Up to now, therefore, we must conclude that all frequency Amateur Amy counts are essentially subjective, since the samples on which they are based are subjectively selected. This can be shown quite easily by comparing the results achieved for the same item in different published accounts. For example, Teen sex site places the words approve and contract in the first 500 of most frequent words, whereas Thorndike places them in the fourth and fifth 500 words respectively, whilst Horn places bread in the fifth 500 and Thorndike in the first 500. As with all quantitative statistical studies the results are as valid as the sampling procedure is valid. Up till now all statistical studies of lan­guage which aim to make general statements about language use are invalid. It follows, therefore, that if relative frequency calculations are to be any use for language teaching they must be based upon data which represent the same range of language use as that aimed at by the learner. There is, however, another snag. A particular Amateur Amy item may receive a high count either because it is found with a general low frequency but distributed evenly throughout component texts of the sample, or because it occurs with very high frequency in one component text of the data but not at all in the other components. In the first case we say that such an item has a wide 'range'. It is not, therefore, a marker of any variety, whereas in the second case the item has a restricted range and is a marker of one or another variety.  

Amateur Amy menu: gallery1 gallery2 gallery3 gallery4 gallery5