Hot answers tagged

169

In comparison to the 538 model, the economist model uses less "fat tails". 538 uses a t-distribution to account for "black swan" events: things that, although they are unlikely, would have a big impact on the polls. This means that 538 assigns a small probability to some very unlikely outcomes (Trump wins California, or Biden wins Utah) ...


81

One purpose of the type of poll that you linked to is to develop a list of potential donors and campaign volunteers and supporters. The one-sided nature of the poll language serves to a) motivate the potential reader to respond and b) screen out anyone likely to disagree or be disinterested. Since the poll form then asks the submitter to include their name, ...


57

Fully-specified polls are perfectly trustworthy in terms of what polls actually indicate. Their topline results are not very reliable. Most people think of polls in various incorrect ways. The most common (and least meaningful) way is as a straight prediction of how people will vote on election day. Polls almost never try to indicate this. Even when hacks ...


48

To demonstrate the difference, I'm going to use the result of the 2008 election as an example. A close race is a race where the difference between the two candidates is small. In 2008, Missouri was the closest race (in percentage point terms), which McCain won by 0.14%. As we know, though, Obama didn't need Missouri to win that election. The tipping point ...


37

I tend to group polls into one of two types: scientific polls that intend to reflect accurately what the population or whichever subgroup it is interested in thinks/believes/desires/knows proof-of-myself polls that intend to show support for whatever my political/sales/religious message is. Of course, they are sometimes hard to distinguish but a lot of ...


33

One term for this effect is "Social Desirability bias" Social desirability bias is a social science research term that describes the tendency of survey respondents to answer questions in a manner that will be viewed favorably by others. It can take the form of over-reporting "good behavior" or under-reporting "bad", or ...


30

Don't mistake polled sentiments for election results. In all likelihood, the next presidential election in the US will be decided by a few swing states. It will not matter how much the Democrats win in a deep blue state. It will not matter how much the Republicans win in a deep red state. All that matters is who wins states like Florida, Ohio, or Virginia (...


28

The survey/poll participants not being all equally likely to answer is a well-known problem called participation bias aka non-response bias. A typical example: A study of nonrespondents from the National AIDS Behavioral Survey (NABS) was conducted in 1990 to attempt to identify potential differences in participants and non‐participants that may influence ...


28

There are some nuances particular to FiveThirtyEight's model. Their dashboard works by simulating a large number of elections, and then presents a sample of 100 of these simulations. The states with the closest races are those where the number of simulations each candidate is predicted to win the state is closest to 50 each. Examples are North Carolina (51 ...


20

Depending on how you became aware of the poll, this may be an example of a push poll. This is a poll where there is no intention of doing anything useful with the results, but where the true goal of the poll is to influence people by asking them loaded questions. A push poll is an interactive marketing technique, most commonly employed during political ...


17

If you look at the date ranges on those "most recent polls" (at the time this question was asked) you will find that they almost all started before July 9th. The Hill/Harris X poll started July 12th. It includes Steyer (0%, rounded). The Economist/YouGov poll started July 14th. It includes Steyer (1%). The premise that Steyer is not included in polls that ...


15

In addition to the caveat about the Electoral College, polls can change pretty fast. Just on Oct 30, another predicted Trump would win the popular vote, against an unnamed Democratic nominee, by a narrow margin. Even more noteworthy, the subjects gave an even higher rating to Trump's chance of actually winning the race (probably because people are aware of ...


12

What's going on is that there are two weekly polls of the Democratic race, one by Politico/Morning Consult and one by Economist/YouGov. The Economist/YouGov poll shows consistently lower support for Biden than other polls of the race. YouGov's last six polls found Biden's support at 25, 26, 25, 25, 22, and 23. By contrast, the last six Morning Consult polls ...


11

Defining an opinion poll as "an assessment of public opinion obtained by questioning a representative sample," the answer to your question is no. Opinion polls are not always overly representative of extreme opinions. If you were to poll 20 people in your company (selected randomly) and they all respond, your poll would not be overly representative of ...


10

A common problem when trying to handicap the Republican race is to place too much reliance on a single statistic: first place votes for each candidate. The problem that Cruz has is that he doesn't do well in other statistics. His favorability is lousy. He has no endorsements from other Senators nor from governors. He is low on most second choice charts. ...


9

This applies to all UK elections. The Representation of the People Act 2002 inserted a section 66A into the initial 1983 of the same name and this read as follows. It is a criminal offence ‘to publish, before a poll is closed, any statement about the way in which voters have voted in that election, where this statement is, or might reasonably be taken to ...


9

Note: links point to websites in italian language. I don't know of any country that forbids polls for a few months, but there is a country, Italy, where polls cannot be published for the last 15 days preceeding the election. The rationale is that forbidding opinions for months would be detrimental to the development of public opinion and the law is already ...


8

One basic notion I was taught when learning about survey methods is that serious surveys ask about something the respondents already have an opinion about. If you invite them to make up an opinion about some topic they don't have any firmly held beliefs or never thought about before, many will do so on the spot but the answers will be fickle and much more ...


8

It really depends on the pollster. Generally, a random sample is taken of the population, using methods such as random telephone number generation, and this sample is then weighted to correct for demographic size and sampling issues. Gallup publishes information about how it samples the population for its weekly U.S. Poll. In this case, the sampling is done ...


7

Election polls definitely affect the election results, but the extent is debatable. Part 1: Study - Do Polls Influence the Vote? This study analyses how do polls influence voters' decisions by examining the impact of polls on the 1988 Canadian election. The results shown in panel A of table 1 confirm that voters did respond to the information provided by ...


7

Real Clear Politics is only listing thirteen of the twenty odd major candidates in their interface. Presumably this is because they lack the horizontal room on the page to show more. This leaves off eight of the candidates who made it to the first debate. If you look down further, the graph shows twenty candidates and Tom Steyer is included there (he ...


7

I don't see any explanation on their Full Methodology Info page, but when percentages don't add up to 100%, the usual answer is that it's due to rounding, so 11% might actually be 11.4%, which was rounded down to 11. Among 4 options, rounding could explain almost all of the missing 2%. Another possibility, is that these 4 options don't represent all ...


7

Although an asterisk usually indicates that a poll is conducted by - or on behalf of - a partisan source, in the case of the RealClearPolitics tracker, this denotes that the polling includes figures for third-party candidates which have been truncated to provide the head-to-head two-party figure. For example, looking at the most recent Trump vs. Biden ...


6

Brexit polling Bloomberg tracked poll results on the Brexit vote. The final value before the election (on June 22nd, 2016) was Remain 46.2% Leave 44.3% with 9.6% undecided (doesn't add to 100% due to rounding). That's a 1.9% margin, well within the margin of error. The actual election was 3.78% in favor of leaving. That's a 5.7% swing. That's large but ...


6

The numbers you're comparing are not "poll results", they're odds of winning based on the polling. The details of their models are spelled out here. They summarize them as: Polls-plus: Combines polls with an economic index. Since the economic index implies that this election should be a tossup, it assumes the race will tighten somewhat. Polls-...


6

In many cases, it appears that major news outlets (CNN, FOX, etc.) are conducting their own polls. Is this correct? Not necessarily. As a general rule, they partner with some polling organization that performs the actual poll. For example, CNN polls often appear as CNN/ORC because they partner with the Opinion Research Council. Partnering in this case ...


6

Per the site's own presentation, its predictions are based mostly on prediction markets and indeed, those appear to put Marco Rubio (or even Jeb Bush, who just withdrew from the race!) way ahead of Ted Cruz so there is really nothing idiosyncratic about PredictWise. And prediction markets are based on multiple bets, not on top secret information or anyone's “...


6

This is just a random twitter poll where anyone can vote regardless of being able to vote regardless of their actual voting status for the election. There is nothing that stops anyone from voting especially bots and the voters are all self selected which means the votes are not reliable. The other poll you mentioned has procedures in place to make sure they ...


6

There are a lot of assumptions baked into this behavior and it may not always achieve the intended effect. Firstly, a single person doing this is unlikely to skew the polling result. Most respectable pollsters strive to get a large and randomized group of participants precisely to diminish outliers cases like these. Unless you organize a massive coordinated ...


5

Shy Trump factor/Clinton effect First, it remains uncertain that there was a shy Tory factor in the Donald Trump election. The popular vote was within the margin of error of the polling. It's only in two states (Wisconsin and Michigan) that the actual result differed from the polling result by more than the margin of error. The greater problem was that ...


Only top voted, non community-wiki answers of a minimum length are eligible