By adding linear, quadratic, or cubic regression lines to your scatter plots, you can see which model best seems to fit your data. Note that the implication of a regression line is that the variable on the Y (vertical) axis is the dependent or response variable, while the variable on the X (horizontal) axis is the independent or predictor variable.
These options are available only if all variables in your scatter plots are numeric. The regression lines are not forced through plot origins.