Answer: D.
The homes were selected at random from all homes the utility serves, so the results can be generalized to those homes. The showerheads were then randomly assigned, so the difference in water use can be attributed to the showerheads.
So each home is estimated to use about fewer gallons per month with a low-flow showerhead, and for all homes the decrease is about gallons per month.
Check: Random selection supports applying the result to all homes, and random assignment supports a causal effect of gallons per home per month; .
A: is , the decrease for only the homes that received the showerhead in the study.
B: is , the decrease if only the selected homes had the showerhead installed.
C: is , which applies the effect to only half of the homes, as if the study’s even split carried over to the whole utility.
Watch for: Random selection lets the result be applied to all homes, and random assignment lets the difference be attributed to the showerheads; multiply the per-home effect by all homes.