UCLA just posted a new brief on rates of completion in STEM broken down by ethnicity (too bad not by gender as well, can't have everything I guess).
Good news first; in 1971 38.4% of Whites/Asians intended to pursue a STEM degree versus 27.9% of what the study calls URM students (underrepresented racial minority). In 2009 however the breakdown was 34.3% White/Asian and 34.1% URM. So the good news is, interest seems to have caught up to near-equal levels.
The bad news? Completion rates. UCLA used two different studies to calculate this. I'm only going to mention their 5-year completion rates here because if any of these students are engineering majors, that's going to be the normal timeframe. Of those who declared a STEM major as a freshman, the following actually went on to complete their degree in five years; White-33% Asian-42% Latino-22.1% Black-18.4% Native American-18.8%.
So I busted out some excel skills on this one(the original study has some nice graphs as well). Below you can see the average number of people indicating an interest in STEM, pretty dire (I think 50% would be a more reasonable interest rate we'd want to aim for). I'm using the UCLA sample size for my arbitrary number of people.
Here's a breakdown in declared major (or interest or whatever UCLA was using) by ethnicity, assuming all races had an equal number of people to begin with which of course they did not.
On the left is a rough population model of the US population in general applied to our sample and not including groups we don't have data for. On the right is the ethnicity breakdown of those completing STEM degrees in five years combined with overall US population ethnicty and with overall college completion rates. One can see how much smaller the URM pie pieces get.
Good news first; in 1971 38.4% of Whites/Asians intended to pursue a STEM degree versus 27.9% of what the study calls URM students (underrepresented racial minority). In 2009 however the breakdown was 34.3% White/Asian and 34.1% URM. So the good news is, interest seems to have caught up to near-equal levels.
The bad news? Completion rates. UCLA used two different studies to calculate this. I'm only going to mention their 5-year completion rates here because if any of these students are engineering majors, that's going to be the normal timeframe. Of those who declared a STEM major as a freshman, the following actually went on to complete their degree in five years; White-33% Asian-42% Latino-22.1% Black-18.4% Native American-18.8%.
So I busted out some excel skills on this one(the original study has some nice graphs as well). Below you can see the average number of people indicating an interest in STEM, pretty dire (I think 50% would be a more reasonable interest rate we'd want to aim for). I'm using the UCLA sample size for my arbitrary number of people.
Here's a breakdown in declared major (or interest or whatever UCLA was using) by ethnicity, assuming all races had an equal number of people to begin with which of course they did not.
So Wikipedia, educate me some on college completion rates by ethnicity.
Let's unscientifically combine that with my previous numbers to see a breakdown of those graduating with STEM degrees in five years by ethnicity (note: Wikipedia's data did not include Native Americans, so I had to not show them here).
On the left is a rough population model of the US population in general applied to our sample and not including groups we don't have data for. On the right is the ethnicity breakdown of those completing STEM degrees in five years combined with overall US population ethnicty and with overall college completion rates. One can see how much smaller the URM pie pieces get.
Lessons learned? I like playing with excel. Also, it's nice that race/ethnicity is no longer a bar to having an interest in STEM. But the washout rates for STEM overall are not good and are really bad if you're an underrepresented minority. This is compounded by the fact that underrepresented minorities have much lower college graduation rates overall. So in tackling this disparity you'd need to focus on not only why STEM washes out more minorities than other people, but also why fewer minorities have both access to and better chance of success at graduating with a college degree.
No comments:
Post a Comment