Archive for July, 2017

The factors influencing students choosing to go into STEM: Economics and gender matter

I saw this in a College Board report, which summarized the paper cited below with these bullets:

  • For both genders, academics played a large part in major choice—passing grades in calculus, quantitative test scores, and years of mathematics in high school were notable.
  • Also important to both young men and women was a student’s own view of his or her quantitative/mathematical abilities.
  • Key drivers in decision making differed between genders. First-generation status correlated with young men being more likely to major in engineering, while a low-income background was associated with young women majoring in scientific fields.

Based on the findings presented here, first generation status leads to a greater likelihood of choosing engineering careers for males but not for females. Financial difficulties have a greater effect on selecting scientific fields than engineering fields by females. The opposite is true for males. Passing grades in calculus, quantitative test scores, and years of mathematics in high school as well as self-ratings of abilities to analyze quantitative problems and to use computing are positively associated with choice of engineering fields.

Source: Choice of Academic Major at a Public Research University: The Role of Gender and Self-Efficacy | SpringerLink

July 31, 2017 at 7:00 am 2 comments

A Weak Argument that Silicon Valley is Pushing Coding Into American Classrooms through Code.org

When the New York Times does an article on Code.org, it’s worth noting.  I had my class on Computing and Society read the essay and critique it, and they were dubious.  They have a bias — they’re all Georgia Tech students in STEM, and almost all majoring in Computer Science.  They tend to think learning to code is a good thing.  Still, they were concerned about the article, with good reason.  They wondered, “Where exactly is Code.org doing something wrong?”

I had similar concerns. I read the quote from Jane Margolis (“It gets very problematic when industry is deciding the content and direction of public education”) and thought, “Jane didn’t just say that.  She would have explained what she meant by ‘problematic.'”  It felt to me like the quote was taken out of context.

Is Code.org really “deciding” what goes into public education?  Or are they simply influencing those who do decide?  Maybe Silicon Valley is having undue influence. This article didn’t really make the case.

Code.org’s multilevel influence machine also raises the question of whether Silicon Valley is swaying public schools to serve its own interests — in this case, its need for software engineers — with little scrutiny. “If I were a state legislator, I would certainly be wondering about motives,” said Sarah Reckhow, an assistant professor of political science at Michigan State University. “You want to see public investment in a skill set that is the skill set you need for your business?”Mr. Partovi, 44, said he simply wanted to give students the opportunity to develop the same skills that helped him and his backers succeed. He immigrated as a child to the United States from Iran with his family, went on to study computer science at Harvard, and later sold a voice-recognition start-up he had co-founded to Microsoft for a reported $800 million.

“That dream is much less accessible if you are in one of America’s schools where they don’t even tell you you could go into that field,” Mr. Partovi said.

Even so, he acknowledged some industry self-interest. “If you are running a tech company,” he said, “it’s extremely hard to hire and retain engineers.”

July 28, 2017 at 7:00 am 5 comments

Helping students succeed in AP CS: GT Computing Undergraduate Female Rising Up to Challenge in CS

There’s a common refrain heard at “CS for All” and BPC events in the US these days. “AP CS A is just terrible. AP CS Principles will fix everything.” The reality is that there are bad AP CS A classes, and there are good ones. There is evidence that just having good curricula doesn’t get you more and more diverse students. The more important reality is that AP CS A accurately matches most introductory computer science classes in the United States. If you want students to succeed at the CS classes that are in our Universities today, AP CS A is the game to play at high school.

That’s why Barbara’s Rise Up programs are so important. She’s helping female and African-American students succeed in the CS that’s in their schools and on University campuses today. And she’s having tremendous success, as seen in the story below about a female high school football player who is now a CS undergraduate.

Barbara’s work is smart, because she’s working with the existing CS infrastructure and curricula. She’s helping students to succeed at this game, through a process of tutoring and near-peer mentoring. This is a strategy to get more female CS undergraduates.

That’s when she discovered Sisters Rise Up 4 CS, a relatively new program developed in Fall 2014 at Georgia Tech by Barbara Ericson. The program was based on Project Rise Up 4 CS, which aims to help African-American students pass the AP Computer Science A exam. Sisters Rise Up does the same for females.The program offers extra help sessions in the form of webinars and in-person help sessions, near-peer role models, exposure to a college campus, and a community of learners.“The program helped me get hooked on computer science,” Seibel said. “I started to actually learn. Seeing that some of the girls in the program had interned at Google or other places like that, and that they really loved CS, it gets you excited about it. They were only a few years older than me, and I was like, ‘Oh. That could be me.’”

Source: GT Computing Undergraduate Sabrina Seibel Rising Up to Challenge in CS | College of Computing

July 26, 2017 at 9:00 am 1 comment

CS Curricula, Standards, and Frameworks will Need to Change: Larry Cuban and Coding as Vocationalism

I just wrote a blog@CACM post (see link) below on a series of essays that Stanford educational historian Larry Cuban has written on “Coding as the New Vocationalism.”  His points are well-taken.  Schools have often been swayed by the needs of industry, and he sees the current “CS for All” effort as mostly being industry-driven.  The questions that he keeps returning to in his posts are, “What are schools for? How does real reform happen?”

It’s the latter set of insights that I think are missing from our current “CS for All” efforts.  I quote Cuban at the bottom of this post with his summary for how reforms succeed. Top-down edicts on what ought to be taught rarely work.  Remember the U. Chicago’s Outlier group research on the landscape of CS education from 2014?  Most professional development is requested by the school or district,  but in CS Ed, professional development mostly sent in by NSF, Google, and Universities (and today, likely, Code.org).  CS education will have to change to achieve the goal of being driven by district and teacher needs.

The most successful reform efforts are those that achieve the top-down goals in a process of mutual adaptation with teachers, an idea developed at Northwestern by a team of learning scientists led by Brian Reiser.

Whatever our curriculum, frameworks, and standards are today, they will change before we achieve CS for All.

Standards change in response to what teachers know, what we can actually teach them (at scale), and what they will actually teach (a process that has already happened in Georgia). We certainly can’t get the curriculum right yet — we’re decades away from reaching 100% of schools in any US state, with many, many teachers to prepare and to work with in a process of mutual adaptation.  I’m not opposed to defining curriculum, frameworks, and standards.  I’m opposed to thinking that we’re going to get it right — not today, when we have such a long road ahead of us.

The lessons that have to be learned time and again from earlier generations of school reformers are straightforward.

  • Build teacher capabilities in content and skills since both determine to what degree, if any, a policy gets past the classroom door.

  • With or without enhanced capabilities and expertise, teachers will adapt policies aimed at altering how and what they teach to the contours of the classrooms in which they teach. If policymakers hate teacher fingerprints over innovations, if they seek fidelity in putting desired reforms into practice, they wish for the impossible.

  • Ignoring both of the above lessons ends up with incomplete implementation of desired policies and sorely disappointed school reformers.

Source: Coding in Schools as New Vocationalism: Larry Cuban on What Schools are For | blog@CACM | Communications of the ACM

July 24, 2017 at 7:00 am 2 comments

Why are underrepresented minorities and poor over-represented in Code.org courses?

Code.org has a blog post describing their latest demographics results showing that they have remarkably high percentages of women (45%) and under-represented minorities (48%). In fact, their students are 49% on free and reduced meals.

Only 38% of students in the US are on free and reduced lunch.  44% of students in the US are Black or Hispanic (using US Department of Education data).

What does it mean that Code.org classes are over-sampling under-represented groups and poorer students?

I don’t know. Certainly, it’s because Code.org targeted large, urban school districts.  That’s who’s there.  But it’s not like the classes are unavailable to anyone else.  If the perception was these are valuable, shouldn’t more suburban schools be wanting them, too?

One explanation I can imagine is that schools that are majority poor and/or minority might be under-funded, so Code.org classes with their well-defined curriculum and clear teacher preparation models are very attractive. Those schools may not have the option of hiring (say) an AP CS teacher who might pick from one of the non-Code.org curriculum options, or even develop his or her own.

The key question for me is: Why aren’t the more majority and wealthier schools using Code.org classes?  CS is a new-to-schools, mostly-elective subject.  Usually those new opportunities get to the wealthy kids first.  Unless they don’t want it. Maybe the wealthy schools are dismissing these opportunities?

It’s possible that Code.org classes (and maybe CS in high school more generally) might get end up stigmatized as being for the poor and minority kids?  Perhaps the majority kids or the middle/upper-class kids and schools avoid those classes? We have had computing classes in Georgia that were considered “so easy” that administrators would fill the classes with problem students — college-bound students would avoid those classes.  We want CS for all.

Code.org has achieved something wonderful in getting so many diverse students into computing classes. The questions I’m raising are not meant as any criticism of Code.org.  Rather, I’m asking how the public at large is thinking about CS, and I’m using Code.org classes as an exemplar since we have data on them.  Perceptions matter, and I’m raising questions about the perceptions of CS classes in K-12.

I do have a complaint with the claim in the post quoted below.  The citation is to the College Board’s 2007 study which found that AP CS students are more likely to major in CS than most other AP’s, with a differentially strong impact for female and under-represented minority students.  “Taking AP CS” is not the same as “learn computer science in K-12 classrooms.”  That’s too broad a claim — not all K-12 CS is likely to have the same result.

Today, we’re happy to announce that our annual survey results are in. And, for the second year in a row, underrepresented minorities make up 48% of students in our courses and females once again make up 45% of our students…When females learn computer science in K-12 classrooms, they’re ten times more likely to major in it in college. Underrepresented minorities are seven to eight times more likely.

Source: Girls and underrepresented minorities are represented in Code.org courses

July 21, 2017 at 8:00 am 11 comments

The General Purpose Blocks Programming Language, GP, is now in beta

GP, the powerful new blocks-based programming language (that I wrote about here, helped show at SIGCSE 2017, and used for MediaComp in a new kind of ebook here), is available for beta-testing as the Scratch 2017 conference starts in Bordeaux, France.  You can access GP at http://www.gpblocks.org.  You can run projects in your browser on the website, or download the application.

GP is a free, general-purpose blocks language that is powerful yet easy to learn.

GP can:

  • generate high-quality graphics computationally

  • manipulate images and sounds

  • analyze text files or CSV data sets

  • simulate physical, biological, or economic systems

  • access the web and use cloud data

  • connect to hardware via the serial port

  • deploy projects on the web or as stand-alone apps

Source: About · GP Blocks

July 19, 2017 at 8:00 am 13 comments

“Algorithms aren’t racist. Your skin is just too dark.”: Teaching Ethics to future Software Developers

In my Ethics class this summer, I had my students watch Joy Buolamwini’s TED talk when we talked about professional ethics and responsibility.  My students had not before considered the possibility that bias is being built into software, but they recognized the importance of her message. Our students who will be software engineers have to be thinking about her message, about the racism that we build into our machines.

She’s been getting a lot of press since her TED talk, including this recent piece in The Guardian.  In her blog post quoted below, she responds to her critics in a careful and respectful tone, which took an enormous amount of maturity and patience.  “Suggesting people with dark skin keep extra lights around to better illuminate themselves misses the point.”  She is more patient and well-spoken than me. I think my response to the critics would have included the phrase, “Are you kidding me?!?” (with perhaps a couple more words in there).

One of the goals of the Algorithmic Justice League is to highlight problems with artificial intelligence so we can start working on solutions. We provide actionable critique while working on research to make more inclusive artificial intelligence. In speaking up about my experiences, others have been encouraged to share their stories. The silence is broken. More people are aware that we can embed bias in machines. This is only the beginning as we start to collect more reports.

Source: Algorithms aren’t racist. Your skin is just too dark.

July 17, 2017 at 7:00 am Leave a comment

Older Posts


Recent Posts

July 2017
M T W T F S S
« Jun   Aug »
 12
3456789
10111213141516
17181920212223
24252627282930
31  

Feeds

Blog Stats

  • 1,460,720 hits

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 5,193 other followers

CS Teaching Tips