Replication Data for: The Shadow of Social Desirability Bias: Evidence from Reassessing the Sources of Political Trust in China

Recent scholarship has largely focused on developing innovative methodologies to mitigate preference falsification in public opinion data within authoritarian settings, but little systematic scrutiny has been dedicated to assessing biases in regression analyses. By conducting both analytical and empirical investigations into the sources of political trust in China, we unveil three novel insights into this issue. First, assessing whether social desirability bias leads to over- or underestimation in regression analysis is inherently challenging to determine ex ante. Second, our statistical analyses of two nationally representative surveys in China, one of which featuring a list experiment, cast doubts about the positive effect of social welfare expansion on political trust. Expanding our analysis beyond social welfare and the Chinese context, we consistently find similar biases in regression analyses relying on direct questions compared to that of list experiment. Third, we demonstrate that some identification strategies could alleviate bias stemming from regression analysis of public opinion data when direct questions are employed.