Believing that one's students have potential—and treating them accordingly—helps them rise to their teacher's expectations. But measuring whether teachers have high or low academic expectations for their students in any sort of objective way is complicated, to say the least.

Several researchers have tackled this tricky issue. Some studies (here and here) use surveys to suss out teachers' expectations for their students. TNTP's The Opportunity Myth did it by having the assignments teachers gave their students independently analyzed.

A new study by Seth Gershenson, released early this month by The Fordham Institute, takes a different tact in which he measures the relationship between teachers' grading and student learning. For all the students who earned a B in a teacher's class, Gershenson calculates the average score on an end-of-course test—one that was outside the teacher's control. Teachers whose B students performed relatively poorly on the end-of-course test were deemed to have "low grading standards," while those whose B students earned a higher score were deemed to have "high grading standards."

The big takeaway is that students whose teachers have higher grading standards learn more. What's more, they have a lasting impact, with students who had teachers with high standards in Algebra I continuing to have higher math performance for at least the next two years. And these findings hold true for students of all racial/ethnic groups examined, and in all school settings. This study makes a compelling case that teachers do their students no favors when they lower the bar—at least if the primary objective is for students to actually learn something!

We offer another research question worth testing. Teacher prep programs themselves tend to give very high grades to their teacher candidates. Does attending a prep program with higher grading standards lead to those teachers using higher standards with their students?