why do you think some equations did a better job predicting height than others?

The reason some equations did a better job of predicting height than others could have to do with how the data was picked. A good strategy could have been to pick data points that span the entire range of heights and tibia sizes found in the classroom (rather than picking points that are all very close together on the graph). So if a group did that, they may have gotten better results than a group that simply picked at random.

Article 1 of the Constitution created the Legislative Branch.