Skip to main content
. 2022 Jul 7;12:11508. doi: 10.1038/s41598-022-15769-4

Figure 3.

Figure 3

Model-implied cross-lagged effects (preventive behavior → conspiracy beliefs) for a range of time lags in months. On average, data for approximately 3 to 4 months were available. Thus, parts of the individual trajectories are simple extrapolations. For short studies, such as 1 and 8, the cross effects at large time lags cannot be reliably estimated. Hence, the largest effects in short studies are shifted to shorter time lags than the meta-analytic curve.