Answer: A) Positive linear association
Notice how line D is fairly close to all of the points. There are two outliers that aren't really that close to the line, compared to the other points, but for the most part the points are really close to the line. So this suggests that we have linear correlation. The linear correlation is positive due to the fact the regression line has a positive slope. As x goes up, so does y.
Something like line C is only close to the far left point, and the other points are left out. The same goes for lines A and B as well. So I wouldn't consider these lines to be good lines of best fit. The idea is to try to get close to every point as best you can.