Some true diamonds are coloured from inclusions or structural differences. Can the prices of these diamonds be predicted with the models we make for traditional clear diamonds?
One website lists a black diamond that weighs 1.44 carats for \(\$ 3050\). Black diamonds are true diamonds but may have formed from impacts of comets, asteroids, or meteors rather than from geological pressures deep underground. They're probably best compared to the most-coloured K-quality diamonds. Here's a scatterplot of Price vs. Carat Weight for K-colour diamonds. The black diamond (shown with a red \(\mathrm{x}\) ) has both high leverage and large residual.

The same site lists a blue diamond that weighs 0.82 carats for \(\$ 16,500\). Blue diamonds are a luxury item, best compared to the best D-colour clear diamonds. Here's a scatterplot of Price vs. Carat Weight for C-colour diamonds with the blue diamond shown as a red \(x\).

QUESTION

What would be the effect of including these exotic diamonds on regression lines fit to each of these scatterplots?

Price 8000 6000 4000 2000 1.0 1.5 2.0 2.5 Carat Weight