Fig 1.
Initial screen when a DoodleTube session begins, with identifying images and information removed.
In this instance, the participant had been randomly assigned to a group in which the order of the videos favored candidate Bill Shorten. The red-outline box above was not shown. To view the unblocked image, visit https://aibrt.org/downloads/VME_PLOS_2024/Fig1.jpg.
Fig 2.
Screen that appears after the participant has clicked one of the videos shown in Fig 1, with identifying images and information removed.
The red-outlined box above was not shown. To view the unblocked image, visit https://aibrt.org/downloads/VME_PLOS_2024/Fig2.jpg.
Fig 3.
Ordering of videos in Experiments 1 and 2.
A: Experiment 1, Group 1 (first bias group, 40 videos in order from Pro Morrison to neutral to Pro Shorten). B: Experiment 1, Group 2 (second bias group, 40 videos in order from Pro Shorten to neutral to Pro Morrison). C: Group 3 (control group, 40 videos in which Pro Morrison and Pro Shorten videos are alternated). D: Experiment 2: The manipulation is masked by swapping the video in Position 2 with the video in Position 39, and the video in Position 3 with the video in Position 38 (shown only for Group 1, Pro Morrison bias group).
Table 1.
Experiment 1: Pre and post opinion ratings of favored and non-favored candidates.
Table 2.
Experiment 2: Pre and post opinion ratings of favored and non-favored candidates.