Oops! I should have mentioned that I already checked the column some time ago. I have a granite square which I placed on the bed and by moving the headstock up and down, I measured the deflection using the dial test indicator bearing against the vertical face. I measured the squareness with the granite square facing forward and then turned the square around and measured with the square facing rearward. By averaging the error, the contribution to the error of the square is zeroed out. It turned out the column is within .0006" over 5" of travel on the granite square. By minimizing the movement of the headstock when making cuts, this error is virtually eliminated.