I am doing fitting a straight line to data when random errors are confined to the x quantity. For the example given in the book, the data is x (2.52, 3.45, 3.46, 4.25, 4.71, 5.47, 6.61) and y (2,4,6,8,10,12,14). The question says: We can perform a least squares analysis assuming the following. (i). The errors are in the x values only. Using equations 6.34 and 6.35 we find, a* = 1.844 and b* = 0.3136. Using equations 6.37 and 6.38 we obtain a = -5.882 and b = 3.189 with equation 6.34 being the equation for the best line fit through the x-y data in this case is given when the intercept, a* is formula for linear regression and the slope b* is also found using the formula for linear regression. 6.34 is a = -a*/b* and 6.35 is b = 1/b*. I got this part of the problem correct and understand it thoroughly using a T84 calculator and number 4 linear regression stat. I do not know how to do the second part of this example. The errors of the y values only. The answer is suppose to be a= -5.348 and b = 3.067 but I do not know what formulas they are using to find it.