Tuesday, July 2, 2013

Concerns from a Devil's Advocate about the Data Supporting the FlippedClass

CC Image courtesy of Flipped Learning Network
If you've been following the flipped class, then you've probably observed the infographic on the left several times. Like dozens of educational reforms, the flipped class has been scrutinized for it's lack of data concerning student outcomes. I must admit that, even though I buy into the flipped class and see it's potential, I worry that some of the data is not convincing - at least not yet.

In my graduate school course, I've read a few papers regarding the flipped class. The Flipped Learning Network, Pearson and George Mason University sponsored a "Review of Flipped Learning." The paper is an extensive review of some preliminary findings and case studies.

The authors were honest at the beginning of the paper when they stated that "quantitative and rigorous data on Flipped Learning is limited...(page 2.)" In the following paragraphs, they further admit that the present research mostly "consists of teacher reports on student achievement after adopting the model" and offer that these teachers "report that their job satisfaction has improved." I have to admit a certain level of skepticism as I continued to read on. My chief concern was the conflating of different pedagogical approaches with flipped instruction. The authors did a wonderful job explaining how Active Learning, Peer Instruction, Priming and Pre-training lead to improved student outcomes. Although I do see how flipped instruction can make these other approaches possible and the overlap between flipped instruction and those pedagogies, I finished this section feeling a bit underwhelmed about the data supporting flipped instruction.

The meat and potatoes of the review paper were the case studies from Byron High School in Minnesota and Clintondale High School in Michigan.

Byron High School
CC Image courtesy of Flipped Learning Network
Bryon High School was plagued with low math scores on standardized exams. According to the review, the flipped class caused the increase in math scores seen in the graph on the right.

However, upon closer inspection, the department chair came up with a plan to replace textbooks with open source materials, rewrite the curriculum and flip their classes. It's almost impossible to tease out the actual cause for the increased test scores. Perhaps not as significant but an interesting observation, the plan to make those changes was created in 2009.  According to the graphic, Bryon already outscored the rest of the state by 16%. Although, the gap did increase beyond 2008.

Clintondale High School
CC Image courtesy of Flipped Learning Network
Clintondale was in a similar situation as Bryon with low achievement and high failing rates. The failure rates appear to drop significantly (33%) and student discipline cases dropped by 74% in two years. The results of the flipped class appear miraculous. The only question I have about the case study is regarding the graph on the left. I wonder if some of these results are actually significantly different. It's hard to interpret the data without knowing the historic trends. I would love to see the same data for the previous five-ten years to see how much the passing rates fluctuate in the traditional model.

Higher learning
The paper continued with citations of similar cases from the College of Westchester. Unfortunately, they only shared the conclusion from an electrical engineering course rather than supplying the data. However, they did share data demonstrating that the flipped class students performed better on a post test than their non-flipped counterparts. Although, I have to admit that both the flipped and non-flipped students apparently scored poorly, 31% versus 24% respectively.

Teacher & student perceptions
CC Image courtesy of Flipped Learning Network
To date, the most convincing data I've read about the flipped classroom comes from teacher and student reports. There is a bit of quasi-scientific anecdotal contamination here but the majority of flipped teachers and students in those classes report positive results and perceptions about the flipped classroom. The infographic at the beginning of this post, courtesy of Classroomwindow, suggests that flipped teachers unanimously agree that they will continue this approach and the majority report increased job satisfaction.  It's rare to get 80% of students to agree on much so I also found the positive reviews on the right intriguing.

Final Thoughts
From this post, you might get the idea that I'm against the flipped class, or at least skeptical of it. It might sound blasphemous coming from a science teacher but I'm not worried about the lack of data. The flipped class is a new grassroots and largely-undefined movement. While we get a sense that most flipped teachers offload direct instruction through video (and despite the F.L.I.P pillars), there is so much variation in this model. This variation makes it difficult to evaluate the effectiveness of the model. I suspect there are "good" and "bad" teachers flipping their classrooms. I would not expect this model to turn a "bad" teacher into a "good" teacher. Even though I criticized the authors for conflating active learning, peer instruction and other models with flipped learning, I understand why the authors used this strategy. The true power of the flipped classroom is NOT the actual flipping, but the real gift is what the flipped class makes possible! There are data supported strategies that work like Mastery Learning, PBL and Active Learning. For me, the point of flipping is to be able to use the approaches that we know to be effective in the classroom.