[xzins]

From ABC: "This ABC News/Washington Post poll was conducted by landline and cellular telephone June 20-23, 2016, in English and Spanish, among a random national sample of 1,001 adults, including 836 registered voters. Results have a margin of sampling error of 3.5 points, including the design effect, for the full sample, and 4 points for registered voters. Partisan divisions are 36-24-33 percent, Democrats-Republicans-independents, in the full sample, 37-27-30 among registered voters."

The weighting of democrats is a 10 point margin for democrats in this poll among registered voters.

Note also that this is about registered voters and not likely voters.

Finally, I can't find the 'N' on any other demographic. Percentages are pointless if you don't know how many people they sampled in that demographic.

Finally, they show Hillary and Trump tied among men with both at 45%. That is contrary to data everywhere else. One really wonders where they get their men.

[LS]

I disagree slightly: if the difference is 10 among registered voters, but your sample includes 20% “adults” who skew it even more D, then your actual D/R split is 11-12 +D.