The scatterplot shows how values of one topic tend to correlate with (increase or decrease at the same time as) values of another topic, among a group of places.

There could be many reasons for this correlation:

The effect is real: one topic really does influence another. For instance, unemployment and income.
The effect is due to some hidden covariate topic that affects both X and Y. For instance, the scatterplot may show a positive correlation between high property values and good health. However, property values do not directly affect health: instead, high income and wealth in a neighborhood may translate into both more expensive housing and better health outcomes.
The effect is due to random chance or magical effects. For instance, emergency department visits and the phases of the moon. Even if a correlation exists, it is safe to assume (in this world) that it is not real.

Share insights

Share insights