An independent variable (also commonly referred to as a 'predictor variable', an 'explanatory variable', an 'input variable', or a 'regressor') is a type of variable used in statistical modelling, the other type being a dependent variable (also commonly referred to as an 'outcome variable', an 'experimental variable' a 'regressand', or an 'explained variable'). Statistical models examine how dependent variables depend on independent variables, or in other words, they look at the extent to which the values of a dependent variable can be explained by the values of independent variables. As an example, one might want to understand the extent to which life expectancy can be predicted by factors such as diet, alcohol intake, and number of cigarettes smoked. In this case, age at death would constitute the dependent variable, and measures of diet, alcohol intake, and cigarettes smoked the independent variables. Independent variables can be either manipulable (for instance, a researcher could experimentally vary a stimulus a respondent sees to understand the effect on attitudes), or non-manipulable (for instance, age and gender).
This entry is called "Independent variable (aka Causal Variable)" but this is misleading because an independent variable can only really be said to be causal under experimental conditions (or if there is some sort of logical or chronological relationship which means it must be causal), and not where it refers to an observed variable in a model (eg age, gender, ethnicity etc) which is the case in much of social research. Therefore, this definition relates to "Independent variables" rather than causal variables.