This article is from the source 'guardian' and was first published or seen on . It last changed over 40 days ago and won't be checked again for changes.
You can find the current article at its original source at https://www.theguardian.com/us-news/datablog/2016/sep/28/donald-trump-poll-la-times-usc-clinton
The article has changed 4 times. There is an RSS feed of changes available.
| Version 0 | Version 1 |
|---|---|
| Trump is four percentage points ahead of Clinton – but the poll is an outlier Trump is four percentage points ahead of Clinton – but the poll is an outlier | |
| (2 months later) | |
| For the past two months, most national polls have put Democrat Hillary Clinton ahead of Republican Donald Trump. The size of that lead can vary significantly from poll to poll - she was seven percentage points ahead according to an NBC poll, then only one percentage point ahead, according to Quinnipiac University – but they still suggest Clinton is winning the majority of support. | For the past two months, most national polls have put Democrat Hillary Clinton ahead of Republican Donald Trump. The size of that lead can vary significantly from poll to poll - she was seven percentage points ahead according to an NBC poll, then only one percentage point ahead, according to Quinnipiac University – but they still suggest Clinton is winning the majority of support. |
| Numbers from LA Times and USC published on Wednesday tell a very different story, though, because they suggest Trump is four percentage points ahead of Clinton. What’s more, LA Times and USC polling have consistently suggested a more optimistic picture for Donald Trump (no wonder the candidate has tweeted their poll results). So, is this poll more or less accurate than the others? | Numbers from LA Times and USC published on Wednesday tell a very different story, though, because they suggest Trump is four percentage points ahead of Clinton. What’s more, LA Times and USC polling have consistently suggested a more optimistic picture for Donald Trump (no wonder the candidate has tweeted their poll results). So, is this poll more or less accurate than the others? |
| To understand, you have to take a peek at their methodology. | To understand, you have to take a peek at their methodology. |
| Unlike most polls, LA Times and USC tracks a panel of roughly 3,000 US citizens that were randomly contacted from across the country rather than trying to find new respondents each time. That means they have a much better sense of who is changing their minds (although it also means that panel of 3,000 people had better be a good sample of opinions, otherwise they’re stuck with it!). The results are readjusted so that they match up with demographic characteristics like race and gender from the US Census Current Population Survey. That last part isn’t surprising, basically every single poll does the same. | Unlike most polls, LA Times and USC tracks a panel of roughly 3,000 US citizens that were randomly contacted from across the country rather than trying to find new respondents each time. That means they have a much better sense of who is changing their minds (although it also means that panel of 3,000 people had better be a good sample of opinions, otherwise they’re stuck with it!). The results are readjusted so that they match up with demographic characteristics like race and gender from the US Census Current Population Survey. That last part isn’t surprising, basically every single poll does the same. |
| But the LA Times and USC poll also adjusts for something else – it takes into account how people say they voted in 2012. Many other poll watchers have pointed out that this is kind of problematic since people might not remember that accurately or else they might fib (for example, saying they voted for the winner when they didn’t). | But the LA Times and USC poll also adjusts for something else – it takes into account how people say they voted in 2012. Many other poll watchers have pointed out that this is kind of problematic since people might not remember that accurately or else they might fib (for example, saying they voted for the winner when they didn’t). |
| Personally, I’m inclined to think that the poll’s strengths are more important than its weaknesses. Rather than just asking the classic “who are you going to vote for?”, they take a more nuanced approach, by asking the following three questions: | Personally, I’m inclined to think that the poll’s strengths are more important than its weaknesses. Rather than just asking the classic “who are you going to vote for?”, they take a more nuanced approach, by asking the following three questions: |
| What is the percent chance that (1) you will vote in the presidential election? (2) you will vote for Clinton, Trump or someone else? and (3) Clinton, Trump or someone else will win? | What is the percent chance that (1) you will vote in the presidential election? (2) you will vote for Clinton, Trump or someone else? and (3) Clinton, Trump or someone else will win? |
| That means they don’t simply lump voters into groups like “voting for Trump”, “voting for Clinton”, “voting for Stein” and “undecided”. They have a much better sense of how voter attitudes (“I love this candidate”/“may as well go with them I guess”) might influence voting behavior (to show up or not show up). | That means they don’t simply lump voters into groups like “voting for Trump”, “voting for Clinton”, “voting for Stein” and “undecided”. They have a much better sense of how voter attitudes (“I love this candidate”/“may as well go with them I guess”) might influence voting behavior (to show up or not show up). |