New Quinnipiac Poll Omits Partisan Breakdown

[ericson00]

https://www.qu.edu/images/polling/ps/ps05102016_Sw4b42d.pdf

I know the story du jour is how Trump is in MoE in these swing states, the results being:

 Florida – Clinton at 43 percent, with 42 percent for Trump and Sanders at 44 percent to
Trump’s 42 percent;
 Ohio – Trump edges Clinton 43 – 39 percent, while Sanders gets 43 percent to Trump’s 41
percent;
 Pennsylvania – Clinton at 43 percent to Trump’s 42 percent, while Sanders leads Trump
47 – 41 percent.

But NOWHERE in the document is how many GOPers, Dems, and Indies were surveyed. Also, note that both Hillary and Bernie are in the margin of error with Trump, in FL and OH, which is + 3%. So Berners, don't gloat! QU has had a pattern of oversampling Republicans to make Hillary look bad.

[JaneyVee]

1. Quinnipiac is rated worst pollster of 2016 so far:

They poll Bill O'Reilly viewers.

[ericson00]

2. proof that being rated well by 538 one day doesn't mean it does well the next

[CrowCityDem]

3. I want to hear...

the explanation for how Bernie's numbers are actually great, despite being almost the same in two of the states, all without having faced ANY real attacks. You would think those unbelievable favorable ratings they're always talking about would give him more breathing room over the candidate they say is the worst we've ever fielded.

[Cheese Sandwich]

4. Shock Poll - Hillary loses Ohio to Trump

Democrats all marching off a cliff.

[woolldog]

5. You want to "unskew" the polls in other words?

[ericson00]

6. your avatar says you should at least be reading what I wrote

because most polls, unlike this one, do not omit partisan breakdown, and QU's polls last fall were always dire news for Hillary.

[CobaltBlue]

7. ericson00—The lowdown on Pennsylvania, Ohio, and Florida

Partisan-voting index:

• Pennsylvania: D+1 to D+3 when the Democratic Party wins. When the Republican Party wins, up to D+5. Status: It is a modest base state for Democrats. Pennsylvania has had a Democratic tilt in every presidential election since after the 1940s. Every Democratic winner since—John Kennedy, Lyndon Johnson, Jimmy Carter, Bill Clinton, and Barack Obama—carried the state of Pennsylvania.

• Ohio: R+1 to R+3 when either party wins. This means, a winning Democrat will carry Ohio if his/her national margin is +2 (as it was with Barack Obama’s re-election, nationally be nearly +4, in 2012; Jimmy Carter won nationally in 1976 by +2 and carried the state by less than a full percentage point). Status: The most reputable bellwether state, it has voted for every winner since 1896 (except 1944 and 1960). It will, with 100 percent certainty, vote for the winner again in 2016. Reason: The genders’ voting support for Obama, in 2012, matched the national—45 percent of males and 55 percent of females—and it was the only state to do so.

• Florida: R+2 to R+4. It’s really close to two percentage points more advantageous than Ohio is for Republicans. (An exception was, of course, the year 2000.) Democrats should win nationally by close to +4, in the U.S. Popular Vote, to carry Florida. (Obama did so with his re-election in 2012.) Status: It has been a bellwether state since 1928—having carried for every presidential winner with exceptions in the elections of 1960 and 1992.

[Armstead]

8. Elections are not counted by Party

A Republican, Democrat and Independent vote is a vote.

(Not factoring rigging, of course. that's always the wild card.)

[thesquanderer]

9. Margin of error does not mean statistically equivalent.

It means they are 95% sure the number will fall within the specified range... but moreover, not all numbers within that range are equally likely to occur. The farther from the quote number, the less likely the result, even if still within the MOE.

Good info at http://www.motherjones.com/kevin-drum/2012/10/one-last-encore-great-statistical-tie-fallacy