USC Daybreak Poll - Change in methodology?

I can't remember if it was 2008 or 2012, but the rolling Rasmussen poll changed it's demographic model midstream. It was noticed on FR first as many people were carefully tracking its day to day polling. No one could figure out what happened. Then a few days later, Rasmussen issued a statement explaining they had adjusted their model. So, it does happen.

If that didn't happen in this case, and a seven day rolling average shows the very significant change reflected in the last two days of this poll, than we should expect to see even more wild shifts in all the other polls out there. We'll know in about 7 days.

What makes you think it will "Start to take hold"? Nobody in the MSM is talking about it. When and if they do it's to defend Clinton. Not really sure Trump cares, actually.

What made anyone believe Trump was going to continue to be given positive press by the MSN, once the conventions were over, and he secured the nomination? It was never a question if Hillary would get it from the DNCC. Sanders was rewarded quite handsomely in his stooge run for the nomination.

I guess that's why sound advice is to never believe you own press.

I think the ‘piling on effect’ is more prominent in rolling average polling. As Trump has made more gaffes and the MSM has all piled on along with the “never Trump” and anti-Trump GOP office holders and the national security experts and the ....blah blah blah, the cumulative effect is pretty obvious as the uncommitted voters waver or turn.

It’s very difficult to watch Trump these days without feeling sorry and exasperated at the opportunity lost.

Why do you think the piling on effect would be more prominent in a rolling average poll? The supposed advantage of a rolling average poll is to tamp down some of the spikes and dips caused by the news cycle. This poll is a seven day average. It should be relatively stable with gradual climbs and descents indicating a week long shift in voter sentiment. Simple math would indicate that to get a very large shift in the published daily number, you'd need a 20-30% difference between the day that fell off the weekly average and the day that was added. In politics, that simply doesn't happen.

I'm not at all arguing that Trump is doing great right now and this poll should be showing him ahead. After watching polls for 8 election cycles now, I have stopped using any of them as a predictor of the actual state of the race on the day they are published. People say "well, they got the 2012 election right". But that isn't entirely true. The RCP average had Obama ahead by .7 on Election Day (or something close to that). He actually won by 3.7ish%. That is significantly off considering the RCP is an average of multiple polls and should have reduced the amount of error inherent in individual polls. Again, based on previous experience with rolling average polls, it is not unprecedented for them to change their demographics as they poll. Rasmussen did it in previous elections. Reuters has already done it in this election. I suspect USC had done it now as well. I wouldn't accuse them of "rigging" their poll. I'd just guess they are trying to fine tune it to meet other data they are collecting.

The latest Reuters poll used 46% democrats and 30% Republicans. It didn’t give the breakdown as to the ratio of likely voters, but it is probably similar. Hillary could only manage 42%, whereas Trump got 36%. Hillary tends to poll 4 to 6% lower than the percentage of democrats in the poll. Trump consistently wins the independent vote. Once and awhile they throw democrats into to the independent bin to make it look better, but it is usually obvious. I think that Trump is actually running a little over 50% at this time. That is why the media is so apoplectic. If Hillary was really significantly ahead they would not be as over the top extreme.

“The media has never been this over the top. No matter where you go — Yahoo, Facebook, Google, Apple, National Newspapers, Local Newspapers, National TV News, Local TV News, Infotainment TV, etc - it’s 100% anti Trump.”

It’s proof that we already live in an Orwellian society. Big Brother has already decreed what we are all to supposed to believe about Trump.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.