Replication Data for: Regularized Regression Can Reintroduce Backdoor Confounding: The Case of Mass Polarization (doi:10.7910/DVN/VEFZXI)

View:

Part 1: Document Description
Part 2: Study Description
Part 5: Other Study-Related Materials
Entire Codebook

Replication Data for: Regularized Regression Can Reintroduce Backdoor Confounding: The Case of Mass Polarization

doi:10.7910/DVN/VEFZXI

Harvard Dataverse

2024-10-31

1

Mellon, Jonathan; Prosser, Christopher, 2024, "Replication Data for: Regularized Regression Can Reintroduce Backdoor Confounding: The Case of Mass Polarization", https://doi.org/10.7910/DVN/VEFZXI, Harvard Dataverse, V1

Replication Data for: Regularized Regression Can Reintroduce Backdoor Confounding: The Case of Mass Polarization

doi:10.7910/DVN/VEFZXI

Mellon, Jonathan (West Point)

Prosser, Christopher (Royal Holloway, University of London)

Harvard Dataverse

Harvard Dataverse

Prosser, Chris

Prosser, Chris

2024-06-13

https://doi.org/10.7910/DVN/VEFZXI

Social Sciences

Regularization can improve statistical estimates made with highly correlated data. However, any regularization procedure embeds assumptions about the data generating process that can have counterintuitive consequences when those assumptions are untenable. We show that rather than simply shrinking estimates, regularization can re-open backdoor causal paths, inflating the estimates of some effects, and in the wrong circumstances, even reversing their direction. Recently, Cavari and Freedman (2022), argued that declining cooperation rates in surveys have inflated measures of mass polarization. We show that this finding is driven by large penalty terms in their regularized regressions, which leads to the estimates being confounded with time. Alternative methods do not show a clear positive or negative effect of declining cooperation on estimated levels of mass polarization.

<strong>APSR Data Editors' Note:</strong> APSR Data Editors have reviewed included documentation for completeness and have successfully reproduced all figures and tables in the article using the code and data included in this deposit. Data editors do not review results presented in appendices or supplementary materials.

This dataset not to be distributed/posted outside of the Harvard Dataverse. All downloads should take place directly on Harvard Dataverse to ensure data integrity.

MELLON, JONATHAN, and CHRISTOPHER PROSSER. 2024. “Regularized Regression Can Reintroduce Backdoor Confounding: The Case of Mass Polarization.” American Political Science Review: 1–9.

10.1017/S0003055424000935

MELLON, JONATHAN, and CHRISTOPHER PROSSER. 2024. “Regularized Regression Can Reintroduce Backdoor Confounding: The Case of Mass Polarization.” American Political Science Review: 1–9.

Label:

american-political-science-review.csl

text/html

Label:

readme.html

text/html

Label:

readme.md

text/markdown

Label:

regularization_backdoor.bib

application/x-bibtex

Label:

regularization_backdoor.Rmd

text/x-r-notebook

Label:

regularization_backdoor.Rproj

application/octet-stream

Label:

regularization_backdoor_do_everything.R

type/x-r-syntax

Label:

regularize.bib

application/octet-stream

Label:

system_and_package_info_published_version.txt

text/plain

Label:

regularization_backdoor_appendix.pdf

application/pdf

Label:

regularization_backdoor_further_appendix.pdf

application/pdf

Label:

cf2_data.RData

application/x-rlang-transport

Label:

cf2_data_note.txt

text/plain

Label:

GSS7218_R1.dta

application/x-stata

Label:

GSS_Codebook.pdf

application/pdf

Label:

gss_var_categorize.xlsx

application/vnd.openxmlformats-officedocument.spreadsheetml.sheet

Label:

regularization lit review.xlsx

application/vnd.openxmlformats-officedocument.spreadsheetml.sheet

Label:

intermediate_data_placeholder.txt

text/plain

Label:

bayes_ridge_equiv_sim.R

type/x-r-syntax

Label:

cf2_lots_of_lambdas.R

type/x-r-syntax

Label:

cf2_reanalysis.R

type/x-r-syntax

Label:

cf2_reanalysis_tables.R

type/x-r-syntax

Label:

cv_functions.R

type/x-r-syntax

Label:

false_positive_sims.R

type/x-r-syntax

Label:

gss_cors.R

type/x-r-syntax

Label:

lasso_ridge_demo_sim.R

type/x-r-syntax

Label:

lasso_selection_sim.R

type/x-r-syntax

Label:

mp_models.R

type/x-r-syntax

Label:

mp_model_tables.R

type/x-r-syntax

Label:

power_sims.R

type/x-r-syntax

Label:

power_smoothers.R

type/x-r-syntax

Label:

random_walks.R

type/x-r-syntax

Label:

simulations_1_plot_smooths.R

type/x-r-syntax

Label:

simulations_2_plot_smooths.R

type/x-r-syntax

Label:

simulations_3_plot_smooths.R

type/x-r-syntax

Label:

simulations_4_plot_smooths.R

type/x-r-syntax

Label:

simulations_main.R

type/x-r-syntax

Label:

simulation_functions.R

type/x-r-syntax

Label:

sim_data_maker.R

type/x-r-syntax

Label:

trust_simulations.R

type/x-r-syntax

Label:

sim_data_placeholder.txt

text/plain