Multiple regression can eliminate internal validity problems by potential third variables. it also sets a variable that acts as the independent variable, or the variable that will influence another variable; and a variable that acts as a dependent variable, or the variable that is hypothesized to be affected by another variable. effect sizes in multiple-regression analyses are represented by which is a standardized effect size, and which is an unstandardized effect size.