I received a nice query by email the other day and I thought readers might find my reply helpful.
Within my district there is quite a debate going on about the difference between formative and summative assessments. Specifically, the administration in my school district has developed a common syllabus for our teachers this year which states that 30% of a student’s grade should be made up of formative assessments (homework, journals, etc.) and the remaining 70% should be made up of summative assessments (quizzes, tests, exams). I am on the side that is arguing that a formative assessment should never be graded – it is supposed to be used as a tool to evaluate teaching so that adjustments can be made to instruction.
I write you in hopes of getting a clear and concise definition of a formative assessment. Can they be graded? If so, how are graded and ungraded formative assessments different from one another?
I’ll need to qualify my answers by distinguishing between ‘theory’ and ‘practice’ and in terms of the meanings of the key terms.
In theory, I would define ‘formative’ assessment as “useful feedback with an opportunity to use that feedback” to perform optimally on later summative assessments. A simple example: a pre-season of 4 games in soccer (as well as ongoing “scrimmages”) in which the games don’t ‘count’ is formative assessment to get teams ready for the games that do count in the regular season. So, your comment about “use the information to evaluate teaching” introduces a slightly different purpose for me. Formative assessments provide feedback – for students first, then teachers; that’s their purpose.
A 2nd simple example, more complicated: Suppose as an English teacher I do a pre-test writing on day 1 of my course, the same test is given twice during the year, and a post-test is given on the last day of the course – and it is the SAME prompt each time (say, a writing prompt on the course Essential Question “How well do we know ourselves?”). I will most likely NOT grade the pre-test, but I might grade it while not ‘counting’ the grade, i.e. tell students where they stand (e.g. what teachers do when they use a practice AP test or use the state rubric). But I also might grade and thus ‘count’ the ‘formative’ writing prompts DURING the year as well as the post-test at year’s end IF I felt that students should be ‘ready’ to be tested on their understanding and writing ability thus far.
Another not so simple example: I give you feedback on your college essay before you submit the final ‘real’ essay to your college. I might “grade” your college essay draft with a B- as well as giving you feedback and advice on how to improve it for the final “summative” version you hand in to the colleges of your choice since we have been working on essays all year. And I might even put that B- in my gradebook as one grade of many for you in English 12 this year. (But, then, I might also raise your grade once I see the ‘final’ version you sent to the colleges and count only the final’ grade on that personal essay in my gradebook.
In short, no matter the pure definition, I don’t think it is accurate to say that formative assessments can’t ever be graded. What matters – what makes a formative assessment formative – is whether I have a chance to get and use feedback in a later version of the ‘same’ performance. It’s only formative if it is ongoing; it’s only summative if it is the final chance, the ‘summing up’ of student performance.
What’s really more irksome for me in these kinds of matters is when people utterly abuse the idea of formative by describing any non-end-of-course assessment as formative. It is completely bogus to declare that ‘homework’ and quizzes are ‘formative’ simply because they are different from quizzes and because they occur throughout the year. If the specific demand only occurs once and you can’t use the feedback from them to do better next time, i.e. if the homework and journals are unique one-time events, then that individual homework assignment is summative. Just because it is not at the end of the year or semester doesn’t make it formative. It’s only formative if it recurs as a task in which I can learn from feedback to improve at the ‘same’ task.
Now, a critic may say – c’mon, Grant: they learn from doing homework how to do homework – so it’s formative and not 1-shot. And we grade it because we want them to be accountable for homework. Fair enough, I suppose (if that is really true). Calling homework assignments formative seems like a stretch to me. The content is unique and we grade the content not just the doing of it in most cases. If the grade is just for turning it in then I might acquiesce. On the other hand, calling a journal ‘formative’ and grading it for development over time seems reasonable.
However, we can say for sure that any truly one-shot assessment – homework, journal, quiz, paper - is summative, no matter when it occurs.
Call me a cynic, but my hunch here is that the makers of this rule are doing it unthinkingly in terms of the pure ideas. They seem to be just throwing trendy language around (at least from what you sent me; that’s all I can refer to). It sounds like they care less about the true meaning of ‘formative’ assessment than they care about making sure that kids do their work and are held accountable for it. FINE! But don’t call homework and journals ‘formative’ then.
Suppose I am wrong; suppose that the point of the plan is to make sure kids do their work AND to provide more truly formative assessment opportunities for kids, AND to expand the different types of assessments to give kids more options for showing what they know and can do. Fine! But then the policy would probably be different than the one proposed, I think.
A policy sensitive to these issues and designed to give kids more opportunities to get good feedback and use it would make that more clear in the policy statement. It would make clear when you should and when you should not include a grade in a kid’s average, what kinds of ‘fair chances’ to learn they need, and that journals and homework are only ‘formative’ if kids can improve at those tasks over time and have the later grades count more than the earlier grades (or some other reasonable rule).
What is really wanted here is a full discussion of the assumption that lurks below consciousness: the longstanding thoughtlessness of averaging grades to compute interim and final grades. This is just a dumb habit that penalizes growth and over-rewards wild swings of performance – though steady growth is what we presumably value in learners. Another day I’ll tackle that beast!
Mike Barth said:
Thanks so much for clarifying the homework issue. Very interesting blog post.
Mike
Kathy said:
My school wants everything graded therefore a child can take homework home, have a parent help them with it, get a perfect score, fail miserably on the end of the unit test and still pass the course. Seems ridiculous to me. In order to get around it, I only assign 5% of the grade as HW. Therefore, my “assessments” in class really tell the story. I can’t wait until we go to standards based report cards and I can get rid of the whole letter grade thing which really tells the student and parents nothing. And I agree with you on the whole averaging thing. One poor score or one really great score can skew the whole thing. What does that tell a student or parent?
grantwiggins said:
Check out my most recent blog entry which offers part 1 of a set of policy ideas for how to handle problems like this
Kathy said:
Great ideas on your most recent entry. I absolutely love reading your blog and often forward your writing on to my principal. It might not change things right away, but at least it keeps them thinking. Thanks for everything you do for education.
Paul said:
I have been chairing a grading committee in my district for over a year now with the intention to reform our grading practices from traditional to standards-based. Our literature review has encouraged us to grade, record and report all aspects of student learning: process, product, and progress; however, each of these components need to be graded, recorded, and reported separately from one another, as opposed to being averaged into one grade. Furthermore, process, products, and progress grades representing students’ mastery of specific learning standards should be recorded and reported separately.
As I read more and more about the difference between formative vs. summative assessments, and examples of each, I find the the argument to be a semantic one. Although, there is clear distinction between the application of formative and summative assessments, the feedback students receive from both types of assessments is summative in nature. Feedback, whether you are providing students with descriptive narratives or a numerical score, represents students’ learning achievement at a particular place in time, which makes it all recordable and reportable. Now, if the feedback or grades are used to inform instructional interventions, then the summative grades are being used formatively. Formative assessment grades should be recorded and reported to improve communication among teachers, students, and parents. Teachers will also have much more data to reflect on when designing instructional activities and specific learning interventions.
Overall, I have found the distinction of assessments as being formative or summative to be confusing among all educational stakeholders, and misleading when discussing students’ performance feedback. I think a better description would be “practice” or “process” assessments and grades, and “product” assessments and grades.
A note about progress grades, only because I mentioned it at the beginning of my rant, would consist of feedback describing the growth between unit pre- and post-assessments. Whether that feedback is better reported qualitatively or quantitatively, I cannot say, for our committee is still reviewing the literature on that one.
grantwiggins said:
Paul: I am confused. All assessments may be graded but that doesn’t mean there is no distinction between formative and summative. A grade is not feedback; feedback is information I can use about what did and didn’t work.
The issue is clear to me, and not just semantic: if you get multiple attempts at a summative task then the initial trials are formative, whether or not they are graded. If you get one shot at the task, it is summative and not formative. Why does the language matter? Because students need feedback and opportunities to use it on recurring tasks, processes and ideas. (None of these 3 can be mastered and tested in one go by almost all learners.) So, unless we deliberately design ‘formative’ assessments, all assessments typically end up as one-shot attempts that highlight content only – neither fair nor a good way to develop mastery.
And why grade everything? Does the coach give grades after every drill and scrimmage? Does the conductor grade the orchestra in rehearsal? ‘Data’ does not equal grades and scores if those grades and scores hide the feedback rather than reveal it. Grades are NOT feedback, strictly speaking; they are value judgments and hence useless as descriptive feedback.