Inference in Linear Regression Models with Many Covariates and Heteroscedasticity

Cattaneo, Matias D.; Jansson, Michael; Newey, Whitney K.

doi:10.1080/01621459.2017.1328360

Cited by 93 publications

(98 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As estimation samples include about four hundred precinct boundaries, the total number of controls in interacted specifications is large, thus reducing statistical power. At the same time, the number of lat-long controls grows with the number of boundaries (and hence with sample size), thus potentially complicating statistical inference (e.g., Cattaneo et al, 2018). For these reasons, I limit the use of interacted specifications to robustness checks.…”

Section: Boundary Fixed Effects With Latitude-longitude Interactionmentioning

confidence: 99%

A Precinct Too Far: Turnout and Voting Costs

Cantoni

2020

American Economic Journal: Applied Economics

View full text Add to dashboard Cite

I study the effects of voting costs—specifically, distance to polling location—using geographic discontinuities. Opposite sides of boundaries between voting precincts are observationally identical, except for their assigned polling locations. This discontinuous assignment produces sharp changes in voters’ travel distance to cast their ballots. In nine municipalities in Massachusetts and Minnesota, a 1 standard deviation (0.245 mile) increase in distance reduces ballots cast by 2 to 5 percent across four elections. During non-presidential elections, effects are three times larger in high-minority areas than in low-minority areas. Finally, I simulate the impact of various counterfactual assignments of voters to polling places. (JEL D72, J15, R41)

show abstract

Section: Boundary Fixed Effects With Latitude-longitude Interactionmentioning

confidence: 99%

A Precinct Too Far: Turnout and Voting Costs

Cantoni

2020

American Economic Journal: Applied Economics

View full text Add to dashboard Cite

show abstract

“…Substantial developments have been made along this line. For example, Cattaneo et al (2018a) and Cattaneo et al (2018b) studied inferring the fixeddimension linear component in a partially linear model, and Lei et al (2018) established asymptotic normality of margins of linear and robust regression estimators in a simple linear model. Their set-up is fundamentally different from ours due to the difference of goals.…”

Section: Discussionmentioning

confidence: 99%

On rank estimators in increasing dimensions

Fan

Han

et al. 2020

Journal of Econometrics

View full text Add to dashboard Cite

The family of rank estimators, including Han's maximum rank correlation (Han, 1987) as a notable example, has been widely exploited in studying regression problems. For these estimators, although the linear index is introduced for alleviating the impact of dimensionality, the effect of large dimension on inference is rarely studied. This paper fills this gap via studying the statistical properties of a larger family of M-estimators, whose objective functions are formulated as U-processes and may be discontinuous in increasing dimension set-up where the number of parameters, p n , in the model is allowed to increase with the sample size, n. First, we find that often in estimation, as p n /n → 0, (p n /n) 1/2 rate of convergence is obtainable. Second, we establish Bahadur-type bounds and study the validity of normal approximation, which we find often requires a much stronger scaling requirement than p 2 n /n → 0. Third, we state conditions under which the numerical derivative estimator of asymptotic covariance matrix is consistent, and show that the step size in implementing the covariance estimator has to be adjusted with respect to p n . All theoretical results are further backed up by simulation studies.

show abstract

“…K-means cluster can create very tight cluster for large number of variables [18]. Whereas linear regression gives very fast results and is used to determine the dependence of one variable on other [19].…”

Section: Hybrid K Based Clusteringmentioning

confidence: 99%

Hybrid K Based Influential Parameter Determination and Design Optimization of 220 kV High Voltage Insulator

Goswami¹,

Sadhu²,

Sadhu³

2019

ELEKTRON ELEKTROTECH

View full text Add to dashboard Cite

Insulator failure generally occurs due to improper design, and if there exists sharp edges over the insulator surface. There is a need to detect the fragile parameters in the insulator design as they can cause large electrical stress, leading to insulation failure. In this paper, machine learning based approach is used to identify the fragile design parameter and thus determine an optimized design of the insulator, which will eventually decrease the voltage stress over it. A 220 kV insulator is designed and electric field intensity and potential distribution for an assembly of live, ground and insulator is calculated for different geometries. The design parameters of the insulator are considered and varied. The electric field is calculated for the whole assembly of insulator to generate the most optimized design. The optimized design reduces the overall stress by 13.12 %.

show abstract

Inference in Linear Regression Models with Many Covariates and Heteroscedasticity

Cited by 93 publications

References 40 publications

A Precinct Too Far: Turnout and Voting Costs

A Precinct Too Far: Turnout and Voting Costs

On rank estimators in increasing dimensions

Hybrid K Based Influential Parameter Determination and Design Optimization of 220 kV High Voltage Insulator

Contact Info

Product

Resources

About