A Comparative Test of Magnitude Estimation and Pair-Comparison Treatment of Complete Ranks for Scaling a Small Number of Equalinterval Frequency Response Anchors