In Part 1 of this three part series, I propose one reason why I suspect some flipped classrooms fail while others succeed. [Go to Part 1: Why Flipped Classrooms Fail] Failure is a broad term and there are many ways a flipped classroom can fail. The type of failure that causes the most tension for me is related to student performance measures, specifically when students do not fare any better on high stakes assessments than they do in traditional classrooms. I can deal with almost all other types of flip class letdowns–such as student resistance or lack of engagement — if students excel. Student success provides a colossal return on my time and energy investments in the classroom. However, when the student performance measure reads “failure” I have to question what I’m doing.
In Part 1, I suggest that one reason for flip class failures on the performance measure is that when students are studying or preparing for high stakes assessments they may not be paying attention, for long enough, to the right things. So the first question I ask myself when I encounter this type of failure is, “how and where am I directing student attention and thinking?” While acknowledging that there are tons of other types of flip class flops, I am going to stick with the student performance measure to explore another key reason why I think flipped classrooms fail.
Reason #2 – Students are not able to apply knowledge in novel situations
On high stakes assessments, especially standardized tests, we typically ask students to demonstrate their ability to apply what they know on questions they have never seen before. To me, this ability to transfer learning across situations and use it to solve novel problems is the holy grail in education. We hope we train students to take what they have learned in the context of our classrooms and apply it to an unfamiliar or new situations off campus. I propose that problems of transfer offer another reason some flipped classrooms fail on the student performance measure.
Application or transfer questions are more than procedural tasks that require plug and chug or rote problem solving. Such questions require students to transfer what they have learned to novel situations and are very common on mid-term and end-of-term exams. They are not as common in homework problems or quizzes. As such, students don’t generally have enough practice preparing for the transfer questions that are going to determine their overall course performance.
Transferring learning is a very difficult thing to do. Especially far transfer or transfer between very different contexts.* 
My favorite example of the complexity of far transfer of learning is demonstrated in the movie The Karate Kid.
To prepare Daniel to compete in an upcoming tournament, Mr. Miyagi has him wax a car using very specific movements. Miyagi varies the context in which Daniel learns the movements with the commands “wax on, wax off,” “paint the fence” and “sand the deck.”
Daniel is flabbergasted at the ridiculousness of all this and feels he is being abused. Moreover, despite Miyagi telling him that he is Karate training by doing these tasks, Daniel never makes the connection that by practicing “wax on, wax off” he is actually learning to
block using his reflexes. Miyagi has to do a big reveal by simulating the connection between the movements in an intense showdown before Daniel gets it (see image to the right).
So, if I am facing a situation where I have invested time and energy in flipping my class and students aren’t performing, another question I am going to ask myself is –“have I trained my students effectively to apply what they have learned?”
Two Strategies for Addressing Problems of Transfer
According to Barnett and Ceci (2002), there are two overarching domains of far transfer: the content domain (what is actually transferred), the context domain (when and where something is transferred to and from) and multiple subdomains within. To help address the complexity of designing for and measuring transfer across far or unfamiliar contexts in education, Barnett and Ceci developed the transfer taxonomy displayed in Figure 1.
To help students apply their learning to novel situations and far contexts I recommend embracing a two fold strategy of variability and telling.
Variability: to prepare students to apply their knowledge of a single concept in novel situations, vary students’ practice by both content and context. Miyagi employees this strategy, he puts Daniel in a variety of situations using very different content and contexts (car waxing, fence painting, deck sanding) to learn the foundational skill of blocking at a deep level. By exposing students to content and contexts with such variety, you help them strip out what is extraneous and hone in on the deep meaning of the underlying concepts, topics, or skills. But as discussed above, varying the context isn’t enough.
Telling: Miyagi is like most of us educators. He carefully and artfully sets up an elegant learning experience and sits back and waits in anticipation for the fireworks to go off. He leaves his student to make the connection that what he is learning in one context can be transferred to another context. He’s almost as flabbergasted that Daniel doesn’t see the bridge as Daniel is that he is doing Miyagi’s yard work. It is critical to teach students how to apply their knowledge to novel situations; that is to do a reveal demonstrating the various connections before the high stakes assessment.
In conclusion, so far I have offered two reasons why I think flipped classrooms fail on the student performance measure:
1) Students aren’t paying attention to the right things for long enough (see Part 1 for more on this)
2) Students don’t know how to apply their learning in novel situations.
As strategies for addressing these two issues, I suggest asking yourself how and where you are directing students attention and how you are teaching students to apply their knowledge in novel situations. In Part 3, we will broaden this narrow focus on student performance measures by exploring why flipped classrooms fail more generally through the viewpoints of several experts in the field.
Go to Part 1: Why Flipped Classrooms Fail
Go to Part 3: Why Flipped Classrooms Fail
*For a discussion of the different types of transfer, see Barnett and Ceci (2002).
Turn to Your Neighbor: The Official Peer Instruction Blog 
allisonsoult
I think a big issue is students not recognizing their own ability to learn in this way. They think that instructors need to lecture for them to learn, despite all evidence to the contrary.