Tuesday, August 12, 2008

Why the focus on standardized math scores are a disservice to science

FSP's post this morning on an article last week in the Chronicle of Higher Education ("How our culture keeps students out of science" by Peter Wood) gave General Disarray and me a great deal to talk about this morning on our way to work. If you have not read either her post or the original article, I suggest you check them out. In short, there are a couple of paragraphs that re-raise the specter of gender differences in scientific abilities. One of the pieces of evidence that is generally brought up in this context (though not in that article) is this underlying assumption that standardized math scores are the hallmark of someone's ability to succeed at science. I know that not all the arguments about why the gender disparity exists in the scientific ranks are focused on the standardized test results but it does seem to keep rearing its ugly head.

Now, I'm no expert on any of this, but I have been a practicing female scientist for a number of years, so I like to think I may have something relevant to say on this topic. I will admit that my SAT and GRE math scores were not, shall we say, stellar. I can't remember what exactly they were, but I was always happy if I landed in the 80th percentile somewhere.  I struggled in math classes and got A's in college - though not in high school.  And I only did that through hard work and sheer stubborn determination. (I always felt like doing math was like writing with the wrong hand - if I focused really hard I could do it but it wasn't necessarily pretty). However, here I am: a female scientist with a good scientific reputation for a young assistant professor. And, I would like to add, with a little glee, more successful (in terms of number of pubs in high quality journals and in terms of citations of said publications) than a not small fraction of my male grad student cohort who undoubtedly scored much, much higher than I did on the math GREs. How can this be?

What I think we miss in this debate is that science is more than math and scientific ability is not quantified by how well one does regurgitating facts on some GRE subject test. Science is about synthesizing facts and ideas, formulating a concept of how the world works, and using logic to make testable predictions of what should be happening if that concept is correct. The scientific process requires creativity, knowledge, and logic.

When I was a graduate student I tried to be more like my male colleagues who had great quantitative math scores. I was bad at what they were good at and we all knew it. One day, I decided that I needed to stop being them and try being me and I knew I was really good at two things: creativity and logic. By logic, I don't mean math. Math is a codified form of logic, but it is not all logic. Much of the logic I employ on a daily basis has nothing to do with math and everything to do with a more verbal type of logic and afterall, isn't that  the core of science? If I think this is happening then I would predict that this other thing must be occurring because of X, Y, and Z. Once I embraced my strengths, my science fell into place.

I'm not saying that math doesn't have an important role in science. What I am saying is that there are many ways to be a great scientist. Insane math skills are only one route. In my short career I have met outstanding scientists who were mathematical geniuses, writing geniuses, conceptual geniuses, experimental geniuses, etc, but lacked strengths on some other axis. And, frankly, while being a genius might make you famous, you don't have to be a genius to have a great career as a scientist (and trust me, I know quite a few people who fall in that category).

The focus on math scores on some standardized test as a symbol of who can be scientists does a disservice to science, education, and our students. In my opinion, the diversity of ways we approach science and the tools that we use to address questions are essential to the scientific endeavor. This focus on mathematical scores is a selective force that focuses on only one type of "intelligence", to the detriment of all the others. It is also a huge disservice to our students who grow up thinking a) that they need to be in the 90th+ percentile on math scores to go into the sciences and b) that if they don't like math then they won't like or be able to do science.  No wonder we have a problem recruiting bright, motivated students. Once I focused on my strengths, I have never found that the fact that was not 99th percentile in math to be a hindrance to my career. Instead, as a scientist, I draw everyday on my abilities to write well, reason clearly, and look at disjointed pieces of the puzzle of life through a creative lens.  And with that, I will leave you with one last thought - we remember Einstein, not because of his sick math skills and 99th percentile score on the SAT, but because of his creativity. And isn't that great.


Anonymous said...

AMEN SISTER! I'm another of the so-so GRE scores who is far and above more successful than my male high GRE score cohorts in publications and grants.

yolio said...

One of the things that gets me about this test score line of argument is that often the proponent will say that if you are willing to be "logical and rational" you will "admit" that women are bad at math. But of course, if you actually use logical, rational thinking to analyze the test score results, you would never come to that conclusion.

I mean, people are trying to use a very small (<2%) mean difference to explain a 40%+ achievement gap. What is more, the test gap is usually measuring something like ability to do raw computations quickly in your head, or 3-D spatial rotations. Does anyone actually believe that the difference between tenured and not tenured is doing arithmetic in your head fast and accurate? These skills may help with computing the tip at lunch and maybe rubix cubes, but come on people: this is the age of computers. The argument doesn't make any sense.

BTW: My GRE was perfect in math and analytics and near perfect in verbal. This does not seem to have exempted me from any of the anti-woman B.S.

Rosie Redfield said...

Math skills help you make up solutions and reactions, but general logic skills help you recognize and do good science.

(And this isn't just sour grapes - my math GRE was in the 99th percentile.)

Gen. Disarray said...

Man, you're telling me. I've spent a good part of my day working on a project with colleagues whose advanced degrees are in math, physics, rocket science, and string theory. Basically a bunch of fields guaranteed to be populated by folks who have aced every single math test they've take since the day they were born. And I can attest to the fact that a good math nerd does not a good scientist make. I mean like wow, it's not even close.

Isis the Scientist said...

Professor Chaos, FSP's post also gave me alot to think about. I thought about posting about gender today but I usually find myself so baffled by this topic. You see, I think there are gender differences and I like that there are gender differences, but I don't know about science and math differences. I just plum don't know, PC.

River Tam said...

Considering how many dinners I've been at where 10 scientists can't seem to calculate the tip and divide the bill total by 10 I have to wonder whether a high GRE score really even helps there!

@ isis

When I was a beginning grad student I believed with every fiber of my being that there was no difference between men and women. With every passing year, I become more and more convinced that there are inherent differences but that these differences do not mean that we are less able to be scientists only that our strengths and weaknesses are different. That was what I was fumbling with trying to communicate - that using one standard of "ability" may be selecting against people with different but important abilities for the scientific process.

@ general disarray
Thanks for joining us! I was beginning to worry that people might think you were all in my head....though I guess there's still no real evidence that I haven't made you up! Maybe we can convince you to guest blog at some point.

River Tam said...

Crap, also meant to respond to Rosie and anonymous -

@ rosie -
I completely agree about relationship between logic and math. The more mathematical types I talk to often refer to "setting up a problem" and then "turning the crank". I think of logic as being able to construct a framework to explain a mechanism and predicted results. If you can then write that as a mathematical equation then it is a clearer statement of the relationships among the elements of your framework.

@ anon - thank goodness someone else here had a so-so GRE math score! However, I find the girls bad at math crap to be increasingly bizarre with all the women here with almost perfect math scores! Hmmm.

Anonymous said...

Hey Prof Chaos,

Google Reader recently directed me to your blog so I'm new around these parts. I frequently wind up ranting about the standardized test myth of academia. (And yeah, I've scored in the upper 95-99% of nearly every math standardized test I've taken so not sour grapes either.)