Do Good Grades Predict Success?

Paul Kimelman lives in Alamo, Calif., and is C.T.O. of the Texas-based microcontroller company Luminary Micro. He is the sort of blog reader we are very fortunate to have. He writes to us now and again with such interesting queries that they’re worth putting up on the blog in their entirety.

Here’s his latest:

I was speaking with a colleague the other day and he was remarking on an accomplishment I have had in my field (of microprocessor design). He assumed I had been a straight-A student all through school.

When I noted that I was far from it, he was shocked. This got me to thinking: we usually just assume that somehow grades in school (at any level) are predictors of future success, or certainly of intelligence; but I highly doubt it. I tried to find some good studies, but found five problems immediately:

1. The very definition of success is elusive.

Is a straight-A student who went all the way through Harvard Business School a success if she sells insurance? If she opens a business, what determines when it is a success? A hardware store in Iowa may not cut it, but creating Home Depot presumably does; what about all the variance in between?

It is even more complex in many other areas. In engineering, being a worker bee is success, but great advancements do not come from those people; so what are we even trying to measure? If we try to put a scale on it, what metric should we use?

Income, even attempts at “earned income,” is tricky for many reasons, but most obviously the inequity of different fields in terms of income potential. A highly successful grade-school teacher (measured by students who become motivated [by that teacher], and [were] thus successful) will always do poorly compared to even a middling professional football player. The highly successful lawyer who does a lot of pro bono work comes off worse than the ambulance chaser.

Hierarchy in the field does not really work, since few fields have clear gradations or career paths, and many such paths are not reflections of success, but only reflections of time.

2. How do you measure validity of grades?

Besides the obvious problem of the A from a poor-quality school being worth less than from a high-quality one, you also have grade inflation, subjective measurements, and, more importantly, subject difficulty.

[Subject difficulty] is more problematic. Someone who gets all A’s in “communications” at a university is probably not working as hard as someone who gets all A’s in physics. Yet this is all subjective. Why would we assume that physics is “harder” than, say, literary critique?

This is especially problematic in high school and middle school, where many “hard” subjects are about memorizing and repeating well-defined steps. Literary critique has no well-defined guide posts (unless you cheat and plagiarize), and so requires a deeper understanding of what is to be done.

3. Most middle schools and high schools put so much emphasis on homework versus actual understanding that they are measuring behavior and compliance far more than what has been learned.

So we end up with two issues: we may well predict success [only] at compliance-oriented fields, and we do not know how many have been trampled so that their possible future success has been lost. Further, this method likely pushes more people towards compliance-oriented behavior, and so reduces their potential for success outside of this narrow measure. We certainly see this in other countries (e.g. Japan).

4. Creativity and creative people tend to mess up metrics at each level.

Creative people tend to do worse on grades at each level of schooling, yet their success measures can be very high in their fields. However, creative people can also be abject failures as a result of their creative natures; so we have no good metric that predicts [how successful] these people [will be]. Even trying to separate out creative people in schools is hard, as much of their behavior is similar to those who are just lazy, have A.D.H.D., or are generally disruptive.

We often do not know the underpinnings of their behaviors until much later, and many may have been crushed under the molding systems of our schools. Further, many of the most successful [people] are specifically creative with high strengths in mathematics and its implementation: in economics, physics, chemistry, engineering (including civil; think of many of the most dramatic bridges and buildings), and so on.

5. Any research I could find was done at some university which tended to bias results using university metrics of success.

This is likely unavoidable for the above reasons, but results from different studies were so contradictory that you have to conclude filtering and selection bias had a very large role to play as well.

What interests me is whether the present system actually produces more success or heavily limits it.

Would a different system with less emphasis on conformity produce more of our best and brightest? Or does the annealing effect of being crushed by the system help to produce those best and brightest?

If you look at those who have commonly advanced our thinking, our abilities, our technologies, and our economy (through business sense), many did poorly in schools, yet they persisted. The persistence may have been the critical element, and it would have perhaps been lost had they been encouraged more.

So does this mean we need more of those mediocre middle school and high school teachers acting as the forge to both create the worker bees we need, as well as the best [and most successful] by trying to destroy them?



Einstein got bad grades


I think that one important contributing factor to this discussion is the definition of "intelligence" and the question of whether there are multiple intelligences that should be measured (along the lines of Howard Gardiner's arguments). In other words, take the kid who genuinely struggles with academic coursework but is a genius in the realm of troubleshooting and fixing mechanical systems (e.g., auto shop). Is he unintelligent? Or the brilliant English professor who is genuinely poor in math; or the brilliant musician who is athletically inept; or the brilliant scientist who is so introverted, uncharismatic, and socially awkward that he couldn't lead a group of 10 people (let alone a city, county, party, or country) if his life depended on it.

I think that the American academic system is heavily influenced by the assumption that the IQ test is the one true measure of intelligence. As a result schools tend to teach to and evaluate only one sort of intelligence (whether one of Gardiner's or not). Consequently, I think, the American academic system measures behavior and compliance over what has been learned, as Paul Kimmelman pointed out. I think that this monolithic understanding of "intelligence" also contributes to what he wrote about the impact (and plight) of "creative people."



In regards to the question on the type of teachers needed--I say both! I had good grades and was in gifted programs, and I owe an enormous debt to two particular teachers (one middle school, one high school). They signed me out of many classes when I was busy, stressed, or had a migraine, and would help me bargain with other teachers to turn work in late, do something different from the rest of the class, etc. At the time, that flexibility made school more bearable for me, and contributed to me being more successful in school.
Once I began working (at 14) and particularly as I transitioned out of the academic world, I began to appreciate the "worker bee" producing teachers more. It takes a lot of endurance to show up every day for work, and do the same thing (roughly) day in and day out. You don't have the changes, the new chapters, the spring breaks that come with academia. And the fact that I never had to show up every day, never had to do all the work, because I was so smart and had a few key teachers backing me gave me unrealistic expectations about what I could 'get away with' because of my intelligence. I'm still very successful, but it was a big adjustment--and it might not have been if 'worker bee' producing teachers had stuck to their guns when I was young! I doubt any of them would have impacted my life as much as my two outstanding teachers, but they would have been a good baseline.



((Someone who gets all A's in "communications" at a university is probably not working as hard as someone who gets all A's in physics))

Quit making fun of Sarah Plain just because her major was Communications.

Chris M.

At #52:
Yes, Einstein does come to mind, though hopefully not for the "fact" that he failed math as a child, which isn't true (See the excellent biography of Einstein by Walter Isaacson for more information on this). To quote Einstein:

"The value of a college education is not the learning of many facts, but the training of the mind to think."

Although I am certain we will never reach a final agreement on what success means, I would agree with Einstein and say it would require a person to be able to think critically and logically about all things, without being overly biased.

I suppose the relevant question is: does public education (or other forms of education) provide this?


to #44:

I think #50's got it -- success is meeting your goals

why? If we measure success by achieving any particular thing, say, money, truth, or happiness, then we are assuming that the goal is good. We have given success a value statement. If we do this, we don't mean anything special by success, but just that it is achieving that supposedly valuable thing. Which might be what many people mean by success, but it isn't very interesting. The science that you hope is close to defining success is really going to fail, because science fundamentally has nothing to say about value statements, anyway.

But if we see success as the achievment of a goal, we've got a much more interesting concept to work with.

Not necessarily useful with grades, however.


1. "A highly successful grade-school teacher (measured by students who become motivated [by that teacher], and [were] thus successful) will always do poorly compared to even a middling professional football player."

This is a horrible argument. The "middling professional football player" is in the middle of the top 0.1% of what he does. I would argue that the expected lifetime pay from an average Joe going into football is probably significantly lower than the average Joe going into education. If you want to compare the most successful teacher at a typical high school to a football player, try the most successful player on that high school's football team.

2. Standardized tests greatly combat this effect

3. "Further, this method likely pushes more people towards compliance-oriented behavior, and so reduces their potential for success outside of this narrow measure. We certainly see this in other countries (e.g. Japan)."

Right - Japanese people are narrow-minded. This is why they have the 3rd largest GDP in the world (after China, which also uses a widespread and systematic test-taking system), and come up with some of the most innovative mechanical and electronic inventions in the world.

Also, where is your justification that this is true? I happen to have learned plenty of concepts in middle school and high school, and it may be surprising to you but I learned a lot of these concepts while in the process of doing homework. I don't want your anecdotal evidence here. Give me facts to back up this point.

4. Hm, so now creative people are lazy, have A.D.H.D, and are disruptive. Interesting. Where's the study on this?

If you were trying to say that SOME creative people fall through the cracks, then you were being a bit misleading. But that's not a solid argument that the system is broken unless it happens on a widespread scale. Again, no evidence to this in your post.

5. Ah, so THAT's why you don't give us any evidence in this post - any studies you could possibly dig up are biased! How convenient.

I gotta say, this was definitely one of your the weaker posts on this blog - basically a rant. The author makes broad sweeping statements with little to no evidence. Did the author's son get a bad grade on a test today or something?

Also, I'm surprised that this post made the cut here, a blog that I read because I like to hear well-constructed, objective arguments. The guy basically shuns objective research by building a strawman case and even gets a little bit racist.



Funny -- today I just got mailed my transcript from college.

Looking back on things, my GPA would not be a very good predictor of my future GPA, let alone more difficult to quantify measures.

Goes like this -- mediocre grades (with inflation) through highschool. Generally better grades in harder classes, because I wasn't so bored, and crappy grades in easy classes. Second semester senior year, all crappy grades, because I was into college.

Freshman year, 1st semester, 1.9. Second semester, had to withdraw from all my classes and get no credit. Year off. Then 3.8 or so sophmore and junior years, and senior year lack of interest and motivation (I'd proven to myself I could do it the two years prior) led to a slip to 3.2 or so.

So it seems that within one metric, life course manages to interfere too much to make grades a decent predictor.


Do we really need to define success? I think as an individual we are most successful if we are happy with whta we are. All the problems come only when we try to be successful in the eyes of other people. They after all need some frame of reference.. be it school grades or awards in some drawing competition. Also, with a high number of applicants for one job, the employers do need a similar yard stick. So we can't really live without the present grading system. Yes, its true that sometimes brilliant people suffer a lot in this sort a compromise/mediocrity, but eventually in the long run they do succeed. So basically I think its kind of a compromise. We know all too well this isn't the best way to educate and evaluate children but this is the convenient one.


Grades may not reflect how talented someone is or how successful he/she may be, but if I had to choose to employ between someone with a great personality and low grades with someone quiet with good grades, i'd choose the one with good grades.

it's hard to tell if someone would be successful through his/her grades but it is a better measurement of how hard he/she is willing to work and the kind of attitude he/she may have built throughout school.

Paul K

@33 (c.o.d), a few comments if I may:

"has found some studies (bad ones)" - actually does not state they are bad, only that they appear to suffer from issues of measurement (quantification).

"discuss the problems with the studies" - really is discussing the problems with the premise/concept of the studies or their measurement, not the studies themselves. If the fundamental problem is that the correlation or outcome quantity/index/value is in fact not properly measurable, then the rest of the study is somewhat meaningless, so method is not relevant. The focus is on the nature of such measurements and whether one can use any specific values to get valid results.


I imagine that grades are strongly correlated with success, however you measure it. There would be some noise, of course (Jay-Z dropping out of high school before going on to be one of the most successful artists in the music industry would be `noise' in this context). But better students will, on average, do better and earn more than their peers with low grades.

Grades are probably better predictors of income for industrial cogs, as opposed to entrepreneurs or specialists, where success is driven by hard-to-measure idiosyncrasies.


I think homeschooling provides the best of both worlds. You can adapt the curriculum to the needs of the student.

I think the best way to measure success in this case is to ask a person's 20 closest friends, family, and co-workers to fill out an anonymous survey.


When I first saw the title of this article, my answer was NO, and Eistein came to my mind. No matter how intelligent one is, school only takes into account of the work done, or right/wrong answers, not what we learned from the mistakes. I understand what JoseCMS that school grades reflect only the effort we put into the work; yet that is not always the case either. Even if we put a lot of effort in studying and understanding the concept/material perfectly, everybody makes little mistakes (especially in math) that lowers the grades (when converted to %, lots of points)greatly. That does not reflect the effort tehy put in. I think the focus of grading should be done on what we learned from the misatkes and not what we got right or wrong. (ex: test corrections - why got wrong and why the new answer right)


Did you work hard on (academic) things in high school that the teachers felt were not important? It sounds more likely that you were lazy in high school but then decided to really apply yourself in college. Why shouldn't it be a surprise then that your grades improved?

I was the exact same way: In high school I only bothered doing work in classes that I enjoyed (because it didn't feel like work), and just got by, while in college I worked hard and graduated summa cum laude. Of course, it also helped that in college I could concentrate on this subjects that I liked. And then in my career I focused on things that I was good at, so I was sucessful.

In grade school, one must study a variety of subjects, some of which you are bound to either lack aptitude for or interest in, but in later life you can limit yourself to what you do best.


Some that do poorly in school need to "try harder" to find and create opportunities, thus the outcome tends to be better then those who do well and don't.

Jackson Miller

I am a 31 year old high school dropout, and I think I am pretty successful. I own a couple of retail stores that have had solid growth for 4 years and I am in the process of launching a new startup. I now work for myself and provide well for my wife and children. Most importantly, I am doing what I WANT to do.

Granted, I always scored incredibly high on tests and had a deep comprehension of what we were studying even though I had dismal school attendance. Still, my lack of documented formal education hasn't held me back and I think I have been able to get a very well rounded "education".

So, put me in the column of "grades are not a good indicator".


This is also an issue pertinent to women's rights: women tend to get much better grades (or to "perform" better than their male counterparts), especially at college. And yet in terms of "success" after college, they earn less, get fewer promotions, and run fewer businesses.

I must say, though, that I resent the idea that creative people get bad grades.

Nad Vega

"we usually just assume that somehow grades in school (at any level) are predictors of future success, or certainly of intelligence"

This is a revelation to me: does anybody actually assume that? School is by nature a conservative institution, and high grades probably mean you are playing by the rules, which is unlikely to make you a success in real life these days. As to intelligence, does it take enormous brainpower to memorize the content of your textbooks (which is essentially what you are required to do at school)? The way I see it, intelligence is the ability to effectively use whatever information you have, not the amount of data you store in your memory. The smartest and most knowledgeable kids are often simply bored in the classroom and therefore don't do that well grade-wise.

As to the definition of success, I believe it is pretty obvious. If you set yourself goals and achieve them, you are a success. If you don't achieve them, you are a failure. If you don't set yourself goals altogether (and living that lifestyle is not a goal in itself for you), the concept of success is probably not applicable to you at all.



My definition of success is utilization of all available resources wisely, efficiently and respectfully, which also requires a sound moral compass.