Which of the following is true of the fundamentals of regression analysis? A fundamental basis of regression analysis is the assumption of a circular relationship between the independent and dependent variables. The differences between actual and predicted values of the dependent variable are known as regression coefficients and are represented by b. Any point that falls on the line of a regression analysis is the result of unexplained variance. The regression coefficient is calculated by squaring errors of each dependent variable. Regression uses an estimation procedure called ordinary least squares that guarantees the line it estimates will be the best fitting line.