Do Online Advertisements Increase Political Candidates’ Name Recognition or Favorability? Evidence from Randomized Field Experiments

Table 4 Treatment effect estimates and confidence intervals of Facebook advertising in study 2

Subgroup	Estimation procedure	DV = heard of candidate	DV = positive impression of candidate	DV = recall of campaign issue	DV = recall seeing online ad
All subjects	Difference-in-means and randomization inference, no covariates	0.011 [−0.036, 0.061]	0.015 [−0.033, 0.067]	0.022 [−0.024,0.069]	0.057*** [0.024, 0.088]
	Difference-in-means and randomization inference, covariate adjustment	0.004 [−0.036, 0.046]	0.011 [−0.028, 0.052]	0.020 [−0.024,0.064]	0.053*** [0.021, 0.083]
	OLS, clustered standard errors and block fixed effects	0.021 [−0.022, 0.065]	−0.000 [−0.047, 0.047]	0.020 [−0.021,0.062]	0.059*** [0.029, 0.089]
	N	3,085	3,085	2,320	2,459
Self-reported Facebook users only	Difference-in-means and randomization inference, no covariates	−0.003 [−0.069, 0.067]	0.011 [−0.045, 0.072]	0.012 [−0.051,0.077]	0.082*** [0.036, 0.126]
	Difference-in-means and randomization inference, covariate adjustment	−0.008 [−0.067, 0.053]	0.011 [−0.045, 0.072]	0.009 [−0.050, 0.070]	0.081*** [0.035, 0.125]
	OLS, clustered standard errors and block fixed effects	−0.013 [−0.082, 0.056]	−0.017 [−0.084, 0.049]	0.001 [−0.059, 0.061]	0.084*** [0.036, 0.134]
	N	1,337	1,337	1,221	1,328

Each cell records the estimate of the effect of being treated with online advertising on the dependent variable at the top of the column. 95 % confidence intervals are shown in brackets below each estimate. The first two rows in each panel employ randomization inference to estimate the uncertainty associated with the main point estimates, with the first row applying the procedure to unadjusted difference in means and the second employing covariate adjusted values. Rosenbaum (2002) 95 % confidence intervals for these results are calculated taking into account the blocked, clustered randomization scheme. The final row shows estimates employing OLS with block fixed effects, with 95 % confidence intervals calculated based on clustered standard errors
* p < 0.05, ** p < 0.01, *** p < 0.001 (one-tailed)