Introduction

In this article, we present a new grading model for teaching comparative vertebrate anatomy and other terminology-heavy STEM courses. This grading model arose from necessity during the early months of the Covid-19 pandemic. Offering a virtual Comparative Anatomy course with traditional exams, quizzes, and laboratory practicals seemed inadvisable, considering the challenges of maintaining course integrity and fairness in an online format. For traditional assessments to be fair, they must be rigorous, have limited vulnerability to cheating, be unbiased with respect to a student's personal situation and demographics, and be accurate measurements of a student's mastery of the material ( Close 2009 ). Proctoring software that monitors a student's computer and testing environment can be effective for reducing some of these issues for remote learning. However, it can exacerbate student testing anxiety and is biased in favor of students who have a stable, private location in which to take assessments. Video-based proctoring tools can even be biased against darker skin tones ( Yoder-Himes et al. 2022 ), which was a particularly grave concern for Farina, the first author, who is a biology professor at Howard University, a Historically Black College and University (HBCU).

This article is a contribution to the January 2025 Society for Integrative and Comparative Biology annual meeting symposium titled Integrative Organismal Biology at HBCUs: Highlighting the effectiveness of HBCUs in training the next generation of organismal biologists , during which faculty, researchers, and students highlighted some of the educational and research strategies that we employ at our home institutions. The uniqueness of an HBCU setting must be underscored. HBCUs are the only colleges and universities specifically created to support the education of diverse Black students from the African Diaspora. With the HBCU mission in mind, instructors make a concerted effort to meet the needs of a student body with different degrees of college readiness from a broad socioeconomic and cultural backgrounds, where over 54% of students are Pell Grant eligible. At the same time, HBCUs like Howard Univer-sity use the talent development model, which was created by Howard professors and researchers and which advocates for having high expectations of students to help them achieve and develop their talents ( LaPoint et al. 2006 ).

Unique challenges to learning biology

Teaching organismal biology courses poses unique challenges when compared with other STEM courses. Courses such as Comparative Vertebrate Anatomy, Ecophysiology, Animal Behavior, Zoology, and Botany require students to learn essential terminology that they may be encountering for the first time, such as species names, clade names, anatomical structures, and physiological terms. This terminology is typically applied in a comparative context across a broad range of different types of organisms and organ systems, and students must develop a baseline lexicon before advancing to more complex topics. Rote memorization and cramming are often used by students to establish this lexicon, and they can be successful short-term strategies for doing well on traditional assessments such as quizzes and exams. While some level of memorization is needed for students to acquire the language of biology, one-time cramming and one-time rote memorization for a quiz are not successful strategies for long-term retention and deeper understanding, and students generally show a preference for deep learning ( Ahmed and Ahmed 2017 ;Spicer et al. 2019 ).

A common strategy for assessing student learning of anatomy is through laboratory practicals, during which students rotate through a series of stations and identify structures on anatomical specimens. This was the strategy used by Farina in her Comparative Anatomy of Vertebrates course prior to 2020. However, practicals were a source of stress both Farina as the instructor and her students, and she found that students took the practicals without adequate preparation, with sometimes up to a third of the class failing with no opportunity to continue engaging with the content before moving on. Practicals were also very difficult to implement virtually during the Covid-19 pandemic. In the section titled "Execution Knowledge," we will discuss a strategy that Farina created and introduced to provide scaffolding for students. This strategy, which she labelled "minipracticals" was implemented after return to in-person instruction. In this approach students can repeatedly take much shorter (5 question) practicals, with open specimen study time between attempts. Farina found that students who failed on the first mini-practical attempt became highly motivated to spend time studying the physical specimens. The growth and learning of these students were evident within 2-3 h of focused lab study time. She observed that within her course instead of one-third of the class failing and never learning the content, most of the students were able to master the material and demonstrate their proficiency in a low-anxiety environment. This begged the question, how can students learn terminology while building to higher levels of understanding? One solution is to redesign assessments and grading schemes to incentivize repeated interactions with the material over time, allowing students to progress from learning basic terminology to tackling more challenging conceptual material more quickly and effectively.

Attending to and framing student learning Anderson (2005) described three stages of skill acquisition. The first is the "cognitive stage," in which participants acquire "declarative encoding" through rehearsal, and where the retrieval of facts is needed to solve problems. However, knowledge remains in declarative form as facts, which makes it difficult to use that information to solve problems. The second stage of skill acquisition is the "associative stage," where continued application leads to learning as the participants make connections and, with repeated errors, can perform better. The third stage is the "autonomous stage," in which procedural knowledge is automated and used with little effort. Unfortunately, the nature of testing and grades does not support skill acquisition even when the tasks and assignments are designed to facilitate growth. Students usually have one test and one opportunity that informs them of a particular skill or knowledge acquisition without further opportunity to grow. McGuire and McGuire's (2015) work with college students in chemistry found that students generally operated in "make an A" mode rather than in learning mode and so were often driven to memorize information. McGuire also found that students' perceptions of studying did not adequately support their learning.

McGuire provided students with learning strategies to support metacognition and build self-awareness and self-regulatory strategies. One of these strategies recommended was to give students Bloom's taxonomy as a tool for learning. Similarly, Quinlan (2012) found that the students in her high school biology class displayed an increase in the connections they made with correct information and biology concepts immediately after they were introduced to the biology concepts but reverted to using keyword searches and incorrect connections 90 days later. Quinlan's study had important implications for the use of schema-based instruction, which involves facilitating students' building and use of a mental model to solve a problem, and for Marshall's (1995) schemabased knowledge types. Marshall found that students Knowledge (IDK) Elaboration Knowledge (ELK) Planning Knowledge (PLK) Execution Knowledge (EXK) What type of IDK do students exhibit? What patterns do students recognize? What schemas are activated? What is the evidence that this particular schema is activated? What did students have to identify in the task? What is the situation, event, or experience? What are the main features of the situation or event? What declarations were made? What specific examples of individual's experiences were made? What general abstractions were described from these experiences? What verbal and visual information is retained? (What is the evidence of retention?) What details from current experiences were fit into a template about the situation? What template was created about the situation? What hypotheses were formed through application of IDK? How do students evaluate hypothesis to determine if sufficient evidence exists to warrant recognition? What new verbal and visual details were students expected to learn? How do students acquire both aspects of ELK? What situations were drawn on/heeded to be elaborated? How can a schema be used to make plans and create expectations? How can the schema be used to set up goals and subgoals? What knowledge is acquired from experiences in using each schema? What necessary knowledge did students develop to help them examine and understand the problem? To what extent did students gain a perspective i.e., broad perspective of making plans that took them from the beginning to the end? What is the evidence that students have PLK and not just IDK and PLK? What knowledge is required for students to carry out the final steps of the plan? What techniques led to actions such as performing the final skills required of the end goal/end task?

To what extent did students understand when and why to carry out various plans?

Guiding questions from Marshall's (1995) schema depiction and knowledge types.

displayed four knowledge types in mathematics-(1) identification knowledge, which pointed to students' pattern recognition formed usually during initial learning or rehearsal, (2) elaboration knowledge in which they abstracted from their experiences or were able to form working hypotheses, (3) planning knowledge, which can be likened to procedural knowledge and the acquisition of schema, which is a better indicator that students have learned, and finally (4) execution knowledge, which can be likened to students' autonomy to execute the skills they are working on. Quinlan (2012 , 2019 ) has applied Marshall's knowledge types to both high school and college students' learning in science and related topics and found that these knowledge types describe students' knowledge acquisition in science. In this paper, applied cognitive theories alongside Quinlan's use of Marshall's (1995) knowledge types are used to unpack how Farina's grading approach was used to support student learning. Table 1 provides a replicated table that will frame the discussion on the grading model that follows. While this framework points specifically to student learning, its use here sheds light on parallels with the grading model created and used by Farina.

Grading model that supports knowledge acquisition

Overview of grading model

The grading model in Figs.

1 and 2 represents two courses taught by Dr. Farina-Comparative Anatomy of the Vertebrates and Organismal Biomechanics (a Course-based Undergraduate Research Experience (CURE)

). The grades represent Howard University's grading policy, which only uses letters, A, B, C, D, and F without minus (-) or plus ( + ) designations. Only A, B, and C options are shown here to reflect the focus on helping students to achieve success with a passing score above D and F. The circular diagrams show three levels or rings, which reflect the nature or type of assignments, with the outer larger ring containing the largest number of assignments and the easiest assignments. The most challenging assignments are in the center of the ring and are fewer. Each filled-in circle represents an assignment. Therefore, students work their way into the center of the ring, where each circle is colored according to the prerequisite module. Course contents are divided into modules, which are generally larger in scope than a single textbook chapter but smaller in scope than a unit that would form the basis of an exam or midterm.

Modules are typically self-contained and organized according to themes within a course. Modules are also useful for scaffolding assignments, in which students can progress through increasingly complex tasks at their own pace, establishing mastery with basic terminology before unlocking assignments that require analysis and synthesis ( Joyner and Parks 2023 ).

The students' goal is to get to the center of the ring with the most challenging assignments, if they want to achieve an "A" final grade in the course. To do so, they must score 80% or higher on each task to progress to the next. Students who complete all of the C-level as-Fig. 1 Comparative Anatomy of Vertebrates grading schematic. This graphic is provided to students to guide them through the grading system. Each module includes two quizzes and a module challenge or lab practical. The quizzes must be completed before the challenge or practical for a module can be attempted. Once a module is complete, it unlocks a synthetic assignment. Students must complete the entire outer ring for a C letter grade, the outer and middle rings for a B letter grade, and all assignments for an A letter grade.

Identification knowledge

Identification Knowledge refers to the pattern-finding mechanisms that acquire knowledge and understand-ing, at the bottom of Bloom's taxonomy. In terms of schema theory, repetition, focus, connections, or associations are some of the ways that students may acquire identification knowledge or begin to see patterns. For example, the lecture or lab quizzes focus on basic knowledge such as names of structures, function of structures, names of groups of vertebrates, relationships among vertebrates, definitions, and specific steps in processes. Questions are drawn from a larger pool of questions, so that each attempt is unique. Lecture and lab quizzes are multiple choice and test basic knowledge. They can be attempted repeatedly with instant, auto-graded feedback. The unlimited opportunities to take low-stakes quizzes encourages students to repeatedly interact with the materials and give them repeated opportunities to make connections that will impact on how they perform later. This is preferable to dropping the lowest grade, which is a common strategy for flexible grading, because students still require knowledge from the dropped assessment ( Close 2009 ).

The "multiple attempts" feature of this model also encourages students to become comfortable with "failing" or earning a low score on an assignment. It is important for students to view "failure" as a learning tool rather than as an end to what they can achieve. In the implementation of the grading model presented here, students can be encouraged to attempt quizzes before lectures, knowing that they will likely fail, as this gives them exposure to the content and introduces some essential terminology to help them better digest the lecture material. This is supported by understandings about the architecture of a schema ( Marshall 1995 ), giving students increased opportunities to make connections with the materials in the lecture. They can then reattempt the quizzes after the lecture as many times as needed to master the terminology before unlocking the challenges and synthetic assignments. The strategy of embracing failure has the advantage of greatly reducing the stigma and anxiety around earning a low grade and fostering a growth mindset instead of deficit mindset in learners and instructors ( Eckstein et al. 2023 ). It also reflects an authentic learning process of trial and error. Furthermore, it gives students insights into the efforts needed to achieve the high expectations required of them.

Elaboration knowledge

Success in the quizzes unlock "module challenges" that apply the terminology to conceptual material covered in the lectures associated with the module, including asking students to make new inferences and connections across different aspects of the material or that were not explicitly discussed in lecture. These challenges are also auto-graded with unlimited attempts from large question pools, so that students receive instant feedback and can assess where they may need to review the material before their next attempt. Module challenges include more difficult types of questions, relative to quizzes, such as "choose all that apply," fill in the blank, and matching. This strategy reflects certain elements of self-regulated learning, such as "adaptive learning" and "scaffolding," in which students progress through self-paced assignments that increase in difficulty and complexity once they master more basic material ( Yen et al. 2018 ).

"Synthetic assignments" are higher-level assignments that span the prerequisite module and often go beyond it, such as asking students to make inferences and connections using the material as a basis for creating a piece of writing or an original graphic. Example synthetic assignments for Comparative Anatomy of Vertebrates are listed in Table 2 . One revision attempt is allowed for synthetic assignments, and students are strongly encouraged to visit office hours if they do not pass a synthetic assignment on the first attempt. Synthetic assignments are the only components of the course with firm due dates (aside from scheduled laboratory practicals). However, students must plan carefully to complete the prerequisite module before attempting the synthetic assignment. Synthetic assignments may also include "Case Studies" in which students are asked to write a short essay based on a randomly assigned scientific journal article related to one or both of the prerequisite modules, which was implemented in the Organismal Biomechanics CURE ( Fig. 2 ). Students are given the following prompt: "Describe two main findings of this paper. For each finding, describe how it relates to a concept that we covered in lecture." Case studies have one resubmission attempt allowed.

Planning knowledge

The Organismal Biomechanics course is highlighted here because of the emphasis on procedural knowledge that could lead students to develop learning structures that lend itself to the development of long-term skills they are less likely to forget. These skills are not gained through memorization but require more planning knowledge no matter how small the task and lead to larger gains such as an understanding of research process or specific software skills, which could be retrieved for later use.

Organismal Biomechanics is a research-based course that is less content-heavy and has fewer modules (see Fig. 2 ). Like Comparative Anatomy of Vertebrates, each lecture module consists of two quizzes and a module challenge. Lab activities include assignments for

Table 2 Synthetic assignment prompts for comparative anatomy of vertebrates.

Prerequisite Module Assignment Prompt

Evolution and vertebrate phylogeny

Choose 16 clades that we have discussed during lecture thus far. Draw a phylogeny by hand that accurately depicts the relationships between these 16 clades (with each of the 16 clades a tip of the tree). Then, correctly label three nodes on your tree with three clade names. Your tree can be any style that you would like (e.g., branches, brackets, etc.), as long as the relationships are accurate.

Vertebrate skeletal system

The two most successful living vertebrate groups, in terms of number of species, are Tetrapods and Teleosts, with Tetrapods transitioning to land from within Sarcopterygii and Teleosts remaining in the water like the rest of the Actinopterygii. What are the major differences between the skeletons of Tetrapods and Teleosts? Describe six differences, and briefly state an advantage of each (for living either in the water or on land). Choose at least one example from each the following categories: cranial, post-cranial axial, and appendicular skeleton.

Development and Histology

Draw two calligrams representing very early stages of vertebrate development (gastrulation and neurulation). A calligram is a drawing or diagram that uses words to form the image. One calligram should be of a gastrula and one calligram should be of a neurula. In the creation of the calligrams, you must use all of the following terms. You do not need to use all terms on both calligrams: ectoderm, endoderm, mesoderm, blastopore, neural tube, notochord, gut/archenteron, somite, and coelom.

Vertebrate muscular system

Design a skeletal muscle (real or imaginary) that can perform a specific function (realistic or extraordinary). What properties would this muscle have? Include at least the following properties: function of the muscle, muscle strength, muscle speed, cross-sectional area, length of muscle, in-lever length, out-lever length, insertion point (proximal or distal?), pennation (pennate or not pennate?), and physiological muscle fiber type. When applicable, you can use relative terms such as "high," "intermediate," and "low" (no numbers required). Briefly explain why you chose each properties, in terms of how it is consistent with the muscle's function. Your essay will be assessed on whether the properties that you assign are consistent with the function of the muscle.

Functional morphology

Define countercurrent exchange, and briefly describe why it is such an effective method of exchange. Give two real biological examples of countercurrent exchange, one involving heat exchange and the other involving gas exchange. How do your two examples differ in structure and function? nervous system You decide to go out for a rowboat ride with your pet dog (or cat, or any other terrestrial vertebrate). Once you're out on the water, a small wave rolls by and gently rocks the boat. You notice that your pet is struggling to stay upright in the boat, but she manages to maintain her balance. Name three of the possible sensory systems that she uses to do this. For each sensory system, name the sensory stimulus, the sensory organ, and the sensory cell used to detect the stimuli and convert it to an electrical signal. Also, briefly describe how the sensory system would be helpful in this scenario. Skeletal lab nodule and muscle lab module Field Trip Assignment (variable, depending on the field trip opportunity).

building research skills, a group project report, and a mock scientific paper. Similarly, students must complete the entire outer ring for a C letter grade, the outer and middle rings for a B letter grade, and all assignments for an A letter grade. A-level assignments include three case studies and participation in a research symposium.

For Organismal Biomechanics lab, students begin the semester by completing three lab activities to develop their skills (LA1, LA2, and LA3 in Fig. 2 ), which consist of workshops on using R statistical software, designing and 3D printing objects, and programming motors for simple robotics using Arduino microcontrollers. After completing these three labs, students complete an asynchronous practical skills assignment, which asks them to interpret R code and uses TinkerCad's browser embedding feature to test their skills with 3D design and Arduino. Students then start a group research project that they will complete by the end of the semester.

For the writing module, in addition to a group project report, students maintain a lecture and lab journal, and they complete an independent paper in the format of a scientific research article ("mock scientific paper"). Note that the writing module is not hierarchical, unlike the other modules that have been discussed so far. Per-haps the most important writing assignment of Organismal Biomechanics is the group project report, based on the semester-long group research project. This is designated a C-level assignment because it is required for passing the course-a student cannot pass without participating in the group research project. This is an example of the "lowest tier" of assignments being utilized for essential assignments that are required for a passing grade. Despite this being considered a C-level assignment, a great deal of planning knowledge is required for students to pass this task. Some of the synthetic assignments in comparative anatomy require planning knowledge as well. For example, while a task such as designing a real or imaginary musculoskeletal system within comparative anatomy reflects an execution knowledge best exemplified by Bloom's highest taxonomy level such as synthesis and evaluation, the nature of planning knowledge depends on the degree of complexity, effort, and output required. Such a task could require students to abstract from their experiences with prior tasks (elaboration knowledge) or use patterns observed (identification knowledge), which may only implicate memory and abstraction, and a schema for combining these experiences and patterns. A design that requires more complex outcomes might require planning knowledge that implicates additional schemas depending on the complexity of students' approach, time spent, and processes for planning.

Execution knowledge

Laboratory practicals are perhaps the most difficult types of assessments to fit into this grading model. One major learning objective for most anatomy courses is to be able to visually identify structures. In the context of comparative anatomy, laboratory practicals are assessments during which students visit stations with anatomical specimens at each station, typically with questions that ask the student to identify an indicated structure and/or provide its function. Students are allowed a set amount of time to complete each station. This format is often difficult for students, especially if it is the first time they are taking a practical exam-it is an unusual format that takes some getting used to. Students often fail such assessments by overestimating their preparedness and underestimating the difficulty. They often neglect to study directly with the physical specimens that will be used on the practical. Therefore, these assessments have a high failure rate and can be demoralizing for students. Due to the difficulty of setting up an extensive practical with many specimens, it is not typically possible to give students multiple attempts on practicals. While there are some highly effective technology-driven alternatives to in-person anatomy labs (e.g., Krishnasamy and Narayan 2024 ), high-quality versions of these tools can be expensive and require strong technological infrastructure to effectively implement.

Despite the limitations of in-person laboratory practical, studying physical specimens offers a hands-on, tactile experience that can enhance student learning. Therefore, to keep this component of the course and align it with her grading model, Farina developed a strategy of using "mini-practicals." Instead of a long-form practical, students take 5-question "minipracticals" with a subset of questions that would be seen on a typical practical ( Fig. 3 ). On the day of the practical, students attend lab and sign up for a mini-practical timeslot, offered every 15 min to a group of five students. Students waiting to take the practical are offered open study time with the specimens. These short practicals are quickly graded, and students who pass with an 80% or higher have completed the assignment. Students who do not get at least an 80% score return to the lab room for at least 30 min of additional study time with the specimens. They can then take the practical again, and they can take it as many times as needed during the lab period. Questions are changed between each attempt by swapping out bones and moving tape and pins to indicate different structures.

The use of mini-practicals has been quite successful in Farina's Comparative Anatomy of Vertebrates course. Five-question practicals assess the same learning objectives and can be administered in less than 15 min. Questions can be changed between attempts by swapping out specimens and/or labels, so that each attempt is unique. Farina has found that this approach encourages students to study with the specimens and reduces the stress of assessment so that students can focus on effectively learning the material. Therefore, this new strategy for the grading schematic and laboratory practicals has led to marked improvements shown by student performance.

Implications for ungrading

Grades are used throughout our educational systems to evaluate student performance and to motivate learning, to varying degrees of effectiveness. Typical grading systems rely on the computation of a final score based on the accumulation of percentage points, which are then mapped to a final letter grade. In recent decades, the concept of "ungrading" has been increasingly used as an alternative to this system ( Blum 2020 ). Ungrading is a wide range of strategies that range from abolishing grades altogether to de-emphasizing the role of grading in a course, typically while also still being constrained by a need to assign a final letter grade ( Stommel 2024 ). These emerging strategies reflect an understanding that grades can be poor motivators for some students, do not promote authentic learning experiences, and disadvantage students who do not thrive in grade-focused courses ( Stommel 2024 ). These strategies have primarily been applied in writing and humanities courses, with limited examples from STEM (Science, Technology, Engineering, and Math) courses in higher education ( Stenson 2022 ;Bonilla and Findley 2024 ;Newell-Caito 2024 ), presumably because of the content-heavy and technical nature of STEM courses. The limited studies of ungrading in the natural sciences have shown promising results that it can be an effective strategy for promoting deeper understanding of the material ( Stenson 2022 ;Newell-Caito 2024 ). In the grading model presented in this paper, grading is deemphasized, and students can be encouraged to choose their final grade at the beginning of the semester, based on what they think they will be able to achieve. This gives students the option of not attempting the most challenging assignments of the course, if they are satisfied with a B or C letter grade. We acknowledge the limitation that we do not have a formal evaluation of this grading model, and our intention is to provide a model that can be tested and implemented in settings where instructors feel that the model is preferable over more traditional grading schemes.

Conclusion

This grading model is a combination of a variety of strategies and teaching philosophies that have been adapted for content-heavy organismal biology courses. This grading model incorporated elements of "contract grading" and "specifications grading," in which students and instructors agree upon a set of criteria at the beginning of the semester for achieving a specific letter grade outcome, and students are evaluated based on whether they meet those criteria ( Leslie and Lundblom 2020 ;Harrington et al. 2024 ). These methods are intended to set clear and achievable standards for students and create a set of criteria that can be used as a checklist for achieving the required elements for a letter grade. Furthermore, these additional efforts support the National Science Board (2018) outcomes, which show that while HBCUs make up only 2% of all institutions, 24% of Black students who obtained doctorates in STEM received their undergraduate education at HB-CUs. Despite the successes of HBCUs in training the next generation of STEM professionals and their disproportionately large support of students from economically and educationally disadvantaged backgrounds, they remain severely underfunded compared to historically white institutions ( Graves Jr. 2025 ), leading dedicated instructors to innovate with limited resources.

The importance of reflexivity

To develop the new grading model presented, Farina, the first author, first spent time reflecting on her own learning when she was a student, and she considered times when she learned terminology most effectively and retained the information for long periods of time. These were experiences when she had repeated exposure to terminology and applied it to increasing complex topics, especially during graduate research, rather than memorizing without context. She also reflected on her own experience with ADHD (attentiondeficit/hyperactivity disorder); one way in which it can manifest (as it did for her) was in a reduction in the effectiveness of grades as an extrinsic motivator ( Smith et al. 2020 ), relative to other intrinsic motivators such as curiosity and extrinsic motivators such as applicably of knowledge to solving a problem, which might even be intrinsic for some. This is what led to her interest in ungrading and self-regulated learning. Even though final grades were important, students could be provided with increasingly challenging and curiositydriven assignments, once they have been given adequate opportunity to first learn the terminology and concepts required to synthesize information. Farina noted the following in the next paragraph, which retains the first person to emphasize the importance of and influence of self-reflections in driving change in the classroom.

Once I decided to abandon traditional assessments, I also felt empowered to develop a grading model that was generally accessible to all students, including neurodivergent students and those experiencing major life disruptions, as we all were in 2020. Therefore, I sought to develop a system that fit the needs of the biology courses and of students. The positive outcomes were so promising that I adopted this strategy for the majority of my courses, even after in-person learning resumed. I continued to refine this system in the years since 2020, with much interest and enthusiasm from the members of the Society for Integrative and Comparative Biology. This and the other strategies presented in this paper have encouraged me to continue to develop a grading model that has the goal of lifting up all students, regardless of their preparation or background .