In the field of statistics, an antecedent (or ‘antecedent variable’) is a variable that explains the behaviour of another (subsequent) variable. Usually the antecedent is a chronologically preceding variable, as seen in auto-regressive and time-series models. In the context of simple regression, the antecedent variable would be one that would explain the behaviour of both the independent and the dependent variables.
The primary intention of using an antecedent in statistical models when applied to the field of social science is to explain the cause-effect relationship between the variables in a phenomenon. This helps to get a clearer picture of why and how that phenomenon occurs when the latter’s mechanism is not fully clear.
However, it may so happen that the relationship so explained by use of an antecedent may not be realistic, even though statistically significant with high correlation. The well-known example of rhythmic increase of ‘ice-cream sales’ and ‘level of crime’ is one such situation. So use of only antecedents and regression models is not recommended for complete explanation of any phenomenon.