Aggregate data yield biased estimates of voter preferences

2021 ◽  
pp. 102604
Author(s):  
Corey Lang ◽  
Shanna Pearson-Merkowitz
Keyword(s):  
Aggregate Data ◽  
Voter Preferences ◽  
Data Yield ◽  
Biased Estimates

Questions to the Article: An Artificial Intelligence Model to Predict the Mortality of COVID-19 Patients at Hospital Admission Time Using Routine Blood Samples (Preprint)

2020 ◽  
Author(s):  
Shu-Chun Kuo ◽  
CHIEN WEI ◽  
Willy Chou
Keyword(s):  
Data Entry ◽  
Ann Model ◽  
Model Accuracy ◽  
Data Format ◽  
Raw Data ◽  
Probability Of Survival ◽  
Testing Data ◽  
Laboratory Results ◽  
Data Yield ◽  
Artificial Neural Network Ann

UNSTRUCTURED The recent article published on December 23 27 in 2020 is well-written and of interest, but remains several questions that are required for clarifications, including (1) 30 feature variables with normalized format(mean=0 and SD=1) required to compare model accuracy with those with the raw-data format; (2)inconsistency in variable numbers between entry and preview panels in Figure 4 and reference typos; and (3) data-entry format with raw blood laboratory results in Figure 4 inconsistent with the model designed using normalized data to estimate parameters. We conducted a study using the training and testing data provided by the previous study. An artificial neural network(ANN) model was performed to estimate parameters and compare the model accuracy with those eight models provided by the previous study. We found that (1) normalized data yield higher accuracy than that with the raw data; (2) typos definitely exist at the bottom review (=32>30 variables in the entry) panels in Figure 4 and typos in Table 6; and (3)the ANN earns a probability of survival(=0.91) higher than that(=0.71) in the previous study using the similar entry data when the raw data are assumed in the app. We also demonstrated an author-made app using the visualization to display the prediction result, which is novel and innovative to make the result improved with a dashboard in comparison with the previous study.

Examining Nonnormal Latent Variable Distributions for Non-Ignorable Missing Data

2021 ◽  
Vol 45 (3) ◽  
pp. 159-177
Author(s):  
Chen-Wei Liu
Keyword(s):  
Missing Data ◽  
Latent Variable ◽  
Bivariate Normal Distribution ◽  
Parameter Estimates ◽  
Missing Responses ◽  
Variable Distribution ◽  
Item Responses ◽  
Nonnormal Distribution ◽  
The Impact ◽  
Biased Estimates

Missing not at random (MNAR) modeling for non-ignorable missing responses usually assumes that the latent variable distribution is a bivariate normal distribution. Such an assumption is rarely verified and often employed as a standard in practice. Recent studies for “complete” item responses (i.e., no missing data) have shown that ignoring the nonnormal distribution of a unidimensional latent variable, especially skewed or bimodal, can yield biased estimates and misleading conclusion. However, dealing with the bivariate nonnormal latent variable distribution with present MNAR data has not been looked into. This article proposes to extend unidimensional empirical histogram and Davidian curve methods to simultaneously deal with nonnormal latent variable distribution and MNAR data. A simulation study is carried out to demonstrate the consequence of ignoring bivariate nonnormal distribution on parameter estimates, followed by an empirical analysis of “don’t know” item responses. The results presented in this article show that examining the assumption of bivariate nonnormal latent variable distribution should be considered as a routine for MNAR data to minimize the impact of nonnormality on parameter estimates.

scholarly journals Comparing the use of aggregate data and various methods of integrating individual patient data to network meta-analysis and its application to first-line ART

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Steve Kanters ◽  
Mohammad Ehsanul Karim ◽  
Kristian Thorlund ◽  
Aslam H. Anis ◽  
Michael Zoratti ◽  
...  
Keyword(s):  
Meta Analysis ◽  
Evidence Base ◽  
Aggregate Data ◽  
World Health ◽  
Coefficient Estimates ◽  
Two Stage ◽  
Individual Patient Level ◽  
One Stage ◽  
Meta Regression ◽  
Empirical Priors

Abstract Background The 2018 World Health Organization HIV guidelines were based on the results of a network meta-analysis (NMA) of published trials. This study employed individual patient-level data (IPD) and aggregate data (AgD) and meta-regression methods to assess the evidence supporting the WHO recommendations and whether they needed any refinements. Methods Access to IPD from three trials was granted through ClinicalStudyDataRequest.com (CSDR). Seven modelling approaches were applied and compared: 1) Unadjusted AgD network meta-analysis (NMA) – the original analysis; 2) AgD-NMA with meta-regression; 3) Two-stage IPD-AgD NMA; 4) Unadjusted one-stage IPD-AgD NMA; 5) One-stage IPD-AgD NMA with meta-regression (one-stage approach); 6) Two-stage IPD-AgD NMA with empirical-priors (empirical-priors approach); 7) Hierarchical meta-regression IPD-AgD NMA (HMR approach). The first two were the models used previously. Models were compared with respect to effect estimates, changes in the effect estimates, coefficient estimates, DIC and model fit, rankings and between-study heterogeneity. Results IPD were available for 2160 patients, representing 6.5% of the evidence base and 3 of 24 edges. The aspect of the model affected by the choice of modeling appeared to differ across outcomes. HMR consistently generated larger intervals, often with credible intervals (CrI) containing the null value. Discontinuations due to adverse events and viral suppression at 96 weeks were the only two outcomes for which the unadjusted AgD NMA would not be selected. For the first, the selected model shifted the principal comparison of interest from an odds ratio of 0.28 (95% CrI: 10.17, 0.44) to 0.37 (95% CrI: 0.23, 0.58). Throughout all outcomes, the regression estimates differed substantially between AgD and IPD methods, with the latter being more often larger in magnitude and statistically significant. Conclusions Overall, the use of IPD often impacted the coefficient estimates, but not sufficiently as to necessitate altering the final recommendations of the 2018 WHO Guidelines. Future work should examine the features of a network where adjustments will have an impact, such as how much IPD is required in a given size of network.

President Obama and the Emergence of Islamophobia in Mass Partisan Preferences

2021 ◽  
pp. 106591292110072
Author(s):  
Michael Tesler
Keyword(s):  
Theoretical Basis ◽  
Independent Predictor ◽  
Survey Experiment ◽  
President Obama ◽  
Significant Independent Predictor ◽  
Cross Sectional ◽  
Panel Surveys ◽  
Voter Preferences ◽  
Nationally Representative ◽  
The Relationship

This article argues that the unusually large and persistent association between Islamophobia and opposition to President Obama helped make attitudes about Muslims a significant, independent predictor of Americans’ broader partisan preferences. After detailing the theoretical basis for this argument, the article marshals repeated cross-sectional data, two panel surveys, and a nationally representative survey experiment, to test its hypotheses. The results from those analyses show the following: (1) attitudes about Muslims were a significantly stronger independent predictor of voter preferences for congress in 2010–2014 elections than they were in 2004–2008; (2) attitudes about Muslims were a significantly stronger independent predictor of mass partisanship during Obama’s presidency than they were beforehand; and (3) experimentally connecting Obama to Democratic congressional candidates significantly increased the relationship between anti-Muslim sentiments and Americans’ preferences for Republican congressional candidates. The article concludes with a discussion of the implications of these results for American politics in the Trump era.

Accuracy development in L2 writing: Exploring the potential of computer-assisted unfocused indirect corrective feedback in an online EFL course

ReCALL ◽  
2021 ◽  
pp. 1-17
Author(s):  
Cédric Brudermann ◽  
Muriel Grosbois ◽  
Cédric Sarré
Keyword(s):  
Foreign Language ◽  
Corrective Feedback ◽  
Computer Assisted ◽  
Mathematics Students ◽  
Language Teacher Education ◽  
Writing Accuracy ◽  
Data Yield ◽  
And Mathematics ◽  
The Impact ◽  
Error Types

Abstract In a previous study (Sarré, Grosbois & Brudermann, 2019), we explored the effects of various corrective feedback (CF) strategies on interlanguage development for the online component of a blended English as a foreign language (EFL) course we had designed and implemented. Our results showed that unfocused indirect CF (feedback on all error types through the provision of metalinguistic comments on the nature of the errors made) combined with extra computer-mediated micro-tasks was the most efficient CF type to foster writing accuracy development in our context. Following up on this study, this paper further explores the effects of this specific CF type on learners’ written accuracy development in an online EFL course designed for freshmen STEM (science, technology, engineering, and mathematics) students. In the online course under study, this specific CF type was experimented with different cohorts of STEM learners (N = 1,150) over a five-year period (from 2014 to 2019) and was computer-assisted: CF provision online by a human tutor was combined with predetermined CF comments. The aim of this paper is to investigate the impact of this specific CF strategy on error types. In this respect, the data yield encouraging results in terms of writing accuracy development when learners benefit from this computer-assisted specific CF. This study thus helps to gain a better understanding of the role that CF plays in shaping students’ revision processes and could inform language (teacher) education regarding the use of digital tools for the development of foreign language accuracy and the issues related to online CF provision.

scholarly journals A Traveler’s Guide to the Multiverse: Promises, Pitfalls, and a Framework for the Evaluation of Analytic Decisions

2021 ◽  
Vol 4 (1) ◽  
pp. 251524592095492
Author(s):  
Marco Del Giudice ◽  
Steven W. Gangestad
Keyword(s):  
Degrees Of Freedom ◽  
A Priori ◽  
Simulated Data ◽  
Style Analysis ◽  
Data Set ◽  
Scant Attention ◽  
Equivalence Type ◽  
The Impact ◽  
Biased Estimates

Decisions made by researchers while analyzing data (e.g., how to measure variables, how to handle outliers) are sometimes arbitrary, without an objective justification for choosing one alternative over another. Multiverse-style methods (e.g., specification curve, vibration of effects) estimate an effect across an entire set of possible specifications to expose the impact of hidden degrees of freedom and/or obtain robust, less biased estimates of the effect of interest. However, if specifications are not truly arbitrary, multiverse-style analyses can produce misleading results, potentially hiding meaningful effects within a mass of poorly justified alternatives. So far, a key question has received scant attention: How does one decide whether alternatives are arbitrary? We offer a framework and conceptual tools for doing so. We discuss three kinds of a priori nonequivalence among alternatives—measurement nonequivalence, effect nonequivalence, and power/precision nonequivalence. The criteria we review lead to three decision scenarios: Type E decisions (principled equivalence), Type N decisions (principled nonequivalence), and Type U decisions (uncertainty). In uncertain scenarios, multiverse-style analysis should be conducted in a deliberately exploratory fashion. The framework is discussed with reference to published examples and illustrated with the help of a simulated data set. Our framework will help researchers reap the benefits of multiverse-style methods while avoiding their pitfalls.

Application of a Rough Set-Based Inductive Learning System

1993 ◽  
Vol 18 (2-4) ◽  
pp. 209-220
Author(s):  
Michael Hadjimichael ◽  
Anita Wasilewska
Keyword(s):  
Machine Learning ◽  
Rough Set ◽  
Presidential Election ◽  
Predictive Accuracy ◽  
Learning Algorithm ◽  
Inductive Learning ◽  
Real Data ◽  
Semantic Content ◽  
Learning System ◽  
Voter Preferences

We present here an application of Rough Set formalism to Machine Learning. The resulting Inductive Learning algorithm is described, and its application to a set of real data is examined. The data consists of a survey of voter preferences taken during the 1988 presidential election in the U.S.A. Results include an analysis of the predictive accuracy of the generated rules, and an analysis of the semantic content of the rules.

Sign in / Sign up

Export Citation Format

Share Document