Hi Chris, the Japanese n-gram data is unfortunately not 
publicly available. Using google hits (or preferably bing, 
as it seems to fudge the numbers less) for the word/phrase 
(in quotation marks) is an acceptable and sometimes even 
preferable substitute (the n-gram data is from 2007).
As for references, you can use free dictionary sites like 
weblio.jp, kotobank.jp and alc.co.jp, etc. for access to 
dictionaries like daijisen (daijs), daijirin (daijr), 
progressive ei-wa chujiten (prog), kenkyusha shin-ei-wa 
chujiten, eijiro (eij), etc. 
There's some more in the editorial policy, if you haven't 
seen it already. 
http://www.edrdg.org/wiki/index.php/Editorial_policy

Thanks Rene. Sorry, for the lack of references, I'm still kind of new at this and not 
quite sure of the process you guys are expecting. Looking at those dictionaries 
you're are referencing, they seem pretty expensive. Is it OK to just invest in one, 
and then use that as a reference when I have something I'd like to submit? 

I've also seen "google n-grams" and "mainichi" referenced before. Is that just a 
matter of searching for the word on the relevant site, and seeing the number of 
hits it get? Or is there something more sophisticated to it than that? I ask because 
as far as I can tell Google's n-gram viewer doesn't include a Japanese corpus 
(https://books.google.com/ngrams).

けんざかい is the only reading in gg5, meikyo, shinmeikai; main reading in daijr, daijs
koj has けんきょう only

references, please

From what I've seen けんざかい is the more common reading.

@@ -8 +8 @@
-<reb>けんきょう</reb>
+<reb>けんざかい</reb>
@@ -11 +11 @@
-<reb>けんざかい</reb>
+<reb>けんきょう</reb>