A linear model predicts that a company will sell 50 units in a given month. The company actually sold 42 units that month. What is the…

The residual is actual minus predicted: $42 - 50 = -8$. It is negative because the company sold fewer units than the model predicted. A point whose actual value is below the predicted value lies below the model's line, since the line gives the predicted value at that $x$. That makes C correct. Choice A has a positive residual, which would mean the company sold more than predicted. Choice B correctly gets the sign but misplaces the point above the line. Choice D is wrong because a residual of zero would require the actual and predicted values to be exactly equal.