Follow
In science we study numbers much much much smaller, and when doing so we compare relative difference.
So in this case let's pretend we are discussing specifically voting against party lines of the two parties in the impeachment trial. We ignore all the numbers not related to this study, 99.3% of the data is not within the scope of our study.
All we are left with is 2% of republicans did not vote YES. 3.9% of democrats did not vote YES. The difference is certainly statistically significant: democrats are 51% more likely to vote outside of party lines than republicans. That's more than double.
@ae Fact remains Not voting is not voting against party lines first off, so stop trying to stretch what little numbers you can, thats not helping you look objective here.
Second you can spin it however you want, you wont convince me that when 99%+ of a party is voting party lines that they are unbias.
Next "statistically significant" just means "unlikely to be random".. So all your saying is that the marginal and insufficient less than 1% difference is probably not just random error.. I mean ok.. they still have 0 credibility if all they have is a marginally insignificant 2 people voting no..
That said we are still about 4 levels deep into meta arguments on this that largely miss the main point anyway.