The school year is ending in the United States and I hear a lot of chatter about grades. The most common complaint I hear is that one kid or another scored high on tests but ended the year with a disappointing grade for not turning in all of the homework. Does that grading system make sense when the point of the homework is to prepare kids for tests?

In the real world, results are what matter the most, as long as you arrive at the results legally and ethically. If an adult makes a sales presentation and nails it, no one cares how many hours she practiced before the meeting.

My suggestion is that schools issue homework grades that are separate from test grades. That way you can get a better sense of what is going on with each kid. Blowing off homework is a valid strategy if you're confident you will ace the test. It's especially valid if skipping homework creates time for a kid to participate in additional extracurricular activities.

Would you rather hire someone who cared little about homework but aced all tests, or someone who was dependable and hard-working but underperformed at test time? The right answer is that it probably depends on the job description. If you're hiring a security guard, you might want the reliable candidate. If you're hiring a research scientist, go for the test scores. If you're hiring a lawyer, you probably want both qualities.

Not all homework is created equal. If an assignment involves writing a paper, for example, obviously that grade needs to be included with test scores. I would only strip out the memorization and practice types of homework assignments and grade them separately.

At one point in our history it might have made sense to blend the scores for homework performance and tests. A combined score probably did a good job of predicting how well a kid might someday run the family farm. But in the information economy, brilliance and reliability go separate directions. We want brilliant people designing microchips and reliable people manufacturing and selling them.

Rank Up Rank Down Votes:  +98
  • Print
  • Share


Sort By:
Jun 28, 2012
Late to the party, so most likely no one will read this. I got roped into being a college instructor late in life and have now taught 4 classes in IS, database,spreadsheet, language, introduction. It's been a fairly high correlation between those who turn in homework and test scores they receive. This is an evening school for returning students at one of those non-profit former business/trade schools. At the college level, for most students (those not continuing on to advanced degrees or to very specialized degrees) what college is a lot about is being able to give your employee an assignment and have him/her complete it within a specified time. Just like most jobs they will get. A generic college degree should tell a prospective employer that the applicant was given assignments and turned in decent work over a period of time. Just like a workplace environment. That's all. I assign everyone homework, but I don't count it against them if they don't turn it in, and if they do, it replaces poor quiz scores (which feed into the unit tests which feed into the final). It's strictly to their advantage to turn it in, strictly (they get higher test scores, and they replace poor quiz scores), and I still see about the same ratio of students turning/not turning in as one would expect. It's NOT the homework, it's the student.
Jun 21, 2012
I like your thinking Scott. However, for me personally, I needed to have homework count for a grade in order to motivate myself to do it, even if it was only a small percent of the final grade. I also seemed to do better in classes that graded homework, because if I didn't do the homework, I wouldn't know the material well enough to get a good grade on the test. Personally, I think homework should never be worth more than 10% of the overall grade because a student shouldn't lose more than a letter grade for not turning in their homework.

Now for applying this scenario in the real world, I agree that whether you want the dedicated student or the one that can ace the tests highly depends on the job; however, I would split it up a little more differently than you. For any project-based job where you have deliverables and deadlines (engineering, jounalism, etc.) you would probably want someone that was more dedicated because for these types of jobs there's alot more homework and not too many tests. Not doing your homework means when a deadline comes up, you're going to have to stay up late and do it anyways.
Whereas with the people who do well on tests would be better for jobs that have more 'tests' than 'homework' (such as sales, marketing, lawers, etc) because if you know your client, you can whip something together last minute and still knock the presentation out of the park.

But all in all, I think your idea is good, but just needs a few tweaks.
+2 Rank Up Rank Down
Jun 14, 2012
I was a student with little interest in doing homework in high school, but who aced almost every test with ease (I would easily have had a 4.0 if not for my seething hatred for tedious assignments). This was due to a combination of factors: I was raised with enough classical rote learning for certain things like arithmetic and spelling to be 'automatic;' I went to private school for grades K-8, so I was ahead of the curve for the first 2 years of public high school; and I had some engaging teachers who made grasping the material fairly easy. So for me personally, doing homework was just a boring chore with little value most of the time, at least during high school.

During college, there was less work and less explanation, and doing the problems became invaluable -- it might have been the 1990's, but you were still on your own and nobody was going to hold your hand for you, at least where I went to school (and the senior, tenured professors were almost all AWFUL lecturers who contributed very little). But in that case, the work was actually teaching something and serving a purpose, not reinforcing something I'd already known for 2 years, or making me regurgitate something that was 90% similar to other things I already knew cold.

My favorite grading structure was a college class that was set up something like this: you could score between 0 and 100 points. There were four "quarters" worth 25 points each. Each quarter had a short weekly quiz and a homework assignment. The midterm was worth 50 points minus whatever you already got for quizzes and assignments. And the final was worth 100 points minus whatever you'd earned in the class to date. The assignments and quizzes were hard as hell and REALLY made you think about the material, but if you did the work and grasped the concepts by the end of it, you could earn an 'A' -- but only if you did the work. Sure, you could probably get a B with minimal effort, but the system was set up to both reward people who learn slowly over time AND those who could nail the material quickly. If you didn't want to do homework, you could still get an 'A,' but only if you nailed the quizzes and exams. I think there was an optional paper for people to make up missing points, too. So, bottom line: you could be a "test" person or a "paper" person and do equally well in that system, yet it was still measuring your knowledge.

My least favorite grading method? "Class participation," especially when it went up to 20%. That's just a load of bull manure - completely arbitrary and subjective, and it proves even less than nailing a test.
Jun 13, 2012
My experience with mathematics (especially at a higher level) is that the students who do the homewqork pass the tests with little diffculty. Those who try to cram before tests do not.
Jun 13, 2012
Here in India, the parents do the kids' homework.
+1 Rank Up Rank Down
Jun 13, 2012
[Your argument supports my idea of separate grades for homework and tests. But I'm sure you memory is excellent. -- Scott]

Well, the separate grade idea's okay, I just took issue with some of the ideas you mentioned en route to that conclusion.

Particularly the idea that tests are good at detecting "brilliance" - my memories of high school tests included people cramming the night before and forgetting it all right after the test, with little understanding at any time. Throwing mnemonic memorisation devices at a multiple-choice wall to see what sticks.

It's possible I just went to a bad school.
+1 Rank Up Rank Down
Jun 12, 2012
No. If the kids don't have to do homework, the ones who need to won't, and will fail the tests because they haven't learned anything. That may be their own fault but teaching is about nurturing, not leaving kids (low-willpower and high-impulse) to their own devices -- which is why we have teachers in the first place and not just quarterly exam administrators. And the smart kids need homework too. Whether in university or in the real world, you need to have developed discipline and study skills regardless of how smart you are and how good your grades are.

[You make me wonder if willpower, also known as emotional intelligence, can be trained into kids via the homework process. I haven't seen any evidence of that. Consider every family with multiple kids that are raised the same and have wildly different outcomes. -- Scott]
Jun 12, 2012
Saying that the point of homework is to prepare for tests is a fallacy. Building the rest of your arguments on top of that is a joke.
Homework shows whether a student can be reliable and follow instructions. It helps the teacher judge whether the student is capable of learning and functioning in the workplace.
While it can also prepare students for tests, so can learning to cheat. Or just being an exceptional test-taker. I could have passed all my high school tests, including the SAT, on my first day of high school because I was good at that kind of thing. You too, probably.
I didn't read all the comments here, but I hope you get ripped by some teachers.

[How'd you score on reading comprehension? I suggested separate grades for homework and tests. That way a college or employer has more information. It's the same reason that individual classes are reported in addition to the overall grade. The more you know about the student, the better. -- Scott]
+3 Rank Up Rank Down
Jun 12, 2012
I'm seeing a lot of additional ideas on this, so reluctantly, I'll add more beyond my original comment.

I had a fantastic college calculus teacher (the late Dr. Harold Wexler) who commented that homework counts 98% if you don't do it, and 2% if you do. I did every problem, and I agree with him. I got a B, but I never worked harder for a grade before or since.

I have been teaching engineering design classes (part time) for about 35 years. In my opinion, design is not a 'test' kind of skill. It involves a commitment of time, energy and enthusiasm to actually 'make' something. Not to mention creativity. Having it actually work is, in itself, a reward. Being able to write it up and/or present it to others closes the loop. No homework, no tests, just 'do it'. But, that's just me.

About 35 years back, a friend of mine and I were halfway up to the top of one of the World Trade Center towers when he commented: "Do you know who C students work for? Often, the 'lowest bidder'. Have you figured out who probably got the contract for this elevator?" Gulp.

My college performance on tests was 'spotty' at best. I didn't keep any report cards where I got all A's, but the card for the quarter that I got 3 C's and a D sits on my nightstand. I show it to my kids, or anyone who wants to see it. I survived, and it's a reminder.

The biggest lesson the university can offer is to teach people to love learning.

Jun 12, 2012
It's a complicated issue. Some people are bad at tests and may do badly even if they know the material. At least if they did all their homework, they probably know at least some of it.

You also are trying to teach kids to prepare in advance instead of waiting until the last minute. Just because some kids are able to get away with cramming does not make that method the most ideal and quality way of performing. Also, a longer period of exposure means you remember the material better long after the test.

Also, even smart people would do better to be able to work with others at least decently. At the work place and even as a boss, you don't always get to do and have things exactly as you would like. Flexability is better.

ON the flip side, I do think homework has gotten to be such a huge chore and has become a bit over rated. Part of being a success is a well balanced personality. Kids should have free time beyond school. In many ways, it's getting so that kids spend almost all their time either at school or working on school projects. There needs to be room for other kinds of development as well.

Anyway, I have come to believe that success is more about attitude than anything else. It's an attitude that you want to and can do it and that you will work at it and look for solutions to problems until you find them and keep trying until you figure it out. I am not sure either homework or tests, the way the system is currently designed, really do much to teach that quality.
Jun 12, 2012
Lets say I'm the guy who programs your pacemaker.

Can I slack off and do piss poor job most of the time and then then pulls everything together at the last moment before shipping when there's no time to make sure everything actually works. Or would you prefer that I work diligently and hand in every part of my work on time to be thoroughly tested?

[I give you a grade of F for your poor analogy, but B+ for the work you put into it. -- Scott]
0 Rank Up Rank Down
Jun 12, 2012
@Ecks - I should have qualified that homework that is designed to rehash what is in a book is not what I meant. Homework should be about taking the lessons learned and applying them to show and develop understanding and expand beyond what was taught in the classroom.

It is also something that has to be age appropriate as people learn differently at different ages.

As to test anxiety, I do not believe it relates to how well someone is likely to work under a pressure environment. In fact, I would believe that you could tell quality individuals by their homework which simulates most job settings in the office better. Writing reports, meeting deadlines, reviewing documentation, following instructions, exceeding expectations, effort and curiosity being attributes I would see homework showing more so than tests do.
+9 Rank Up Rank Down
Jun 12, 2012
No disrespect Scott, but I never thought the point of homework was particularly to prepare kids for tests - and if anything, I would have thought tests were *more* biased to favour memorise-and-spit-back-out type students than brilliant ones.

Being able to store a set of facts over a short-or-medium term makes you about as "brilliant" as a sheet of rapidly biodegradable note paper, and is a completely irrelevant skill in a smartphone enabled world.

[Your argument supports my idea of separate grades for homework and tests. But I'm sure you memory is excellent. -- Scott]
Jun 11, 2012
Hi Scott, interesting comment there. I remember in secondary school receiving an "87% fail" for "not meeting mandatory requirements". My exam mark was 87% so I should have received an A grade but I failed to hand in an internal assessment that was worth nothing but was mandatory.

I also remember a class where the homework ("internal assessment") was worth 25% of the years mark, scored by if you did it or not, so even if you scored a "fail" in the exam, the total COMBINED mark meant that if you scored at least 30% in the exam and did all the homework you would pass the class! Crazy? Yes! Because if you didn't hand in any of the homework it was an instant -25% to your years mark regardless of if you aced the exam, which made it nearly impossible to pass!
Jun 11, 2012
I think the more important factor is the subject matter that one is trying to evaluate in the student.

For example, if you are trying to evaluate a students writing ability, then it is much better to grade the student on papers that they turn in. If on the other hand you are trying to grade a students reading comprehension skills, then assignments are a poor metric, and a formal test is much better.

The purpose of formal schooling is not to endow students with all of the skills and knowledge they will need for the future. Rather it is to evaluate a students ability to learn and adapt. Post-secondary schooling then follows by training students on how to learn those skills. Most of the skills a person will use will be learned on the job.

Heres some more examples of whether a student is better evaluated on a test or on homework:

Math: Test
Biology: Test
Physics: Homework
Chemistry: Test
Writing: Homework
Reading: Test
History: Homework
Sociology: Homework
Shop/HomeEc: Homework
Geography: Homework

+6 Rank Up Rank Down
Jun 11, 2012
Two kinds of Students? Hmmm . . . there are two kinds of people. Those who classify students in one of two buckets and those who don't.
Jun 11, 2012
> In the real world, results are what matter the most, as long as you arrive at the results legally and ethically.

Disagree with your premise - the purpose of homework is not always to prepare for a test. I can remember in English and History class being assigned homework essays. The purpose was to prepare students to do work, in this case writing on an assigned topic, which in and of itself was the learning.

I'd also submit that teaching students a good work ethic, even if the work doesn't directly affect their test scores, is valid - often one must "show their work" to receive credit. This is true as well in the real world, where one's pronouncements are often challenged.
+4 Rank Up Rank Down
Jun 11, 2012
I am a high school math teacher, and I have few students who can not do homework and still make an A on every test. Of those who COULD do it, almost all care too much about their grades to blow off their homework. However, for the occasional student who can make an A on every test, I don't penalize him/her in the end for not doing the homework, because, in my opinion, the grade should primarily reflect what the student knows.

The next big question though, is what do grades mean in high school? I would say almost nothing. Sure, you need decent grades to get into a decent college, but no college can vet every single high school. A 3.7 at one school is completely different than a 3.7 at another school. A 3.7 from a student who took no advanced classes is completely different than a 3.7 from a student who took all advanced classes, etc. And this is why colleges rely so heavily on standardized tests like the SAT and ACT--even though SAT/ACT scores are far from perfect predictors of college success.

I predict (and hope) that we will see nation-wide reform of grading practices in high schools. With most states in the nation switching to using a common set of standards (Common Core State Standards), there is a much better basis for comparing state to state, if, for example, students received grades for specific standards they should have mastered vs. a single grade to summarize an entire course.
Jun 11, 2012
I suggest a simpler compromise that was used in one of my upper-division math courses in college:

- Homework is nominally part of the grade BUT
- If you get an A on all tests you get at least an A- for the course, regardless of your homework grade.

This lets the brilliant students skip unnecessary practice and also lets the plonkers pad their grade with hard work. Incidentally the professor warned us that in 20 years only one student had ever overcome a bad homework grade this way.
0 Rank Up Rank Down
Jun 11, 2012
Around the globe lots of different systems are tried.
One, at many universities:
All assignments are combined into one mark, the end-of-term test into another and the average is it. So the test mark counts as much as all homeworks together.
Get the new Dilbert app!
Old Dilbert Blog