates as one responds, reading the general surveys, emphasize data on sex and age of respondents, and are used as a proxy for gender. In many general surveys, the proportion of transsexual and transgender individuals is low, making the sample and correlation reasonably good.
Also, the original country or place of birth may be used as a proxy in the classification of humans or races.
wiki
