Word frequency data

The differences between the iWeb and the COCA frequency lists are due to the composition of each corpus. COCA is balanced between spoken, fiction, popular magazines, newspapers, and academic journals. iWeb, on the other hand, comes just from web pages (but web pages dealing with almost every imaginable topic).

The two lists below give some examples of what types of words are more frequent in each corpus. In the 60,000 word list, there are 1,522 words (like those to the left), where the word is much more frequent in iWeb than in COCA. (The rank order in COCA is at least twice that of iWeb, e.g. word #1010 in iWeb, but word #2103 in COCA). And there are 672 words (like those on the right), where the word is much more frequent in COCA (same calculation as above).

Notice that many of the iWeb words deal with technology and commerce (e.g. update, blog, battery, provider, submit, offer, delivery, purchase), while those from COCA come either from fiction (e.g. sing, stare, cry, smile, nod, kiss) or news (violence, debate, reform, troop, nuclear).

In addition, there are about 1,050 words in the iWeb top 60,000 list that are not in the COCA top 60,000 list at all. Some of these are non-American words or spellings (centre, organisation, flat, tyre, kilometre, tonne, litre), since COCA is just from the US (but iWeb comes from six English-speaking countries). But the majority of these words relate to technology, since iWeb comes from the Web: e.g. IT, email, LED, CD, ipad, IP, smartphone, plugin, USB, AC, google, SQL, GPS, API, screenshot, blog, AI, byte, linux, volt, LCD, SEO, javascript, wifi, FM, webinar.