Case Study: SBG in a large class -- Jennifer Momsen's introductory Biology
This week, we’re back with another case study from our upcoming book, “Grading for Growth.” This case study is one answer to a common question: How can I use alternative grading in a big class?
As always, please leave us feedback about this case study in the comments. Do you find it helpful? Is there something more you’d like to know? What could make it better?
Jennifer Momsen is a professor of Biological Sciences at North Dakota State University, a large state university with high research expectations. She regularly teaches General Biology II, a survey course “focused on biological information flow, exchange, and storage.” It is taken by biology majors as well as other life science majors, and is a prerequisite for most upper level Biology courses.
General Biology II typically has 135 students in each section. Momsen teaches it in a highly structured, activity-based format. She has two or three “learning assistants” who provide students with feedback during class activities, although they do not assist with grading.
Momsen’s course is entirely standards-based. She uses a relatively small list of 15 standards that are grouped into related categories. For example:
1. Explain how genetic information is stored in living systems.
(1a) Use a model or series of models to illustrate the structures of DNA and RNA.
(1b) Create a model to illustrate the relationships among DNA, chromosome, gene, and allele.
7. Explain how biological information is communicated using chemical, electrical, and/or behavioral signals to elicit biological responses within and between individuals.
(7a) Construct a model based on evidence to explain how living systems use communication to acquire and exchange information about their environment.
(7b) Use evidence to produce a model or series of models that illustrate the core components of communication (signal/reception/feedback) and how these components of communication impact biological function at different biological levels (within an individual, within and between species).
In these examples, statements 1a, 1b, 7a, and 7b are the standards that are actually assessed.
Momsen emphasizes that figuring out a good list of standards is hard, and that her list is still evolving. Several colleagues helped her develop Momsen’s standards, beginning with the Next Generation Science Standards as a starting point. She especially recommends taking advantage of colleagues and others in your discipline for support when first creating a list of standards. Momsen developed these standards, as well as the example activity and assessment below, along with her colleague Tim Greives, who is an associate professor in the Biology department.
There are also “syllabus repositories” for some disciplines that contain examples of course designs and lists of standards for commonly taught courses.1 These are excellent places to begin when creating your own list of standards.
Momsen describes her overall goal for assessments this way:
[M]y focus was on creating opportunities for failure followed by success. This meant really coupling formative assessment with feedback. In addition, I incorporated many ways for students to retest each standard, whether in writing or verbally during office hours.
Students complete in-class activities in small, permanently assigned groups. They submit a group paper for feedback. This is purely formative feedback, and provides students with their first opportunity to practice with new ideas. Momsen recommends this approach for giving feedback in large classes: “I can give feedback on 45 papers from one class period to the next, but I can’t do that for 135 individual students.” Students have time in class to review their group’s feedback, and Momsen also gives overall feedback from group activities in class.
Here’s an example of one part of an in-class formative activity that addresses standard 7a, which we saw above. After a pre-class video lecture about biological information and feedback systems, students see a specific example related to chin coloration in one species of lizard.
Activity 2. Building a model of communication
Create a model that describes how biological information within a male lizard is communicated to melanin-producing cells
In this case:
What cell is the emitter/sender?
What cell is the receiver? What is the receptor?
What is the signal?
What biological change induced by the receiver cell?
Use your responses to the above questions to create a model.
This is followed up with additional activities that involve extending the model in specific ways, and using it to make predictions.
Students also have purely formative weekly homework – sometimes in the form of a short quiz, or some extended-response items, all posted on a learning management system. Momsen provides a rubric or answer key so that students can self-assess after completing the homework. After attempting these, students fill out a short reflection form (similar to what we read about in Joshua Bowman’s SBG Calculus 1). Again, only completion of the reflections matters. Momsen uses these reflections for whole-class feedback:
With [over] 100 students, I did not provide feedback individually, but would review the responses and then process globally in class the next class period (e.g., I noticed a lot of folks struggling here, let's pause and talk about that).
Momsen’s only summative assessments are take-home “problem sets” (in particular, there are no traditional in-class timed exams or quizzes). There are four or five problem sets, which are assigned every few weeks on a schedule clearly communicated in the syllabus. Each item on the problem set is a detailed and authentic question that is aligned with one of the standards. Each standard is graded for proficiency: “Met” or “Not yet met” based on all relevant work on the problem set. Problem sets each cover four standards, with a mix of new and repeated standards. As newer standards are added to problem sets, older ones “fall off.”
Here is an example of a question from a problem set. This assesses standard 7a, which we saw above. The question is similar to–but in a different context from–the previous in-class activity example. The overall goal is to provide a deep and authentic assessment of the standard, and so the question is detailed and involves each part of the standard.
Students were earlier provided background information about the FOXP2 gene and animal vocalizations, which isn’t reproduced here.
Rodents can make vocalizations that humans can hear, such as a squeal. But they also can communicate during social interactions using ultrasonic vocalizations - vocalizations at a frequency higher than the human ear is capable of hearing. Sex differences in vocal communication, similar to observations in songbirds, are observed In rats, with adult male rats eliciting a typical type of ultrasonic vocalizations not observed in females.
As noted elsewhere in this Problem Set, FOXP2 is involved in the development and neural control of vocalizations in a broad spectrum of species, including rodents. In humans it is linked to human speech and language disorders. In rats, FOXP2 has been related to sex-specific vocalizations, like the male-typical ultrasonic vocalizations. FOXP2 expression in neurons in the brain is greater in males than females. Further, recent research demonstrated that FOXP2 is a target of activated androgen receptors (i.e. specialized receptors that bind androgens such as testosterone and di-hydrotestoterone) (Bower et al., 2014).
Create a model that describes how biological information within a male rat contributes to the development and production of the behavioral phenotype of male-typical ultrasonic vocalizations.
As you work through the model, be aware of the the following:
What cell is the emitter/sender?
What cell is the receiver? What is the receptor?
What is the signal?
What biological response is induced by the receiver cell?
How does this biological response contribute to the development and production of male-typical ultrasonic vocalizations?
Because these are significant and authentic problems that are assessed after multiple opportunities for formative feedback, it makes sense to require a standard to be met only once.
Momsen creates an assignment for each standard in the course’s learning management system, and students upload their work to the corresponding standard. This ensures that all of a student’s attempts on a standard are visible together in one place.
Interestingly, these problem sets only have “suggested deadlines,” rather than hard due dates. Momsen guarantees feedback only if students hand them in on time, and she has found that students generally used this ability wisely. Few, if any, abuse the deadlines or get too far behind.
Momsen also makes the first problem set very small–it covers just one standard–and assigns it early in the semester. The goal of both of these features is to help students get used to SBG and build habits for success in the class.
Momsen offers two options for reassessments. One is the usual form for SBG: Subsequent problem sets include new assessments of previous standards. Students only need to attempt problems covering standards that they have not yet met. Momsen emphasizes that, to keep grading manageable, it’s important to limit the number of standards on each problem set. There is one final problem set that is comprehensive, offering a new opportunity on every standard (although by this point, most students don’t need to attempt many of the standards).
The other reassessment option is an office hour meeting. Office reassessments are available in four or five specific weeks during the semester. These are the weeks when a problem set is neither assigned nor returned. Students can sign up for a 10 minute meeting to reassess 1 or 2 standards. When they sign up, they must also fill out a “pre-reflection” with three questions:
(1) what you got wrong or omitted or were unclear about in your response on [the problem set];
(2) which class weeks correspond to this material; and
(3) how you would change your answer.
The goal is to help students prepare for the reassessment, and also ensure that students know which material to review in preparation for it. At the reassessment, Momsen provides a new problem that the student attempts “live.” Momsen determines a new mark based on the office hour work and the pre-reflection (“met” or “not yet met”). She notes that most students choose to attempt new problems on a subsequent problem set, which keeps office hour reassessments from taking up too much time.
Momsen’s 15th standard is a bit different from the other 14. It has to do with “community development”--that is, taking part in the key activities that enable students to learn and succeed in the class. As with other standards, Momsen emphasizes that it’s important to work with colleagues to develop standards, and so she credits Katie Mattaini with the idea for Standard 15. Here is the standard:
Contribute to classroom community development. We are a learning community, focused on living systems. This community works when we are all prepared to explore new ideas, apply our knowledge, and learn from mistakes. Meeting this standard involves both:
Individual accountability. Reading text, watching videos, and successfully completing knowledge checks prior to class meetings.
Group accountability. Engaging in class meetings by contributing ideas, critically thinking, and responding to feedback.
This standard is met when a student completes a total of at least 80% of the otherwise ungraded assignments, including attending class, completing formative homework, and taking part in group activities. By combining many different forms of participation, students have a variety of ways to meet this standard and also some cushion for when the semester gets busy, or unexpected issues arise.
Final grades are based purely on the number of standards met on at least one problem set by the end of the semester. Specifically:
F: 0 standards met
D: 1-3 standards met
C: 4-7 standards met
B: 8-11 standards met
A: 12-15 standards met
North Dakota State University only assigns whole letter grades.
Notice that this final grading scale does not set 60% performance as a failing grade. Momsen does this intentionally, inspired by Joe Feldman’s discussion in Grading for Equity about the disproportionate penalty provided by a 60% failure line. As she says: “An F is not doing anything at all,” but beyond that the scale is focused on representing what students have accomplished.
How do students react?
Students are initially stressed - this is a very novel system - but that quickly fades as they realize no single test or test question is going to tank their grade in my course.
Momsen also emphasizes that this puts students in charge of their own learning. With flexible due dates, students can choose what to focus on.
Students appreciated this flexibility - they said outright that they didn’t study or even turn in a problem set because they just didn’t have time, but they felt safe doing so knowing they could pick it all up on the next problem set.
This may sound strange, and it requires the instructor to modify their mindset as well. Students know what is required of them, and can make informed decisions for themselves based on their own circumstances. In other words, they take much greater responsibility for their own learning. The group tasks, homework reflections, and expected deadlines all help Momsen identify students who are falling behind or need encouragement, which leads to more targeted and useful interventions.
What about academic dishonesty? With problem sets taken home and submitted online, there could be plenty of opportunities for cheating. Momsen hasn’t had trouble with this. Indeed, she grades one problem at a time, and notes that this would make copied or overly-similar solutions stick out. She says that “I generally think this approach helped academic honesty”. Why?
Informal feedback from students underscored their feeling of less pressure, less worry about losing points. They trusted in me and the new system, and responded in kind. I do think that the alignment of class activities to the problem sets also helped. Students knew that what they did in class was what they would do for the problem set, just in a new context. They could use those class resources, too - the notes, the assignments, all of it.
Students also reported that they liked coming to class. One student wrote “I found myself learning more since I felt free to make mistakes and did not have to worry about penalties from them.”
What advice does Momsen have for others teaching a large class like hers? First: Keep assessments at a manageable size. You don’t need to allow students to reassess every standard on every problem set!
She also advises instructors to keep the number of standards small. Her 15 standards are about as large as she recommends. Keep in mind that “standards” are what you assess, which is different from finer-grained “learning objectives” you might cover during class. Momsen says: “A lot of learning objectives go into a single standard.” This is one reason why problem sets include only longer, authentic questions: They allow Momsen to assess a broader standard thoroughly.
Momsen highly recommends using small groups in an active learning class format. The groups help enable formative feedback and keep it at a manageable scale. Notice that the groups are permanent: They can become another form of support for students in and out of class, and enable opportunities for peers to help each other “iron out” individual difficulties. More than one of our interviewees who teach large classes recommend using permanent, possibly assigned groups in this way.
As described in the introduction, Momsen recommends using as much help and support as you can. She especially advocates for in-class Learning Assistants who can help with groups, although many instructors for large classes do without in-class support at all. She also encourages instructors to let teaching or learning assistants “specialize,” becoming experts in a few standards so that they can help with retesting.
Above all, “know your limits.” Keep it simple, and do what you can, in your situation, to make a difference. She is especially grateful to her colleagues and others in the field who supported her, gave feedback on her list of standards, and helped her prepare her standards-based class.
Momsen’s class shows that it is absolutely possible to use alternative grading in a large single-section class, even without help from graders. It’s worth noticing how she has carefully streamlined potential rough spots: a small list of standards keeps grading manageable; automatic reassessments on a subsequent problem set make reassessments simpler and reduces extra grading; office hour reassessments are limited to weeks when neither Momsen nor students have a problem set to work on or grade; permanent groups help enable targeted formative feedback without being overwhelming.
As always, there isn’t one right way to do alternative assessments. Think carefully about the four pillars for improving assessments and how they might apply to your situation. Momsen gives an excellent example of how to make these pillars work for her large class.
If you’d like to see our other case studies (so far!), here are some quick links:
Thanks for reading this case study! Do you find it helpful? Was there something you especially liked? What would you add or remove? Let us know in the comments!
Here are four that we know of: math, physics, biology, chemistry.
I have two thoughts for details to address in possible future posts: 1. a case where someone uses Team-Based Learning, 2. a case where someone uses flexible assignments (for example, I let students and teams self-select topics of interest and then ask them to tell me how they align with my learning outcomes and micro-outcomes... I am trying to find room to make this approach fit into an alternative grading scheme.
I am curious if you have any social science examples in your book.