This paper presents a comparison of unsupervised methods of hypernymy\nprediction (i.e., to predict which word in a pair of words such as fish-cod is\nthe hypernym and which the hyponym). Most importantly, we demonstrate across\ndatasets for English and for German that the predictions of three methods\n(WeedsPrec, invCL, SLQS Row) strongly overlap and are highly correlated with\nfrequency-based predictions. In contrast, the second-order method SLQS shows an\noverall lower accuracy but makes correct predictions where the others go wrong.\nOur study once more confirms the general need to check the frequency bias of a\ncomputational method in order to identify frequency-(un)related effects.\n