rating by genre: a 16% spread

Breaking rating down by genre on movie-ratings, because the headline average is 3.62 and no segment is actually there.

• romance — mean 3.96, median 4.00 (1,950 rows)
• documentary — mean 3.73, median 4.00 (675 rows)
• horror — mean 3.60, median 3.50 (832 rows)
• comedy — mean 3.60, median 3.50 (1,330 rows)
• drama — mean 3.57, median 3.50 (843 rows)
• animation — mean 3.54, median 3.50 (459 rows)

Top to bottom that is 3.96 against 3.41, a spread of 16.1%. The pooled average is 3.62.

select genre,
  count(*)                                            as rows,
  round(avg(rating)::numeric, 2)                      as mean,
  percentile_cont(0.5) within group (order by rating) as median
from movie_ratings
group by 1
order by mean desc;

The groups are close enough that the pooled average is a fair summary. That is worth confirming rather than assuming: the check costs one query and the failure mode is invisible.

Notice the mean and median columns disagree only slightly here. Always compute both in the group-by. The comparison between them per segment is free and tells you whether you are looking at a level difference or a tail difference.

This is the setup for Simpson's paradox — the case where every segment moves one way and the total moves the other.