At the beginning of the discussion of cost estimating we issued a warning that accounting magnitudes may not correspond to economic measures. However, all the regression analyses just reviewed employed accounting data and, through various adjustment procedures, attempted to make the numbers correspond more closely to the cost concepts meaningful to the economist. In addition, economists have tried to use methods that are not dependent on numbers drawn from companies’ accounting records. Two techniques of analysis are discussed briefly: the engineering cost method and the survivorship technique.