Given this huge caveat, the title of the article is at best extremely deceiving. It's bonkers to me that this was the comparison they chose (watched at 1x speed -> 1 week delay -> test) vs (watched at 2x speed -> 1 week delay -> watched 2x speed -> test). Full description of this is found on page 7 at https://www.researchgate.net/publication/356194270_Learning_...