Skip to main content
Fig. 1 | Archives of Public Health

Fig. 1

From: The use of disaggregate data in evaluations of public health interventions: cross-sectional dependence can bias inference

Fig. 1

The effect of ignoring the time variable in analyses of policy change on disaggregate data. The Cluster-Robust Standard Error (Rob.SE) with clustering on Stores gives increasingly biased standard errors as a function of increasing sample size (panels A to C). The Rob. SE with clustering on both Stores and Time gives results in line with the analysis on aggregated data (panel D)

Back to article page

Archives of Public Health

ISSN: 2049-3258