Yet the preponderance of studies demonstrates that homework conceptualized correctly does improve learning outcomes. More specifically, any beyond the most minimal doesn’t work for younger children; for older students it can’t demand too much time (the “10-minute rule” – that much for each grade daily across all classes being a practical manifestation of this); it can’t be busywork; it mustn’t introduce new things but must reinforce recent learning; and, it can’t assume completers have access to outside sources (like parents or the Internet) beyond those providing minimal assistance.
Instead, LPSD should create a random experimental approach, where this semester chance selection occurs picking classes where teachers grade homework and don’t in others (ideally, teachers will grade half their sections). When administering standardized tests, an additional coding would occur to indicate whether that student had graded homework in that subject to allow for comparison. Further, for each teacher, evaluators could compare sections with and without grading to find any significant differences in graded performance on exams.