Why I flipped
ATM601 is cross-listed with CHEM601 and stacked with ATM401. In principle, these are three different classes to a certain degree. However, none of these classes would have high enough enrollment when taught as a separate class. Since these three classes have different learning goals, there are different homework assignments tailored to the respective class the student is enrolled in.
The pre-requisites for ATM401 are CHEM 105X, CHEM 106X, MATH 302 and PHYS 212X, while those for ATM601 and CHEM601 are graduate standing. ATM601 is also taken by many graduate students from engineering. Thus, there is a broad spectrum of pre-knowledge regarding to atmospheric sciences ranging from close to nothing to a BS in atmospheric sciences.
In a lecture style class, this spectrum leads to student frustrations on both ends. The students with a BS in atmospheric sciences are bored when concepts are explained that they already know from their undergraduate work. Students, who have no atmospheric pre-knowledge at all, struggle when the pace is in sync with the students, who have atmospheric sciences background. Thus, flipping offers me the opportunity to teach all three groups from the level they are at, to the level that is the goal of the respective class they are actually taking. Think of it like the one-classroom schools in rural areas.
As a first step towards flipping, I wrote a textbook (Lectures on Meteorology) with a colleague of mine (Gerhard Kramm). This textbook derives every equation from the physical or chemical fundamental basics. Thus, it is easy to see where concepts come from, i.e. they can be related to material all three groups of students should know. Since the equations do not “fall from the sky” the students, all students have anchor points they already know and can relate to.
Flipping transfers the knowledge collection to a certain degree outside the class. Thus, students who struggle can read and watch the material outside the class as often as needed. This means those who struggle feel less stressed about capturing the material and less overwhelmed than when sitting through a lecture.
I have divided the class material into 26 units. Each unit has its own page that provides the learning goal and assigns the tasks to the students plus a student learning assessment form to be filled out prior to class. For each unit I have created a short video that presents the material I want the students to learn. These videos are either Pencasts or Powerpoint Presentations run through Camtasia to put the audio on. You can find an example of a unit here.
The unit page assigns which sections of the book are to be read as homework and to fill out the Student Learning Assessment form. Since the answers are graded, this form ensures that the students read and learn the material. If they cannot answer a question because they did not understand the material to the fullest, they can send emails that provide a precise question that demonstrates that they read the material. In the email they have to state what they understood, and what not.
Since the forms and questions should be provided at 1500 AST the day before class (but I accept the forms still until onset of class time), I can see where the students had difficulties and go over the material in class, and answer the questions. Sometimes I use group discussions to clarify material. Doing so helps the students to advance in science communication. The clarification section typically is about 30 minutes.
Then the students work on applications of the learned material, either alone or in groups for 30 minutes. During that time I go around, ask where they are stuck or to explain their ideas, and ask questions to help them learn to manage the problems. Doing so gives me insight on which material the various students are struggling and to provide extra help for success/understanding/learning.
The task are text problems with a twist to prepare them for their thesis and life as professionals. These problems require the students to make reasonable assumptions based on the learned material to solve the problem. The students also have to assess how the assumptions limit the validity of the results.
I alternate between assigning applications/exercises to groups and as single work. This way I can prepare the students well for their exam and the comps as well as for their future professional life which most likely will require teamwork experience at least.
In the last 30 minutes, the students present the solutions, and I discuss the results in class. The former trains the presentation skills and helps the weaker students to understand the results. In the latter, I can target the areas that I have seen causing trouble for some of the students.
These student learning assessments make up a notable part of their grade. The other parts relate to the in-class exercises, and participation in discussions and peer-to-peer learning, and the final exam.
What I’ve Learned
In fall 2013, I started flipping some of my class hours using the peer-to-peer and the group discussion approach. I only applied it to disseminate the results of the homework assignments or in-class exercises. I had not announced in my syllabus that I would occasionally flip the class. Doing so is of great importance as students expect lecture style.
According to the student comments, about 50% like (dislike) flipped classes. I believe this (dis)like is similar to the one I have seen in the last big change in eduction in the early 2000s. Back then, about 50% of the students commented power-point lectures were too fast. However, once they got used to the new style and saw its advantage they complained about any faculty who still used the “old-fashioned’ green/white board.
In fall 2014, I announced in the syllabus that the class will be taught in flipped style. In my introductory class, I explained why I flipped and why it is of advantage for them.
The students prefer working alone because they have to do the exam and written comps alone too. However, to foster their ability for teamwork and their communication skills I alternate team and single work.
Groups of up to three students have worked best for me. It is important to ensure that all three present parts of the results so all students have equal chances to improve their presentation skills, to undermine free-riding, and to ensure they all understood it at the level to explain it to their fellow students who worked at the same time at a different task. If the enrollment allows it will be useful to create the groups yourself rather than letting the students always pick their group mates.
It is important to convey/ensure that the students come prepared to class, and that you will give fellow classmates first a chance to answer a question a classmate has prior to answering it (at a deeper level).
At the beginning, it takes some energy to get them going. Just to get an idea what I mean with this. The first time I asked them to build two groups, it took several requests. Then I assigned each of the groups a task to work on and asked them to use the board. They were very hesitant to come up the stage to the assigned part of the board.
The most rewarding part was when I realized how much more they learned as a group. I had two groups working on different assignments at the board. They were to compare two different mesoscale systems – one group was to write down what the systems had in common, the other group was to list what makes them different. When I saw they run out of ideas, which was about the same time for both groups, I switched the two groups to review the work of the other group and work on the other assignment. In the end, both lists were pretty complete. In a normal class setting, the complete list is impossible to convey as in lecture style one looses them in such a comparison quickly.
The more outgoing students and strong students liked flipping and the active learning. Shy and students on the left tail of the distribution did not like flipping at all. This group of students especially disliked being called to present the group work or parts of it to the other group.
Despite every student claimed to have a laptop and most students used their laptop to take notes during lectures they were not willing to use the google shared document option to give answers anonymously in a shared document that I intended to project for discussion of answers to questions.
When producing the videos first a script has to be written including action marks like clicking to the next page or to activate an animation. The videos have to be edited to blend the video and the powerpoint and remove noise, or other mishaps.
The Student learning outcome forms must be design in such a way that the answers cannot be googled. This means one has to formulate them that the students have to combine learned material to be able to decide whether the statement is true or false or the answer is yes or no.
It is also important that the students understand that from a statistical point of view one can get 50% of the answers correct by flipping a coin. When I brought this up my students suggested that it would be fair to require more than 50% right to get a grade better than F!