Graders can improve accuracy by grading similar submissions consecutively.
Using code similarity measures can significantly enhance the accuracy of human grading in programming courses.