/trump-exposure-code

Primary LanguageR

Deplatforming Donald Trump Reduced His Media Coverage But Not Audience Engagement

This repository stores the needed data and files for replicating our paper.

Files

  • data/: stores the data for replication
  • code/: stores the code for replication
  • README.md: this file, providing an overview of the repository

Replication

  1. Clone / download this repository, and cd into the directory
  2. Go to the directory of code/, e.g., cd ./code/
  3. Run the R-script in replication.R to replicate the results, each replication will be commented with its DV and Platform

Example Results

R version: 4.4.1 (2024-06-14)

Loading required package: Matrix

Attaching package: ‘dplyr’

The following objects are masked from ‘package:stats’:

    filter, lag

The following objects are masked from ‘package:base’:

    intersect, setdiff, setequal, union

...

============= Replicating: Media Coverage / Instagram =============

Call:
   felm(formula = eq, data = df) 

Residuals:
     Min       1Q   Median       3Q      Max 
-0.39837 -0.07515 -0.01398  0.05269  0.60592 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept)  0.19013    0.03512   5.414 1.72e-07 ***
g            0.20674    0.03355   6.162 3.78e-09 ***
s            0.16937    0.04589   3.691 0.000287 ***
l            0.13145    0.04060   3.238 0.001407 ** 
log1p(pres)  1.29100    0.13840   9.328  < 2e-16 ***
log1p(gop)   1.23594    0.15296   8.080 5.59e-14 ***
g:s         -0.17556    0.06074  -2.890 0.004264 ** 
g:l         -0.31652    0.04428  -7.148 1.54e-11 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1338 on 204 degrees of freedom
Multiple R-squared(full model): 0.7182   Adjusted R-squared: 0.7085 
Multiple R-squared(proj model): 0.7182   Adjusted R-squared: 0.7085 
F-statistic(full model):74.26 on 7 and 204 DF, p-value: < 2.2e-16 
F-statistic(proj model): 74.26 on 7 and 204 DF, p-value: < 2.2e-16 

...

Desensitization

Data in are aggregated and do not contain any sensitive information nor violating the terms of usage required by sources.

Contact

If you have any questions, please contact our corresponding author, **; or **, who maintains the repository. Their emails are provided in the paper.

Citation

  • Link to our paper: TBD
  • APA TBD
  • BibTex TBD