Hello. May someone teach me how to post a thread here, in a way it appears the title and the body text? Just like whuber and AndyW did in previous comments. Or is this a privilege? Thank's.

test message please ignore

@AndreSilva As I did above :) Just copy the URL from your web browser and paste it in a separate comment box (in this case, http://meta.stats.stackexchange.com/questions/995/has-any-methodological-research-begun-with-a-crossvalidated-question).

Hello gentlemen

Could you please spare me a second and tell me if there is any glaring error in this post?

@chl, got it. Thank's.

@badp Overall it looks like a good analysis, well-explained for the audience. You got the right chi-squared statistic at the outset. Removing the "outliers" and re-running the test is not justifiable but it didn't lead you astray. The correlation analysis is neither necessary nor advisable: a glance at the scatterplot says it all, anyway. If you wanted to drive home your message and make the lack of association even clearer you could start the y-axis values at zero.

@whuber Thanks. It's easy to get this stuff wrong.

Especially when it agrees with your argument and you're sort of following handouts semi-blindly