Standards-Based Testing

How to drop alternative grading into coordinated or traditionally graded classes

Photo by Jeremy Bishop on Unsplash

Continuing the theme of “small alt-grading”, this week I’ll describe a way you can add alternative grading into almost any class: Standards-Based Testing, or SBT. This works even in classes that are built on traditional points and partial credit, are coordinated, or where you don’t have full control over assessment decisions.

What is Standards-Based Testing?

Standards-Based Testing, or SBT, is a variation on Standards-Based Grading that focuses only on one kind of skills-focused assessment, usually tests or quizzes. It’s also known as Mastery-Based Testing.

In a class that uses SBT, each test is organized around a few clear standards that represent critical concepts in the class. These are typically broad standards, each one covering the same amount of topics as one section of a textbook. Each standard is assessed on one page of the test, perhaps containing multiple questions. Students earn a single mark on each standard – usually “Satisfactory” or “Not yet” – based on whether they’ve consistently demonstrated a thorough understanding of the standard across all questions on that page.

For example, an SBT class might have 16 standards and four midterm tests. The first test has one page of questions for each of the first four standards. The second test has questions covering four new standards, and new questions on the previous four standards. The third test has four new standards and new attempts at all 8 previous standards, and so on. Students typically need to earn satisfactory once per standard; after that they can ignore the standard on future tests.

By including new attempts on subsequent tests, SBT builds reassessment without penalty directly into the system in a streamlined way. This simplicity is one of the big advantages of SBT.

The unique feature of SBT comes in the final grade calculation. A student’s “exam grade” can be calculated as the percentage of standards they’ve completed. For example, a student who earns satisfactory on 14 of 16 standards would have an exam grade of 14/16 = 87.5%. Then this percentage can be used in a traditional weighted average calculation, along with other items such as problem sets, quizzes, online homework, etc.

Some instructors don’t calculate this percentage directly, but rather make a lookup table that lets them adjust the percentages to be less punitive, or map those percentages to points. For example, a class with 16 standards might use this table:

This allows you to give a bit of grace – such as allowing students to miss one standard and still earn a high test grade – or set a certain bar for earning a passing grade.

SBT classes typically aim to give students at least one attempt on every standard before the final exam. Then the final exam includes no new standards, but does have new questions on every past standard. It basically acts as one last chance to reassess.

Instructors may identify a small subset of “core” standards that students are required to “recertify” on the final exam. This addresses concerns about students who succeed on a standard early in the semester, and then forget it later. Work on these core standards can be used to modify the final grade. For example, if there are four core standards, earning satisfactory on all of them might result in the final grade increasing with a “+”, e.g. B to B+, or adding a certain percentage to the overall exam grade. If 0 or 1 are satisfactory, that could drop the final grade with a “-” or subtract a certain amount from the exam grade. It’s wise to leave a wide range in between that doesn’t change the final grade, since final exams are usually a high-stress setting that can’t be reassessed.

There are endless variations on SBT, which can be tweaked to fit many circumstances. For example, if one standard is especially difficult, you could give a quiz between regular tests, covering just that one standard. Likewise, instructors can choose to allow reassessments in other ways, such as scheduling an office hour meeting to attempt a new question. Some instructors use a three-level set of marks, including an option like “Revisable” that specifically allows students to revise their past work if it showed only minor errors, rather than having to make a new attempt. The goal of many of these variations is to reduce the number of standards that students need to attempt on later tests.

What are the advantages of SBT?

For instructors, SBT is especially easy to add into traditionally graded or coordinated classes, without changing anything else. The SBT grade can simply be “dropped in” to a traditional grade calculation. This also makes SBT a good choice for instructors who face required percentage ranges for each letter grade, a requirement that’s otherwise incompatible with alternative grading. 

SBT makes it possible to convert only the tests (or quizzes, etc.) of an existing class to use alternative grading, a much less daunting process than converting all assessments. In addition, SBT streamlines the reassessment process, by focusing it entirely on regularly scheduled tests. Instructors need to come up with multiple new questions for each standard, but they can predict the timing and workload involved in grading these new attempts.

For students, SBT brings many of the same advantages as other forms of alternative grading. One large group of co-authors (Collins et al., see below) created a survey about student experiences with SBT and administered it across a wide range of SBT math classes at multiple institutions.

In the survey, students strongly agreed that SBT helped them understand the material deeply and prepared them to address a variety of problems. Students also overwhelmingly agreed that SBT was a fair assessment of their knowledge (over 80% agreement), and strongly disagreed that they attempted to memorize solutions to past tests as a means of studying.

In written feedback, the authors’ students also described how SBT benefitted them by encouraging them to keep studying until they learned material, rather than treating tests as “one and done” assignments. They also cited reduced stress during tests. The authors quote one student who described how SBT

… alleviated a lot of the pressure brought upon me as a college student from exams normally, and also re-enforced learning concepts I had difficulty mastering in a different way, helping to solidify knowledge of concepts that I did not understand as well. (Collins et al., p. 455)

What if…

Here are some common concerns about SBT:

What if I don’t use tests? The name “Standards-Based Testing” can be a bit misleading. This approach works well on many kinds of assessments. In addition to tests, I often use SBT with small weekly quizzes, adding only one or two standards (and letting older ones “drop off” until the final exam).

But SBT works beyond that too. For our book, I interviewed Jennifer Momsen who uses a system that is essentially SBT with biweekly homework problem sets in a large (135 person) Biology class. Essentially, any type of assignment that can be broken up into separate questions about clearly defined standards could be used with SBT.

What if I don’t teach intro level courses, or my courses go beyond discrete skills? SBT isn’t for every situation – nothing is. But SBT can be used in more places than you might expect. It works best with topics that can be divided up into discrete, independent standards. This often makes SBT fit best in introductory courses. But, many classes include some discrete skills that must be learned and practiced. Since SBT is only one part of the assessment system, you can decide to use SBT only to assess discrete skills. For example, some instructors have used SBT with very high-level math classes like Real Analysis to test essential discrete skills, while also using other assessments such as proof portfolios that test higher-level conceptual learning and synthesis.

What if I don’t want to give standardized tests? Be careful not to confuse SBT with “standardized testing”, which is a completely different idea. Standardized testing is a requirement that many K-12 schools give the same test to all students in a certain grade for comparison and evaluation across classes and districts. This is quite different from SBT, in which an instructor creates tests for their own class, with a focus on learning and growth.

Where can I see some examples of SBT?

SBT seems to be most commonly used in math courses. This might be due to the influence of George McNulty, a math professor at the University of South Carolina who helped popularize it (and mentored many grad students in its use). However, under a wide variety of names, SBT has been used in a variety of disciplines. Here are a few citations to get you started:

J. B. Collins, A. Harsy, J. Hart, K. A. Haymaker, A. M. Hoofnagle, M. K. Janssen, J. S. Kelly, A. T. Mohr, & J. OShaughnessy (2019): Mastery-Based Testing in Undergraduate Mathematics Courses, PRIMUS, 29:5, 441-460. https://doi.org/10.1080/10511970.2018.1488317. Many of these authors have subsequently published additional papers about their experience with SBT and various studies and surveys of SBT classes.

J. Ring (2017): ConfChem conference on select 2016 BCCE presentations: Specifications grading in the flipped organic classroom. Journal of Chemical Education, 94(12), 2005-2006. https://doi.org/10.1021/acs.jchemed.6b01000 While this is called “specifications”, the actual system used is essentially SBT.

In addition, see our interview with Jennifer Momsen about her large intro Biology course that uses a system that's closely related to SBT.¹

Final thoughts

If you teach a constrained or coordinated class, or want to give alternative grading a try without jumping all the way in, SBT is a good way to dip your toes into alternative grading. This is what I almost always do when I teach a new class, especially if it’s a primarily skills-based class where I can come up with a list of standards relatively easily. SBT lets me focus on teaching the new class and learning about the tough spots and difficult topics, while giving my students some amount of grace in their assessments.

As always: find what works for you! If SBT sounds interesting, take a look at the resources linked above or ask for more suggestions in the comments. But if you think SBT doesn’t sound right for your situation, then that’s fine too. There are dozens of variations on alternative grading, and you can find a different one that fits your situation.

1. Momsen uses only problem sets as assessments, which means that you could more accurately call her grading system “pure standards-based grading” (since the SBG portion isn’t inserted into a traditionally graded structure). But the shape of the structure is the same as SBT: Using broad standards on regularly scheduled assessments, where reassessments are built in through new attempts. Plus, I’m not too interested in policing terminology!