The question of whether we should use marks, ranging from 0 – 100, or grades, usually A+ to F is a topic that often vexes teachers and can lead to heated discussions. There is scientific evidence that speaks to the topic.
The question is complicated by marking rubrics and isn’t a stand-alone topic. The reason being is that when grading assessments, best practice requires that a teacher produces a detailed rubric to guide the students in their work. It makes sense for a teacher to tell the students what they will be looking for when they are grading their work.
Common aspects of a rubric depend on the type of assessment. The example I will use is an essay or report of some kind. A rubric for an essay or report will have sections that include some of the following: ideas, clarity, development, originality, grammar, organization, poise, voice, coverage, and on and on. Often, a teacher will have (at least) seven or eight different aspects of a rubric to cover while they read the work.
An expectation of the student is that the teacher will mark every aspect separately, providing the student with guidance as to what the teacher thought of each aspect and feedback for what they need to work on in the future. In fairness, if you as a teacher say in your rubric that these are the aspects that will influence a student’s grade, these are the aspects that a teacher needs to provide some form of feedback on.
The rubric that is given to the students when the assessment is set is usually closely related to the rubric that is returned to the student after marking. The primary difference will be the judgment that the teacher makes on each of the various aspects of the work. In returning the work, it is expected that there will be an indication of the quality of each aspect. At the end will be an overall grade or mark for the piece of work will be given that is derived from the quality of the various aspects.
The question of grades or marks for the assessment lies in the detail of the rubric. Often, the aspects of the rubric are divided into four or five categories – excellent, good, adequate, poor, and fail. I have seen some rubrics that have thirteen categories (A+ through to F) and some rubrics that give a percentage mark for each aspect. All of these rubric types are used for assessments of all types.
The Science of Learning speaks to this topic through basic research looking at the number of items a person can hold in their short-term memory. The number is about 5±2 items – between three and seven, depending on the individual.
If we look at marking rubrics, even in their simplest form, there are too many items for a human being to keep in their short-term memory. If we consider a rubric with five aspects having five categories for each aspect, that gives us 20 items that must be kept in short-term memory while a teacher is reading a piece of work. A rubric that uses 13 categories means a teacher must keep 65 items in short-term memory, while a rubric that gives a percentage mark for each aspect requires that a teacher holds 500 items in their heads while marking a paper. Even if a teacher is only keeping seven items in their memory while grading a piece of work, the requirement to grade any paper using the simplest rubric is extremely complex – a feat called a simultaneous multidimensional judgment. This kind of skill is found in very few situations. It is something that takes years of practice to master, even when working at it for many hours a day.
What this speaks to is the need to simplify marking assessments. Rubrics should have a minimal number of aspects and a minimal number of categories. From this perspective, a rubric using few aspects and broad grading categories (A, B, C, D, F) is more accurate than marking.
Teachers will argue that using marks that range from 0 – 100 provides reassurance to students and parents that there is some real precision and fairness to their marking. That is true but brings to question the integrity of those making the argument. We really can’t do it but we can lie and pretend that we can do it in order to provide reassurance to the various stakeholders.
Why would we risk our integrity over something as inconsequential as whether to use grades or marks? Even using a marking scheme that has a minimal number of categories is an extremely difficult task. To pretend that we can do otherwise is both foolish and dishonest.
I doubt any of you want to know what the research really says about how teachers mark work. The biggest single influence when it comes to grading a paper is the previous knowledge of the student (not a problem when you have 350 papers to mark). The second biggest influence on grading a paper is the paper that immediately preceded the one being marked. The third biggest influence is the opening paragraph of the work. We simply don’t have internal standards of quality that are not influenced by the correlations of these three forces. What we can do, with a great deal of accuracy is to order papers in terms of the quality of the work as a result of a general impression gained from reading them, but that takes at least three readings of each paper – something that very few of us are willing to do.
In the meantime, the science tells us that we should use a simple a rubric as possible and use broad categories (grades) and even then, know that we are only making a best guess.