As it happens, however, we already know why . The explanation is found in the truly seminal piece of educational research called the “Beginning Teacher Evaluation Study”. Employing intensive, repeated observations of 25 second- and 25 fifth-grade classrooms, this study found that, on average, 2 hours and 15 minutes of the second-grade school day was devoted to academic activities (which were defi ned as instruction in reading, mathematics, science, and social studies), whereas 55 minutes was devoted to nonacademic activities (such as music and art), and 44 minutes was “wasted” on things such as waiting for assignments and conducting class business.
Taking math and reading as the two primary academic subjects of interest, the researchers found that, on average, the 25 second-grade teachers allocated 2 hours and 6 minutes per day to instruction. Their students were actually engaged in learning for 1 hour and 30 minutes (or 71 % of the time). What was even more telling, however, was the fact that the top 10 % (approximately) of the teachers allocated 50 minutes more to instruction than did the bottom 10 % , and their students were actually engaged in learning these subjects for about the same amount of extra time (50 minutes).
Although this may not sound like a great deal, it means that, in these two crucial subjects, some children could receive 150 hours more instruction during a school year than other students. And, since the average amount of time actually allocated to teaching these subjects was 2 hours and 6 minutes, this means that some children received 71.4 days more instruction than others, or a total of over 14 weeks of extra schooling !
To put all of this in context, the investigators contrast two hypothetically average students, one of whom (Student A) receives a grand total of 4 minutes per day of relevant instruction and one (Student B) who receives 52 minutes. Since these students are average, they would start the year at the 50th percentile on the standardized tests, yet by midyear Student A would decline to the 39th percentile, while Student B would improve to the 66th percentile! The authors go on to justify the feasibility of their analyses as follows:
It may appear that this range from 4 to 52 minutes per day is unrealistically large. However, these times actually occurred in the classes in the study. Furthermore, it is easy to image how either 4 to 52 minutes of reading instruction per day might come about. If 50 minutes of reading instruction per day is allocated to a student (Student A) who pays attention a third of the time, and one-fourth of the students’ reading time is at a high level of success [these authors defined “a high level of success” as instruction administered at an appropriate level of difficulty], the student will experience only about 4 minutes of engaged reading at a high success level. Similarly, if 100 minutes per day is allocated to reading for a student (Student B) who pays attention 85 percent of the time, at a high level of success for almost two-thirds of that time, then she/he will experience 52 minutes of Academic Learning Time per day.
So, the moral here is that massive differences exist in both the amount of instruction that different teachers deliver, as well as in the amount of relevant instruction students receive . (We’ve already mentioned some work 50 that found that the variability in the amount of instruction received by typical students on a school wide basis can be as much as 50 % , which borders upon a criminal offense in my opinion.)
So while I haven’t seen these studies even mentioned in the value-added literature, in my opinion they constitute the only theoretical rationale of which I am aware for why we should be able to differentiate teachers who produce more learning from those who produce less of it. And by simply monitoring classroom instruction by continuously recording it on digital cameras (assuming that provisions were made for constantly identifying opportunities for improvement and then providing sufficient professional development to show teachers how to teach more intensely) we could go a very long way toward either reducing teacher differences in performance or weeding out those teachers who consistently teach less. At the very least we could combine these data with value-added procedures, which in turn might improve the latter’s present woeful ability to identify teacher differences that were consistent over time.
Read More : Beyond Value-Added Teacher Assessment
