Let's say we'd like to assess the relative frequency of the same way versus in the same way.

This wouldn't yield a true frequency comparison: (the same way),(in the same way) because (the same way) would return true when it encountered "...in the same way..."

Is there something like (-in the same way) -- meaning 'the same way' not immediately preceded by 'in'?

  • 2
    the same way,the same way-in the same way ... Or you could express it as a proportion: in the same way/the same way Mar 15, 2015 at 13:06
  • So this query the same way-in the same way would subtract the {in the same way} subset from the {the same way} set?
    – TimR
    Mar 15, 2015 at 13:19
  • @TRomano - Yes, you can do some elementary "algebra" in Ngrams. Have a look.
    – J.R. Mod
    Mar 15, 2015 at 19:40
  • 1
    In this case, actually, you have to bracket the terms as (the same way)-(in the same way). There are limits on the number of tokens you can treat. Mar 15, 2015 at 22:31


You must log in to answer this question.

Browse other questions tagged .