Elections are random events. From individuals deciding whether to vote, to individuals deciding for whom to vote, to election authorities deciding what to count, the outcomes of competitive ...democratic elections are rarely known until election day or beyond. Understanding Elections Through Statistics: Polling Prediction, and Testing explores this random phenomenon from two points of view: predicting the election outcome using opinion polls and testing the election outcome using government-reported data.
Written for those with only a brief introduction to statistics, this book takes you on a statistical journey from how polls are taken to how they can-and should-be used to estimate current popular opinion. Once an understanding of the election process is built, we turn towards testing elections for evidence of unfairness. While holding elections has become the de facto proof of government legitimacy, those electoral processes may hide the dirty little secret of the government illicitly ensuring a favorable election outcome.
This book includes these features designed to make your statistical journey more enjoyable:
vignettes of elections, including maps, starting each chapter to motivate the material;
in-chapter cues to help one avoid the heavy math-or focus on it;
end-of-chapter problems designed to review and extend that which was covered in the chapter; and
many opportunities to turn the power of the R Statistical Environment to the enclosed election data files, as well as to those you find interesting.
From these features, it is clear that the audience for this book is quite diverse. It provides the mathematics for those interested in mathematics, but also provides detours for those who just want a good read and a deeper understanding of elections.
Ole J. Forsberg holds PhD degrees both in Political Science and in Statistics. He currently teaches mathematics and statistics in the Department of Mathematics at Knox College in Galesburg, IL.
US election polls: a quick postmortem Forsberg, Ole J.
Significance (Oxford, England),
February 2021, 2021-02-01, 20210201, Volume:
18, Issue:
1
Journal Article
Open access
How did the 2020 US presidential election polls really do? Ole J. Forsberg gives his assessment
How did the 2020 US presidential election polls really do? Ole J. Forsberg gives his assessment.
In 2020, as in 2016, US presidential candidates may dismiss some election polls as fake while lauding others as trustworthy. But don't just take their word for it! Ole J. Forsberg offers information ...and advice to help voters decide which polls and polling houses to trust
In 2020, as in 2016, US presidential candidates may dismiss some election polls as fake while lauding others as trustworthy. But don't just take their word for it! Ole J. Forsberg offers information and advice to help voters decide which polls and polling houses to trust.
Bayesian regression has emerged as a viable alternative for the estimation of curriculum-based measurement (CBM) growth slopes. Preliminary findings suggest such methods may yield improved efficiency ...relative to other linear estimators and can be embedded into data management programs for high-frequency use. However, additional research is needed, as Bayesian estimators require multiple specifications of the prior distributions. The current study evaluates the accuracy of several combinations of prior values, including three distributions of the residuals, two values of the expected growth rate, and three possible values for the precision of slope when using Bayesian simple linear regression to estimate fluency growth slopes for reading CBM. We also included traditional ordinary least squares (OLS) as a baseline contrast. Findings suggest that the prior specification for the residual distribution had, on average, a trivial effect on the accuracy of the slope. However, specifications for growth rate and precision of slope were influential, and virtually all variants of Bayesian regression evaluated were superior to OLS. Converging evidence from both simulated and observed data now suggests Bayesian methods outperform OLS for estimating CBM growth slopes and should be strongly considered in research and practice.
Bayesian techniques have become increasingly present in the social sciences, fueled by advances in computer speed and the development of user-friendly software. In this paper, we forward the use of ...Bayesian Asymmetric Regression (BAR) to monitor intervention responsiveness when using Curriculum-Based Measurement (CBM) to assess oral reading fluency (ORF). An overview of Bayesian methods and their application to the problem-solving model is first presented, which is further illustrated by a case example. We conclude the paper with a Monte Carlo simulation study demonstrating the validity of BAR, as compared to the current standard of practice for CBM decision-making, ordinary least squares (OLS) regression. Results suggest that BAR is most advantageous with studies using small-to-moderate sample sizes, and when distributional information (such as the probability of intervention success) is of interest.
Impact and Implications
Disagreement persists regarding how long one needs to progress monitor to make a confident decision regarding intervention responsiveness. We forward a possible resolution: the use of a robust Bayesian estimator. Our findings suggest that the Bayesian method is competitive with traditional methods and offers significant advantages to decision making within a problem-solving model.
This article explores the performance of major political polls as reported on a well-known political website. The presidential elections of 2004, 2008, and 2012 are examined, and polling results for ...battleground states are compared to the actual margins in each state. These data are combined into an analysis of variance (ANOVA) model and differences in the various polls assessed. No significant differences in polls were found in 2004 or 2008. In contrast, very large differences were found in 2012, and all polls were found to underestimate President Obama's performance in the battleground states. In addition to individual polling results, a model using the statewide and national polling data from all polls is created. These results are compared to the established polls. We present a possible explanation as to why the polls in the recent (2012) election are all biased toward the Republican side.
Differential Invalidation Forsberg, Ole J.
Understanding Elections through Statistics,
2021
Book Chapter
There are a couple other methods to cheat in an election, both are examined in this chapter. One method is differential invalidation; the other, ballot box stuffing. Both methods leave evidence that ...can be detected using regression. This chapter examines methods for detecting that evidence, using the 2010 Ivoirian presidential election that led to a flare-up in their civil war.
Specifically, this chapter looks at four regression methods that can test for both: ordinary least squares regression, weighted least squares regression, binomial regression, and beta-binomial regression. The first is typically taught in introductory statistics courses, so it may be familiar. The other three are more advanced in their mathematics, though not in their application here. The computer allows us to select the best method without having to worry about the internal calculations that must be done.
This chapter examines the six Sri Lankan presidential elections between 1994 and 2019 for empirical evidence of persistent electoral unfairness in favor of the government-supported candidates. While ...regression tests for differential invalidation are the primary methods used, geography will help to illustrate some of the findings — and to provide additional questions.
The results raise the specter that Sri Lankan elections fail to fully meet the requirements of a free and fair election. However, as the years have passed, the elections have gotten better. With that said, the Tamil question arises from the evidence of differential invalidation against Tamil-dominant regions of Sri Lanka.
Combining Polls Forsberg, Ole J.
Understanding Elections through Statistics,
2021
Book Chapter
Thus far, we have been drawing conclusions from a single poll. While this is a useful skill to have, elections are polled frequently; that is, there are multiple polls taken in a given election. Can ...we better estimate the population support, π, if we combine those multiple polls? Clearly, since this chapter exists, the answer is yes. The more interesting question concerns how we can combine the polls in a meaningful way.
Using the 2017 South Korean election as a backdrop, this chapter looks at methods for combining and weighting polls to properly predict the election result. While elections are random, much like the world, these methods do improve on estimates.
This chapter applies the theory and methods from the first half of the book to a single election: The 2016 Brexit referendum. This chapter raises a lot of questions — some of which are undecided in ...polling. However, it should give you a taste of what poll analysts do to produce their estimates. By the end, this chapter provides some suggestions for both polling houses and the media for performing and covering polls.
Since 1993, various factions of various British parties have pushed for the United Kingdom to remove itself from the EU. Arguably, the most vocal Euroskeptic party was the UK Independence Party, which formed in 1991, changed its name in 1993, and increased its vote share in Parliament in each election until after the Brexit vote in 2016. The interesting thing about this graphic is that polls of “likely voters” became very erratic in the waning weeks of the election period. The position of “all adults” was much more stable. Furthermore, the “all adult” surveys came closer to the actual Brexit vote. The 72% turnout was the highest for any British election since 1992, and for any British referendum ever. This high turnout was not constant across all groups. The older voters turned out at a much higher rate than did the younger, 90% of the former vs. 64% of the latter.