Mark Guzdial

Meet Mark Guzdial

Back to Perspectives

CIRCL perspectives offer a window into the different worlds of various stakeholders in the cyberlearning community — what drives their work, what they need to be successful, and what they think the community should be doing. Share your perspective.

Mark Guzdial

Mark Guzdial is a professor in the Electrical Engineering and Computer Science department at the University of Michigan (starting fall 2018). Before that, he was a professor in the School of Interactive Computing at the Georgia Institute of Technology for 25 years.

Tell us about your work in computer science education, particularly on the K-12 levels.

I’ve worked in six areas of computer science education:

Curriculum: I invented Media Computation, an approach to introducing computing by having students manipulate low-level data to create high-level media effects. Students manipulate pixels to create Photoshop-like effects, samples to generate sound effects, and frames to implement digital video special effects. Media computation curricula are used at many schools, and appear in AP CSA through Barbara Ericson’s Picture Lab. I was on the original CS Principles definition committee.

Technology and Media: I led the effort to implement JES, a Python IDE used for Media Computation, which has been downloaded tens of thousands of times. Barbara Ericson and I have been developing new kinds of ebooks to teach computing (AP CSP and AP CSA).

Assessment: My student, Allison Elliott Tew, developed FCS1, the first validated test of introductory CS knowledge designed to be multi-lingual. Her work was replicated by my student, Miranda Parker, in SCS1, which is a freely available research instrument.

System Analysis: I’m interested in using data to get a better understanding of the influences on CS education. A few years ago, Barbara and I wrote a paper analyzing what influences the number of AP CSA exam-takers in a state. We have a new paper that helps us to understand how socioeconomic status influences success in CS. Currently, Miranda is developing a quantitative model that predicts why schools in Georgia decide to teach CS.

Learning & Teaching Theory: We’re trying to get a better understanding of how individuals learn computer science and the role of teachers. My student Amber Solomon is publishing a paper on the role of gesture in computer science teaching. My student Katie Cunningham and I are interested in how students develop their understanding of how programs work. Last year she published a paper on how students use pen-and-paper when tracing programs, and how their notes and sketches give us insights into how they think programs are executed.

Public Policy: Barbara and I, with Rick Adrion and Renee Fall at U. Massachusetts Amherst, have led an NSF Broadening Participation in Computing (BPC) Alliance to help states improve and broaden their computing education. The Expanding Computing Education Pathways (ECEP) Alliance works with state leaders and policy makers in 16 states and Puerto Rico on a range of issues, from defining standards, to developing teacher certification, and to setting policy for making CS education valuable and funded in their states.

What has surprised you most about your research and outreach in CS education? What has not surprised you?

I’ve been most surprised at how little Schools of Education have been involved in the national growth of CS Education. There are few education faculty studying CS Ed. There are few classes or programs for pre-service teachers to learn computer science. In hindsight, I get why — Schools of Education are underfunded in the US, and they already have a lot to do. But we do have to figure out how to involve Schools of Education (e.g., to create pre-service pathways into CS education) to have sustainable, equitable CS education in the United States.

I’m not surprised that there’s a gap in the middle in CS Education. Kids are excited, parents and teachers want more CS education, and technology companies are funding CS education. But principals and departments of education are less excited. The middle has to prioritize. Is funding more CS education more important than funding more reading or science education? Is computational thinking more important (or even different) than engineering thinking or science literacy? There is excitement for more computing education, and there’s funding available. When administrators have to choose, we researchers discover that we don’t have enough knowledge to tell them how to make choices. We haven’t done the research to be able to make the case for trade-offs between CS education and other educational needs..

What is the biggest challenge for K-12 CS education going forward?

The biggest goals right now should be sustainability and systematicity. We have successful classes like CS Principles and Exploring Computer Science, but they’re in a small percentage of schools. How do we get everywhere? How do we get to every student? Teachers don’t last long in the classroom. We will likely lose half of our current CS teachers in the US in the next five years. How do we prepare more CS teachers so that we both sustain and increase access?

The biggest challenge, then, is research. We don’t have any large scale data on CS teacher retention. We know little about less-privileged students learning CS, e.g., students with special needs and low-SES students.

What should U.S. education leaders and policy makers know about your research & outreach in K-12 CS education?

The first thing to know is that the inequities in computer science are far greater than leaders and policy makers realize. Computer science is even more male-dominant than physics. Most primary and secondary schools do not offer any computer science to their students.

The second thing to know is more positive: Computers are an amazingly flexible medium. If they are hard to learn, we can change them to make them easier. Computer science is the newest STEM discipline, and we’re still figuring out what to teach and how.

The most important thing to know is that we can make things better. We can design curriculum, teaching practices, learning tools, programming tools, languages, and degree programs to improve computing education and make it more equitable.

  • I developed Media Computation because liberal arts, business, and design students were failing their required CS class at a rate of about 50% a semester. Since 2003, we have maintained an 85% success rate in that course.
  • Female participation in computing can be equitable. At Georgia Tech, our BS in Computational Media degree is over 40% female. Our PhD in Human-Centered Computing is over 50% female.
  • Our research on Parsons problems and subgoal labeling show that we can change how we teach to dramatically improve learning, retention, and transfer.

We need to focus on research and design to making a CS education than can and does reach all students.