Fig. 5.

Data loss over the course of the entire experiment. A) The continuous experiment consists of 2 unique videos, repeated 3 times. After each video a short break is experienced during the starting up of the new video. Clear upward trends can be spotted over the course of each separate video. This upward trend is reset after each break. A slight upward trend over the course of the entire experiment can also be detected. Both 5 and 10-month-old infants showed similar data loss over time. B) The ERP experiment consists of 96 trials, with mandatory breaks after every 24 trials, during which a video is shown. Here also a clear effect of break can be seen, with stark decreases in data loss after each break, especially later in the experiment. Contrary to the continuous experiment, however, the improvement does not last for much longer than 2 trials, which indicates that data loss did not reset. 5-month-old infants seem to outperform 10-month-old infants, but data loss increased more sharply over the course of the experiment in the 5-month-old infants.