Though the corpus includes 160,000 texts and 450 million words from a variety of sources, you could argue that some words are overrepresented due to the proportion of texts. For example, only 85 million words in the corpus are transcripts of spoken English from radio programs or television shows. That's only about 19% of the total. At the same time, another 81 million words were taken from academic journals, roughly the same percentage. I believe that's likely why there is an over representation of certain words like "education" compared to what we might guess, given that most of our guesses are probably based on spoken English and not words we'd find in academic journals.
Common English Verbs
Common English Adjectives
If you're still looking for more quizzes after that, here are a few more of my personal favorites that don't get as much traffic as some others:
Censorship Around the World
Oldest Cities by Country
Good point. My bad on the numbers not matching. There were actually 18 words on their list not included on yours, and 20 on yours that weren't on the COCA site. So I guess I counted wrong and came up with 16 instead of 18. Then also they had a "$" and "Mr." for two of their "words". Forgot to count those two as part of their 20. So 16 plus the two I missed plus their two non-words makes 20.
Not sure what you did but the site says it was last updated in the summer of 2012 and your quiz looks like it was created in April 2013. Didn't really mean to come across as upset.
Not that big of a deal though really. I'm just going to delete the other comment. Happy trails.