Question

Without a correlation coefficient from a linear regression analysis, how might you determine the quality ( precision and accuracy) of the data?

Answer 1

One way is by visual inspection of your fit, checking if the curve really fits your date (i.e. the points stand on or near the regression line)

Another way is to input your experimental values into the equation of your regression and check if what you get is close to the value you got in your experiment. Suppose you are measuring a calibration curve, you have a set of values of concentration and the corresponding absorbance. You draw your calibration curve and you don't have the correlation coefficient. You can plug your concentration values in your regression equation and check if the absorbance values you get are coincidential with your experimental values. If they are, you are somewhat certain that your fit is good, if they don't, then there's a problem.