1. A number or symbol multiplied by a variable. 2. In a regression analysis, the estimated numerical association between one variable and another, usually taken to represent the sign and size of the causal effect of one on the other.