897 "likely" voters. That is a damn small sample size.
And it's a Zogby poll.
Not even realclearpolitics will include Zogby in its list of polls because they're so unreliable.
But my favorite part is how the herd cherry picks which ever poll has the most favorable ratings for trump. If they're leaning on Zogby now, that can only mean Rasmussen no longer has the highest job approval rating.
Don't you think it's highly significant, though, that Trump has high approval ratings amongst Walmart shoppers?
Not so much. Conservatives love Walmart. Nascar too. His high ratings among those two groups is in line with strong support from Republicans.
But where Zogby fails is in his desire for balance where no such balance exists. He invited thousands of volunteers to participate but included only 897 responses and weighted the results at 36% Democrat, 36% Republican and 28% Independent/unaffiliated. But that is neither representative of the public in general or how voters turned out in the last presidential election, which was 36% Democrat, 33% Republican and 31% Independent. So between over sampling Republicans and omiting everyone who's not a likely voter, and by breaking support down by strongly approve and somewhat approve, Zogby was able to show trump is at 51%, the slimmest possible majority.
But even worse for Zogby is his methodology... he uses online interactive polling which is the least accurate polling...
How accurate are online polls?
By contrast, most online polls that use participants who volunteer to take part do not have a proven record of accuracy.