A study of online recommendation algorithms found that users presented with personalized suggestions consumed a narrower range of content over time, even when reporting interest in diverse material. Users without algorithmic recommendations, who browsed catalogs manually, eventually sampled a wider variety of genres. Both groups rated their satisfaction similarly, though the manually-browsing group reported encountering more unexpected items.
Based on the passage, which inference is most strongly supported?
- A
Personalization tools always reduce overall user satisfaction
- Bcheck_circle
Stated preferences for diverse content may not always translate into actual consumption patterns under personalization
- C
Algorithmic recommendations consistently surface unexpected items
- D
Manual browsing produces no benefits compared to algorithmic recommendations
Explanation
Reported diversity interest paired with narrower actual consumption supports B. A contradicts equivalent satisfaction; C and D contradict the findings.