The Public Interest Research Group (PIRG) surveyed 197 credit reports and concluded that one in four contained serious errors. However, the survey's results were called into question: "Besides objecting to the way the survey was conducted, with PIRG surveying its own members, Norm Magnuson of the Consumer Data Industry Association protested that the group ‘unilaterally decided what is a serious error' in presenting its findings."3 Is Magnuson suggesting there is bias because the individuals are not representative of the larger population, or bias in the assessment of variables' values, or both, or neither?