A district comparison finds that schools with garden programs have higher science scores. The comparison does not measure parent involvement or community partnerships. Which factor could explain the observed association without assuming one variable directly caused the other?