SAN FRANCISCO -- New research from Google shows that black students are less likely to have computer science classes in school and are less likely to use computers at home even though they are 1.5 times more interested in studying computer science than their white peers. The findings are part a report released Tuesday by Google in partnership with Gallup that puts the spotlight on the racial and gender gap in K-12 computer science education. Google says its aim with the research, which surveyed thousands of students, parents, teachers, principals and superintendents, is to increase the numbers of women, blacks and Latinos in computer science. Computer science classes are popping up in K-12 schools around the country. The growing effort is coming from many quarters -- the National Science Foundation, the College Board, Freada Kapor's SMASH Academy, Black Girls Code, Girls Who Code, Code.org and major tech companies such as Google -- all searching for the best way to put computers and computer know-how in the hands of kids from all racial, ethnic and socioeconomic backgrounds.
The idea that university CS programs are taking bright young minds and fashioning them into algorithm and data structure whiz-kids defies the observations of almost any incoming CS student or their instructor. Many CS freshmen enter college already having a passion for computers and likely a privileged amount of access to technology and mentorship. Like myself, they were given computers as children by parents who were themselves close to technology. They have computer usage skills (how to configure your machine, how to fix basic computer problems) and have parents (or tutors) who introduced them to programming. For those without that background, freshman CS can prove very challenging.
Notably, as with many courses taken during one's educational career, computer science also teaches many generalizable skills. Computer science is much more than learning to code, and its benefits go beyond knowing a particular programming language. Computer science teaches students about logic, understanding systems and engineering and design basics, all of which are applicable to other academic and career fields. Perhaps this is why correlational data show that learning computer science is associated with higher math achievement. Computer science coursework also naturally lends itself to 21st century skills like collaboration, problem-solving and creativity, which are valuable and highly sought-after skills in the modern workplace.
Among those pledging their support are Apple CEO Tim Cook, Facebook CEO Mark Zuckerberg, Facebook COO (and bestselling author) Sheryl Sandberg, Bill and Melinda Gates, IAC Chairman Barry Diller, Walmart CEO Doug McMillion. The tech names are joined by California Governor Jerry Brown, Arkansas Governor Asa Hutchinson, and 26 other state governors, evenly split across party lines. The educators on board include Oakland schools Superintendent Antwan Wilson, NYC Department of Education Chancellor Carmen Fariña, and NAACP President and CEO Cornell Brooks, among others. "The breadth of support shows that computer science isn't just a tech problem anymore, it's an America problem," Code.org "And it's not just a Democrat issue, it's the most bipartisan issue in the U.S."Earlier this year, President Obama also hoped to reignite the country's "spirit of innovation," calling for 4 billion to improve computer science curriculums in every K-12 school across the country.
One way to foster these skills in students is through the use of digital simulations or microworlds, which allow students to conduct authentic inquiry of scientific phenomena the way scientists do. Students generate a hypothesis, test it by changing values in the simulation, interpret their data, warrant claims with data, and communicate findings about their discovery. Although such digital environments are not new, what is new is the development and application of educational data mining to analyze students' data for assessment and real time tutoring of core disciplinary knowledge and inquiry practices. Educational data mining offers more than the traditional statistics used on typical, multiple-choice tests. These high-fidelity data are in the form of log files from mouse clicks within the digital learning environment.