Exit Slip: Are Marks Accurate?
In Leonard Mlodinow's book The Drunkard's Walk Mlodinow explores the way randomness is present in and affects our everyday life. This is especially true for people who struggle with math and numeracy, but even for those of us who have studied math, randomness can be an extremely difficult notion, especially as a species built to recognize and work with patterns.
I'm sharing this book not only because it is a book that I enjoyed reading, but also because in it Mlodinow dedicates a section to discussing marking and assessment with percentage grades. My personal experience, and having read this book, lead me to say no, percentage grades are not accurate. Useful? Sure, but not accurate. Or at least not to the point of individual percentages.
I am of the opinion that it is impossible to be consistent in assigning number grades to students' work. I might be biased in this because I spend a lot of time thinking about how we assess the arts, because drama is my other teachable subject, and because music and theatre are things that I'm passionate about.
In my drama cohort language around assessment is often on a scale of "Does not meet expectations" "Approaching Expectations" "Meeting Expectations" and "Exceeding Expectations." Because it is hard to put a number value on performance and drama. However, it is often required that teachers translate this assessment into number values. It feels somewhat arbitrary to choose a number for each of those levels of assessment, so how could they possibly be "accurate"?
In pop culture, reviewers often use numbers to assess how good music, films, books, etc. are. So it exists in the "real-world" outside of the classroom, and once again, it seems somewhat arbitrary. On one hand, these reviews aren't assessing how well something sells tickets etc. but on the other hand, if it is a respected publication it can have an effect on how well the art sells. One example that is relevant to me is the website Pitchfork as they review new music releases and it's fun to see if they've decided that something new is better or worse than something else that I enjoy. But really, it's not the specific number that I care about, it's whether it's higher or lower than the number assigned to something else. It has no intrinsic value to me, only comparative value.
So what does all of this say about the accuracy of math or science assessment? One might argue that you can give a number of questions to a student and their grade would be reflected in the number they answered correctly. The percentage is a direct expression of their performance. And on the surface level this seems to be okay, and indeed this is what I have seen a lot in education. But there are a lot of questions below the surface. Who decides what questions are tested? Who decides how long the test is? Are there partial marks, or questions worth multiple points?
All of these factors contribute to the subjectivity of percentage grades. It's obviously a very complex issue. There are a lot of things that can be changed in assessment that will influence student success, but what determines what changes are better or worse? Is it good enough to say "at least everyone is assessed according to the same rules? Do the rules have biases?
I'm sharing this book not only because it is a book that I enjoyed reading, but also because in it Mlodinow dedicates a section to discussing marking and assessment with percentage grades. My personal experience, and having read this book, lead me to say no, percentage grades are not accurate. Useful? Sure, but not accurate. Or at least not to the point of individual percentages.
I am of the opinion that it is impossible to be consistent in assigning number grades to students' work. I might be biased in this because I spend a lot of time thinking about how we assess the arts, because drama is my other teachable subject, and because music and theatre are things that I'm passionate about.
In my drama cohort language around assessment is often on a scale of "Does not meet expectations" "Approaching Expectations" "Meeting Expectations" and "Exceeding Expectations." Because it is hard to put a number value on performance and drama. However, it is often required that teachers translate this assessment into number values. It feels somewhat arbitrary to choose a number for each of those levels of assessment, so how could they possibly be "accurate"?
In pop culture, reviewers often use numbers to assess how good music, films, books, etc. are. So it exists in the "real-world" outside of the classroom, and once again, it seems somewhat arbitrary. On one hand, these reviews aren't assessing how well something sells tickets etc. but on the other hand, if it is a respected publication it can have an effect on how well the art sells. One example that is relevant to me is the website Pitchfork as they review new music releases and it's fun to see if they've decided that something new is better or worse than something else that I enjoy. But really, it's not the specific number that I care about, it's whether it's higher or lower than the number assigned to something else. It has no intrinsic value to me, only comparative value.
So what does all of this say about the accuracy of math or science assessment? One might argue that you can give a number of questions to a student and their grade would be reflected in the number they answered correctly. The percentage is a direct expression of their performance. And on the surface level this seems to be okay, and indeed this is what I have seen a lot in education. But there are a lot of questions below the surface. Who decides what questions are tested? Who decides how long the test is? Are there partial marks, or questions worth multiple points?
All of these factors contribute to the subjectivity of percentage grades. It's obviously a very complex issue. There are a lot of things that can be changed in assessment that will influence student success, but what determines what changes are better or worse? Is it good enough to say "at least everyone is assessed according to the same rules? Do the rules have biases?
Comments
Post a Comment