One interesting debate of note in the flipped teacher community is the order of hands on work and viewing of the videos. I first became aware of this debate while listening to the Flipped Learning Network podcast hosted by Troy Cockrum. In episode number #3 with Ramsey Musallam, Ramsey offered a critique of the mastery flipped model because he thought it encouraged plowing through content. He also argued that science classes lend themselves to the inquiry model of instruction. Specifically, his major contention was that students ought to engage in hands on exploration before watching a flipped video. Students ought to have a reason to watch videos; in his model, a higher Bloom's exploration would cause cognitive dissonance. This sense of discomfort would provide motivation to watch the video, which would provide tools to successfully complete the exploration and extension activities.
Recently, a study at the Stanford Graduate School of Education further supported Ramsey's work. A group tested the impact of sequencing of direction instruction and hands on exploration. The results demonstrated that students, who engaged in exploration prior to direct instruction, outperformed those who participated in the same activities in reverse. The speculation is that students who engage in hands on work prior to instruction are familiar with and have built context for facts delivered in videos or textbooks. My interpretation is that students generate a schematic or framework when engaging in these activities and even though they don't master the content during exploration, these students are primed for the content. They are also able to relate the video or text content to their prior experience from the exploratory activity.
A summary of the study