When analyzing the relationship in non-linear data as a linear association, when should the strategy of ignoring outliers be used? If ignoring half of the data points allows the remaining data to be represented by a straight line. If ignoring one or two data points allows the remaining data to be represented by a straight line. If ignoring half of the data points allows the remaining data to be represented by a curved line. If ignoring one or two data points allows the remaining data to be represented by a curved line