andrew_gelman_stats andrew_gelman_stats-2010 andrew_gelman_stats-2010-364 knowledge-graph by maker-knowledge-mining
Source: html
Introduction: Nate Silver and Justin Wolfers are having a friendly blog-dispute about momentum in political polling. Nate and Justin each make good points but are also missing parts of the picture. These questions relate to my own research so I thought I’d discuss them here. There ain’t no mo’ Nate led off the discussion by writing that pundits are always talking about “momentum” in the polls: Turn on the news or read through much of the analysis put out by some of our friends, and you’re likely to hear a lot of talk about “momentum”: the term is used about 60 times per day by major media outlets in conjunction with articles about polling. When people say a particular candidate has momentum, what they are implying is that present trends are likely to perpetuate themselves into the future. Say, for instance, that a candidate trailed by 10 points in a poll three weeks ago — and now a new poll comes out showing the candidate down by just 5 points. It will frequently be said that this
sentIndex sentText sentNum sentScore
1 When people say a particular candidate has momentum, what they are implying is that present trends are likely to perpetuate themselves into the future. [sent-5, score-0.38]
2 Say, for instance, that a candidate trailed by 10 points in a poll three weeks ago — and now a new poll comes out showing the candidate down by just 5 points. [sent-6, score-0.684]
3 They create the impression that — if the candidate has gone from being 10 points down to 5 points down, then by next week, he’ll have closed his deficit further: perhaps he’ll even be ahead! [sent-9, score-0.406]
4 But, as Nate points out, this ain’t actually happening: Say that a candidate has improved her position in the polls from August to September. [sent-10, score-0.625]
5 Sometimes, a candidate who has gained ground in the polls continues to do so; otherwise, the trend reverses itself, or the race simply flatlines. [sent-18, score-0.695]
6 There is also no sign of momentum we look at the change in polling between other periods. [sent-22, score-0.58]
7 In general elections, the direction in which polls have moved is not predictive of the direction in which they will move . [sent-26, score-0.548]
8 Nate is saying that the change from A to B does not predict the change from B to C (or, to be precise, that the change from A to B is a very weak and negative predictor of the change from B to C). [sent-32, score-0.488]
9 Public opinion is not a random walk Nate does slip up at one point, when he writes: In races with lots of polling, instead, the most robust assumption is usually that polling is essentially a random walk, i. [sent-47, score-0.826]
10 , that the polls are about equally likely to move toward one or another candidate, regardless of which way they have moved in the past. [sent-49, score-0.518]
11 For many races, you can use a forecast from the fundamentals to get a pretty good idea about where the polls are going to end up. [sent-51, score-0.367]
12 Or, if you want to focus on congressional elections, take a look at the work of Erikson, Bafumi, and Wlezien , who find predictable changes in the generic opinion polls in the year leading up to the election. [sent-54, score-0.538]
13 First, individual polls are noisy, and so any immediate changes are likely to be noise. [sent-56, score-0.509]
14 Second, the random walk is such a standard paradigm for statistical noise that it’s natural for Nate to use it as a default. [sent-57, score-0.434]
15 If a team has an unexpectedly good record midway through the season, it’s likely that they will slip in the standings during the second half. [sent-65, score-0.367]
16 (I was going to write “the random walk story contributes to . [sent-68, score-0.405]
17 ) The natural accompaniment to the random walk model is the idea that, if you can shift the polls by X percentage points at any point during the campaign, this will give you an expected X percentage point advantage when the election comes around. [sent-72, score-0.983]
18 Another implication is that George Bush’s campaign was so awesome and Michael Dukakis’s campaign was so horrible to explain the big swing in the polls in the months leading up to the 1988 election. [sent-73, score-0.658]
19 ( We think a much more plausible story is that the polls were gonna swing toward Bush big-time, and the perceived incompetence of Dukakis’s campaign was a consequence, not a cause, of this polling shift. [sent-74, score-0.713]
20 As Nate emphasizes, such a prediction is only appropriate in the context of real-world information, hypotheses of “factors above and beyond the direction in which the polls have moved in the past. [sent-77, score-0.554]
wordName wordTfidf (topN-words)
[('nate', 0.444), ('polls', 0.367), ('momentum', 0.344), ('walk', 0.228), ('justin', 0.222), ('candidate', 0.178), ('random', 0.133), ('poll', 0.124), ('change', 0.122), ('trends', 0.118), ('polling', 0.114), ('dukakis', 0.111), ('campaign', 0.103), ('races', 0.096), ('standings', 0.088), ('swing', 0.085), ('bush', 0.085), ('likely', 0.084), ('team', 0.08), ('points', 0.08), ('baseball', 0.08), ('noise', 0.073), ('wolfers', 0.072), ('slip', 0.069), ('deficit', 0.068), ('moved', 0.067), ('scales', 0.065), ('prediction', 0.063), ('season', 0.062), ('ain', 0.062), ('percentage', 0.062), ('predictable', 0.06), ('changes', 0.058), ('direction', 0.057), ('times', 0.056), ('standpoint', 0.056), ('ground', 0.056), ('series', 0.056), ('opinion', 0.053), ('correct', 0.052), ('shift', 0.051), ('unemployment', 0.051), ('noisy', 0.051), ('race', 0.05), ('record', 0.046), ('elections', 0.045), ('george', 0.044), ('contributes', 0.044), ('incompetence', 0.044), ('reverses', 0.044)]
simIndex simValue blogId blogTitle
same-blog 1 0.99999976 364 andrew gelman stats-2010-10-22-Politics is not a random walk: Momentum and mean reversion in polling
Introduction: Nate Silver and Justin Wolfers are having a friendly blog-dispute about momentum in political polling. Nate and Justin each make good points but are also missing parts of the picture. These questions relate to my own research so I thought I’d discuss them here. There ain’t no mo’ Nate led off the discussion by writing that pundits are always talking about “momentum” in the polls: Turn on the news or read through much of the analysis put out by some of our friends, and you’re likely to hear a lot of talk about “momentum”: the term is used about 60 times per day by major media outlets in conjunction with articles about polling. When people say a particular candidate has momentum, what they are implying is that present trends are likely to perpetuate themselves into the future. Say, for instance, that a candidate trailed by 10 points in a poll three weeks ago — and now a new poll comes out showing the candidate down by just 5 points. It will frequently be said that this
2 0.43631455 1512 andrew gelman stats-2012-09-27-A Non-random Walk Down Campaign Street
Introduction: Political campaigns are commonly understood as random walks, during which, at any point in time, the level of support for any party or candidate is equally likely to go up or down. Each shift in the polls is then interpreted as the result of some combination of news and campaign strategies. A completely different story of campaigns is the mean reversion model in which the elections are determined by fundamental factors of the economy and partisanship; the role of the campaign is to give voters a chance to reach their predetermined positions. The popularity of the random walk model for polls may be partially explained via analogy to the widespread idea that stock prices reflect all available information, as popularized in Burton Malkiel’s book, A Random Walk Down Wall Street. Once the idea has sunk in that short-term changes in the stock market are inherently unpredictable, it is natural for journalists to think the same of polls. For example, political analyst Nate Silver wrote
3 0.3797968 131 andrew gelman stats-2010-07-07-A note to John
Introduction: Jeff the Productivity Sapper points me to this insulting open letter to Nate Silver written by pollster John Zogby. I’ll go through bits of Zogby’s note line by line. (Conflict of interest warning: I have collaborated with Nate and I blog on his site). Zogby writes: Here is some advice from someone [Zogby] who has been where you [Silver] are today. Sorry, John. (I can call you that, right? Since you’re calling Nate “Nate”?). Yes, you were once the hot pollster. But, no, you were never where Nate is today. Don’t kid yourself. Zogby writes: You [Nate] are hot right now – using an aggregate of other people’s work, you got 49 of 50 states right in 2008. Yes, Nate used other people’s work. That’s what’s called “making use of available data.” Or, to use a more technical term employed in statistics, it’s called “not being an idiot.” Only in the wacky world of polling are you supposed to draw inferences about the U.S.A. using only a single survey organization. I do
4 0.31221026 270 andrew gelman stats-2010-09-12-Comparison of forecasts for the 2010 congressional elections
Introduction: Yesterday at the sister blog , Nate Silver forecast that the Republicans have a two-thirds chance of regaining the House of Representatives in the upcoming election, with an expected gain of 45 House seats. Last month, Bafumi, Erikson, and Wlezien released their forecast that gives the Republicans an 80% chance of takeover and an expected gain of 50 seats. As all the above writers emphasize, these forecasts are full of uncertainty, so I treat the two predictions–a 45-seat swing or a 50-seat swing–as essentially identical at the national level. And, as regular readers know, as far back as a year ago , the generic Congressional ballot (those questions of the form, “Which party do you plan to vote for in November?”) was also pointing to big Republican gains. As Bafumi et al. point out, early generic polls are strongly predictive of the election outcome, but they need to be interpreted carefully. The polls move in a generally predictable manner during the year leading up to an
5 0.27439547 1570 andrew gelman stats-2012-11-08-Poll aggregation and election forecasting
Introduction: At the sister blog, Henry writes about poll averaging and election forecasts. Henry writes that “These models need to crunch lots of polls, at the state and national level, if they’re going to provide good predictions.” Actually, you can get reasonable predictions from national-level forecasting models plus previous state-level election results, then when the election comes closer you can use national and state polls as needed. See my paper with Kari Lock, Bayesian combination of state polls and election forecasts . (That said, the method in that paper is fairly complicated, much more so than simply taking weighted averages of state polls, if such abundant data happen to be available. And I’m sure our approach would need to be altered if it were used for real-time forecasts.) Having a steady supply of polls of varying quality from various sources allows poll aggregators to produce news every day (in the sense of pushing their estimates around) but it doesn’t help much with a
6 0.26202172 391 andrew gelman stats-2010-11-03-Some thoughts on election forecasting
7 0.24510908 210 andrew gelman stats-2010-08-16-What I learned from those tough 538 commenters
9 0.22076593 1634 andrew gelman stats-2012-12-21-Two reviews of Nate Silver’s new book, from Kaiser Fung and Cathy O’Neil
10 0.20802191 300 andrew gelman stats-2010-09-28-A calibrated Cook gives Dems the edge in Nov, sez Sandy
11 0.17019962 2343 andrew gelman stats-2014-05-22-Big Data needs Big Model
12 0.15845527 250 andrew gelman stats-2010-09-02-Blending results from two relatively independent multi-level models
13 0.14751725 43 andrew gelman stats-2010-05-19-What do Tuesday’s elections tell us about November?
14 0.13980912 654 andrew gelman stats-2011-04-09-There’s no evidence that voters choose presidential candidates based on their looks
15 0.13687511 987 andrew gelman stats-2011-11-02-How Khan Academy is using Machine Learning to Assess Student Mastery
16 0.13571209 100 andrew gelman stats-2010-06-19-Unsurprisingly, people are more worried about the economy and jobs than about deficits
18 0.12392515 1610 andrew gelman stats-2012-12-06-Yes, checking calibration of probability forecasts is part of Bayesian statistics
19 0.12148759 1695 andrew gelman stats-2013-01-28-Economists argue about Bayes
topicId topicWeight
[(0, 0.189), (1, -0.054), (2, 0.138), (3, 0.076), (4, -0.033), (5, 0.008), (6, -0.077), (7, -0.04), (8, 0.004), (9, -0.056), (10, 0.09), (11, 0.02), (12, 0.049), (13, -0.093), (14, -0.133), (15, 0.022), (16, -0.016), (17, 0.028), (18, 0.048), (19, 0.007), (20, -0.072), (21, 0.111), (22, -0.031), (23, 0.105), (24, 0.023), (25, 0.046), (26, 0.023), (27, 0.012), (28, -0.014), (29, 0.086), (30, -0.034), (31, 0.032), (32, -0.017), (33, -0.079), (34, 0.047), (35, 0.102), (36, 0.022), (37, -0.007), (38, 0.038), (39, -0.055), (40, -0.024), (41, 0.113), (42, -0.016), (43, -0.043), (44, -0.088), (45, 0.065), (46, 0.0), (47, -0.057), (48, 0.074), (49, 0.033)]
simIndex simValue blogId blogTitle
same-blog 1 0.95389855 364 andrew gelman stats-2010-10-22-Politics is not a random walk: Momentum and mean reversion in polling
Introduction: Nate Silver and Justin Wolfers are having a friendly blog-dispute about momentum in political polling. Nate and Justin each make good points but are also missing parts of the picture. These questions relate to my own research so I thought I’d discuss them here. There ain’t no mo’ Nate led off the discussion by writing that pundits are always talking about “momentum” in the polls: Turn on the news or read through much of the analysis put out by some of our friends, and you’re likely to hear a lot of talk about “momentum”: the term is used about 60 times per day by major media outlets in conjunction with articles about polling. When people say a particular candidate has momentum, what they are implying is that present trends are likely to perpetuate themselves into the future. Say, for instance, that a candidate trailed by 10 points in a poll three weeks ago — and now a new poll comes out showing the candidate down by just 5 points. It will frequently be said that this
2 0.89259768 270 andrew gelman stats-2010-09-12-Comparison of forecasts for the 2010 congressional elections
Introduction: Yesterday at the sister blog , Nate Silver forecast that the Republicans have a two-thirds chance of regaining the House of Representatives in the upcoming election, with an expected gain of 45 House seats. Last month, Bafumi, Erikson, and Wlezien released their forecast that gives the Republicans an 80% chance of takeover and an expected gain of 50 seats. As all the above writers emphasize, these forecasts are full of uncertainty, so I treat the two predictions–a 45-seat swing or a 50-seat swing–as essentially identical at the national level. And, as regular readers know, as far back as a year ago , the generic Congressional ballot (those questions of the form, “Which party do you plan to vote for in November?”) was also pointing to big Republican gains. As Bafumi et al. point out, early generic polls are strongly predictive of the election outcome, but they need to be interpreted carefully. The polls move in a generally predictable manner during the year leading up to an
3 0.89103174 300 andrew gelman stats-2010-09-28-A calibrated Cook gives Dems the edge in Nov, sez Sandy
Introduction: Sandy Gordon sends along this fun little paper forecasting the 2010 midterm election using expert predictions (the Cook and Rothenberg Political Reports). Gordon’s gimmick is that he uses past performance to calibrate the reports’ judgments based on “solid,” “likely,” “leaning,” and “toss-up” categories, and then he uses the calibrated versions of the current predictions to make his forecast. As I wrote a few weeks ago in response to Nate’s forecasts, I think the right way to go, if you really want to forecast the election outcome, is to use national information to predict the national swing and then do regional, state, and district-level adjustments using whatever local information is available. I don’t see the point of using only the expert forecasts and no other data. Still, Gordon is bringing new information (his calibrations) to the table, so I wanted to share it with you. Ultimately I like the throw-in-everything approach that Nate uses (although I think Nate’s descr
4 0.8533408 131 andrew gelman stats-2010-07-07-A note to John
Introduction: Jeff the Productivity Sapper points me to this insulting open letter to Nate Silver written by pollster John Zogby. I’ll go through bits of Zogby’s note line by line. (Conflict of interest warning: I have collaborated with Nate and I blog on his site). Zogby writes: Here is some advice from someone [Zogby] who has been where you [Silver] are today. Sorry, John. (I can call you that, right? Since you’re calling Nate “Nate”?). Yes, you were once the hot pollster. But, no, you were never where Nate is today. Don’t kid yourself. Zogby writes: You [Nate] are hot right now – using an aggregate of other people’s work, you got 49 of 50 states right in 2008. Yes, Nate used other people’s work. That’s what’s called “making use of available data.” Or, to use a more technical term employed in statistics, it’s called “not being an idiot.” Only in the wacky world of polling are you supposed to draw inferences about the U.S.A. using only a single survey organization. I do
5 0.84698033 1570 andrew gelman stats-2012-11-08-Poll aggregation and election forecasting
Introduction: At the sister blog, Henry writes about poll averaging and election forecasts. Henry writes that “These models need to crunch lots of polls, at the state and national level, if they’re going to provide good predictions.” Actually, you can get reasonable predictions from national-level forecasting models plus previous state-level election results, then when the election comes closer you can use national and state polls as needed. See my paper with Kari Lock, Bayesian combination of state polls and election forecasts . (That said, the method in that paper is fairly complicated, much more so than simply taking weighted averages of state polls, if such abundant data happen to be available. And I’m sure our approach would need to be altered if it were used for real-time forecasts.) Having a steady supply of polls of varying quality from various sources allows poll aggregators to produce news every day (in the sense of pushing their estimates around) but it doesn’t help much with a
6 0.83961046 1512 andrew gelman stats-2012-09-27-A Non-random Walk Down Campaign Street
7 0.82927185 2343 andrew gelman stats-2014-05-22-Big Data needs Big Model
8 0.81382364 43 andrew gelman stats-2010-05-19-What do Tuesday’s elections tell us about November?
9 0.76212811 391 andrew gelman stats-2010-11-03-Some thoughts on election forecasting
10 0.73687053 292 andrew gelman stats-2010-09-23-Doug Hibbs on the fundamentals in 2010
12 0.71320218 210 andrew gelman stats-2010-08-16-What I learned from those tough 538 commenters
13 0.70450288 508 andrew gelman stats-2011-01-08-More evidence of growing nationalization of congressional elections
14 0.70244926 478 andrew gelman stats-2010-12-20-More on why “all politics is local” is an outdated slogan
15 0.68632573 934 andrew gelman stats-2011-09-30-Nooooooooooooooooooo!
17 0.67108268 237 andrew gelman stats-2010-08-27-Bafumi-Erikson-Wlezien predict a 50-seat loss for Democrats in November
18 0.67064863 1000 andrew gelman stats-2011-11-10-Forecasting 2012: How much does ideology matter?
20 0.65097409 2181 andrew gelman stats-2014-01-21-The Commissar for Traffic presents the latest Five-Year Plan
topicId topicWeight
[(5, 0.204), (9, 0.058), (16, 0.078), (21, 0.021), (24, 0.073), (42, 0.011), (50, 0.013), (63, 0.039), (65, 0.012), (86, 0.03), (95, 0.018), (99, 0.27)]
simIndex simValue blogId blogTitle
1 0.97563505 422 andrew gelman stats-2010-11-20-A Gapminder-like data visualization package
Introduction: Ossama Hamed writes in with a new dynamic graphing software: I have the pleasure to brief you on our Data Visualization software “Trend Compass”. TC is a new concept in viewing statistics and trends in an animated way by displaying in one chart 5 axis (X, Y, Time, Bubble size & Bubble color) instead of just the traditional X and Y axis. . . .
Introduction: As I wrote a couple years ago: Even though statistical analysis has demonstrated that presidential elections are predictable given economic conditions and previous votes in the states . . . it certainly doesn’t mean that every election can be accurately predicted ahead of time. Presidential general election campaigns have several distinct features that distinguish them from most other elections: 1. Two major candidates; 2. The candidates clearly differ in their political ideologies and in their positions on economic issues; 3. The two sides have roughly equal financial and organizational resources; 4. The current election is the latest in a long series of similar contests (every four years); 5. A long campaign, giving candidates a long time to present their case and giving voters a long time to make up their minds. Other elections look different. . . . Or, as I said in reference to the current NYC mayoral election: Et selon Andrew Gelman, expert de l’universi
3 0.96511817 513 andrew gelman stats-2011-01-12-“Tied for Warmest Year On Record”
Introduction: The National Climatic Data Center has tentatively announced that 2010 is, get this, “tied” for warmest on record. Presumably they mean it’s tied to the precision that they quote (1.12 F above the 20th-century average). The uncertainty in the measurements, as well as some fuzziness about exactly what is being measured (how much of the atmosphere, and the oceans) makes these global-average things really suspect. For instance, if there’s more oceanic turnover one year, that can warm the deep ocean but cool the shallow ocean and atmosphere, so even though the heat content of the atmosphere-ocean system goes up, some of these “global-average” estimates can go down. The reverse can happen too. And of course there are various sources of natural variability that are not, these days, what most people are most interested in. So everybody who knows about the climate professes to hate the emphasis on climate records. And yet, they’re irresistible. I’m sure we’ll see the usual clamor of som
4 0.96509051 1250 andrew gelman stats-2012-04-07-Hangman tips
Introduction: Jeff pointed me to this article by Nick Berry. It’s kind of fun but of course if you know your opponent will be following this strategy you can figure out how to outwit it. Also, Berry writes that ETAOIN SHRDLU CMFWYP VBGKQJ XZ is the “ordering of letter frequency in English language.” Indeed this is the conventional ordering but nobody thinks it’s right anymore. See here (with further discussion here ). I wonder what corpus he’s using. P.S. Klutz was my personal standby.
5 0.95638412 665 andrew gelman stats-2011-04-17-Yes, your wish shall be granted (in 25 years)
Introduction: This one was so beautiful I just had to repost it: From the New York Times, 9 Sept 1981: IF I COULD CHANGE PARK SLOPE If I could change Park Slope I would turn it into a palace with queens and kings and princesses to dance the night away at the ball. The trees would look like garden stalks. The lights would look like silver pearls and the dresses would look like soft silver silk. You should see the ball. It looks so luxurious to me. The Park Slope ball is great. Can you guess what street it’s on? “Yes. My street. That’s Carroll Street.” – Jennifer Chatmon, second grade, P.S. 321 This was a few years before my sister told me that she felt safer having a crack house down the block because the cops were surveilling it all the time.
6 0.95044982 87 andrew gelman stats-2010-06-15-Statistical analysis and visualization of the drug war in Mexico
same-blog 7 0.94647408 364 andrew gelman stats-2010-10-22-Politics is not a random walk: Momentum and mean reversion in polling
8 0.94567096 1512 andrew gelman stats-2012-09-27-A Non-random Walk Down Campaign Street
10 0.93741149 1286 andrew gelman stats-2012-04-28-Agreement Groups in US Senate and Dynamic Clustering
11 0.92263132 951 andrew gelman stats-2011-10-11-Data mining efforts for Obama’s campaign
12 0.92024469 1052 andrew gelman stats-2011-12-11-Rational Turbulence
13 0.91399896 228 andrew gelman stats-2010-08-24-A new efficient lossless compression algorithm
14 0.90982395 1606 andrew gelman stats-2012-12-05-The Grinch Comes Back
15 0.90386546 224 andrew gelman stats-2010-08-22-Mister P gets married
18 0.88402468 1207 andrew gelman stats-2012-03-10-A quick suggestion
19 0.88391769 2194 andrew gelman stats-2014-02-01-Recently in the sister blog
20 0.88273925 123 andrew gelman stats-2010-07-01-Truth in headlines