Extended Data Table 3 Heterogeneous Treatment Effects for Exit and Voice
From: Companies inadvertently fund online misinformation despite consumer backlash
- Notes: This table shows OLS regression results for our full sample (n = 4039) where Treatment is a binary variable that takes a value of 1 if a respondent is randomized into any of the company-specific treatment groups (T1, T3 or T4) and zero otherwise. In columns 1 to 4, the dependent variable is switch in gift card choice from the respondent’s top choice company (i.e. “exit”). In columns 5 to 8, the dependent variable is clicking on a link to sign a petition that suggests that companies like the respondent’s top choice gift card company need to block their advertisements from appearing on misinformation websites. Female is a binary variable that takes a value of 1 if a respondent reports being female and zero otherwise. Biden voter is a binary variable that takes a value of 1 if a respondent reported voting for President Biden in the 2020 US Presidential election and zero otherwise. Frequent user is a binary variable that takes a value of 1 if a respondent reported using their top choice gift card at least once a month. Consumes misinformation is a binary variable that takes a value of 1 if a respondent reported using one or more misinformation news outlets (out of a list of 26 popular news outlets) in the past 12 months and zero otherwise. No adjustments were made for multiple comparisons. Robust standard errors in parentheses. P-values derived from two-sided t-tests reported below standard errors.
