Capacity for Computer Science in Wisconsin
The first component addressed by the CAPE framework is the “Capacity” for CS education. This is measured by the availability of resources such as faculty, funding, and policies that support high-quality CS instruction. Without first building these resources to create the capacity for CS education, Wisconsin cannot hope to improve the access to, participation in, or experience of CS education. Fortunately, the last decade of innovations in computing and the push for CS education, has made tools for teaching CS more affordable and widely available. It is now feasible for most schools to provide students routine access to computer hardware such as chromebooks or other inexpensive PCs, laptops and tablet computers. Additionally, specialized and expensive hardware is no longer a prerequisite to teaching CS in the classroom, and many high quality teaching materials and CS curriculum options are now available at little or no cost. This means that even districts with less funding have the ability to provide a high-quality CS education.
For school districts seeking to adopt a CS program, it is easier than ever to obtain the right tools. However, these tools are little use without a proper wielder. While attempting to measure Wisconsin’s capacity for CS education, we discovered a concerning trend in how new CS teacher production compares to the number of new schools adopting CS programs. Despite the state requirement for CS-specific licensure, nearly 30% of public high schools with CS enrollment do not report a teacher with proper certification. Furthermore, while an increasing number of Wisconsin schools are adopting CS with each new year, the number of new CS teachers we are producing is abysmally low by comparison. Although there have been a number of successful professional development programs to boost certified teacher production in the past few years, it is evident that it has not been enough to combat the new growth in CS education and the number of teachers leaving the system due to retirement or other factors. If Wisconsin hopes to increase capacity for CS education, we need to focus our immediate attention toward professional development for CS educators.
The number of highly qualified and active CS teachers in the state has steadily grown since 2014, due in part to multiple rounds of grant funded interventions sponsored by federal, non-profit and industrial sources. CS professional development opportunities in the state have included university teacher preparation programs, in-service training of current elementary and middle school teachers, as well as alternative programs such as The Urban Learning Collaborative (ULC)’s computer science certification. Support for Marquette University’s PUMP-CS Project alone has included:
In addition, scores of individual schools and districts have provided space for on-site professional development workshops for their teachers, or have supported their teachers as they travel to participate in centrally located professional development workshops.
The cost of high-quality teacher professional development varies widely based on scale, location, and depth of coverage. The PUMP-CS Project has provided approximately 32,000 hours of CS professional development between 2012 and 2021, and thus can provide insight into typical costs of CS professional development workshops.
For non-specialist, general classroom K-5 teachers, the PUMP-CS Project recommends 14 hours (2 days) of in-service training for the Code.org CS Fundamentals (“CSF”) curriculum – one introductory day, and a second follow-up workshop once teachers have experience starting to implement and integrate the curriculum in their classroom. The goal of the professional development is to prepare K-5 classroom teachers with no prior CS background to integrate approximately 20 hours of grade-specific, plugged and unplugged lesson plans into the school year. At scale, such training by a qualified facilitator, including curriculum supplies, can be accomplished with roughly $35 / day / teacher. This estimate assumes that the curriculum itself is provided at no-cost. This estimate does not include refreshments or meals for the participants, parking, travel reimbursements or lodging costs, or compensation for the teachers’ time, all of which may be considered part of professional development norms based on location and context.
For middle grades teachers, the goal of CS professional development is often preparation to teach all or part of a year-long, stand-alone course in CS, such as Code.org CS Discoveries, or to integrate specific computing-based modules into an existing STEM course, such as Project GUTS. Preparation to teach such a course is necessarily more involved because of the higher level of the course and the greater volume of content. Teacher professional development for a stand-alone, year-long CS course is often at least 60 hours of training. Because this is more time than can be practically released during the school schedule, professional development workshops of this type normally take place during summer months and during Saturdays during the school year.
High school teachers normally attend CS professional development targeted to teaching a specific stand-alone course, such as:
Given the large investments in teacher time and effort required by many of these professional development options, it becomes apparent why external grant funding for scholarship dollars have been an essential ingredient for growing the CS teacher population of the state. The full cost of nine days of face-to-face professional development for a typical high school course, (including teacher compensation and travel costs to spend a week of the summer and four school-year Saturdays in another city,) costs approximately $1400 / teacher at scale. While often measurably less effective, distance learning professional development can offer the advantage of greater geographical reach and lower cost.
Given the prevalence of tight discretionary professional development budgets in the state, many schools – particularly those in less affluent districts that do not already have CS – cannot afford the direct cost of CS teacher professional development. As a matter of course, schools and teachers are accustomed to regular professional development to keep their skills current in many other fields. As a relative newcomer in this context, CS is often regarded as a new, unfunded mandate from the state. With fixed budgets, allocating support for CS professional development leads to reducing support for some other field. Viewed through the lens of this zero-sum game, it is no surprise that CS professional development is usually only occurring at scale where new, external funding has been made available.
For fiscal year 2022, other states have allocated more than $65 million for K-12 CS education, more than in any previous year [Code.org et al., 2021]. In Wisconsin, the governor’s proposed biennial budget included $750,000 for CS education, but this item did not make it to final passage. The problem of how to pay for expanding, widespread CS teacher professional development across WI remains a vexing challenge.
Code.org’s CS Fundamentals curriculum for grades K-5 consists of six freely available, grade-specific courses designed to be integrated into the regular elementary classroom by a teacher. Each course comprises approximately 20 hours of instruction in a mix of plugged and unplugged lessons. Code.org reports over 11,000 teacher accounts in WI, and more than 486,000 WI student accounts on their platform in 2021. (This is across all grade bands and Code.org courses, K-12.)
More than 2,300 elementary grade teachers across 530 WI schools in 231 districts have participated in 1-day CS Fundamentals professional development workshops between 2014 and 2021. Those trained teachers have delivered over 132,000 student-years[1] of CS instruction through the Code.org platform in grades K-5 to date. With roughly 29,000 kindergarten and elementary teachers in the state, about 7% of K-5 teachers have now received at least one day of training on this specific curriculum. While other excellent K-5 CS options exist from a variety of vendors, there do not appear to be any other efforts of similar scale to the CSF rollout, nor are there comprehensive training and implementation statistics available for alternatives.
While 7% of WI K-5 teachers have participated in a CSF professional development workshop, only 45% of those teachers (1039) were subsequently able to successfully integrate that content into their classroom, a percentage that is consistent with implementation rates seen across the U.S. Only 15% of in-service teachers trained in CS at the K-5 level (350 teachers) actually completed the course they were trained in with their students.
The takeaway for this landscape report is that even when CS curriculum is freely available and no-cost professional development is plentiful, it remains difficult for K-5 teachers to systematically integrate CS instruction into their classroom. Barriers typically cited by teachers in follow-up surveys include not enough instructional time with students, not enough preparation time to become comfortable with the material, and lack of support from administration.
Teachers who are successful in integrating CSF into their classroom are more likely to teach in a context where other teachers in their grade level also participated in the professional development, and the group were given time and permission to develop a coordinated plan for rolling out the curriculum. Successful CSF teachers were also more likely than their peers to view CS instruction as a “new take” on learning standards they were already implementing in their classroom, rather than as totally new content they needed to make room for in an already packed day.
This focus on integration of CS into the existing day, rather than making room for it by teaching less of something else, seems to be a hallmark of successful CS teaching at the K-8 level. PUMP-CS Project’s newest collaboration with Sacred Heart University in CT is predicated on this observation, and is two years into a five-year study examining long-term integration strategies with 30 elementary schools and hundreds of teachers. It is hoped that this research study will validate new, scalable strategies for improving implementation rates and sustained integration in the K-5 classroom.
Other instructional models we have seen in use focus on after school coding activities for students, or rely on a dedicated specialist teacher to deliver CS content in the school day, rather than classroom teachers. After school coding clubs can be very impactful, and are an easy solution for schools looking to start their first CS experience for students. However, from the perspective of long-term, sustainable CS education experiences for all students, it is important to aim for classroom exposure during the school day.
Many schools have success dedicating a specific building staff member to rolling out CS experiences for their students. Tech integration or library media specialists can be enthusiastic proponents for CS education, and may have the luxury of seeing all of the students in a particular building for dedicated “specials” time. Much depends on context, however. Prior studies of school structures with dedicated elementary science specialists, for example, found that students would often receive less exposure to science content than in the more traditional classroom integration model. When classroom teachers know that it is some other teacher’s responsibility to ensure their students’ exposure to science or CS, they adjust accordingly. Models that assume a dedicated specialist is both running their own classroom full of students – while also planning integrated lessons with classroom teachers – similarly result in a disappointing reduction in actual overall coverage. Full integration by classroom teachers, with the support of dedicated specialists, remains the preferred target for CS education.
Wisconsin provides multiple, well-documented pathways for individuals to become licensed teachers [WPDI-Pathways, 2021]. We focus on three specific pathways here:
Wisconsin requires teachers to hold a 405/1405[2] CS license to teach CS courses. However, any licensed regular education elementary teacher may teach computer literacy to their own self-contained classroom at their licensed grade levels, and computer application and literacy courses within specific disciplines may be taught by any licensed teacher in that discipline. Updated guidance from DPI now defines CS courses in terms of content and the CS academic standards, rather than by the deprecated guidance that a CS course contained “at least 25% computer programming.”
Often considered the “traditional” way of becoming a teacher, initial teacher pathways include:
As can be seen from Figure 2, the number of institutions that have produced initial teachers licensed in CS over the past nine years has been dismal, and is declining.
Figure 2: Wisconsin initial CS Teacher production data by year and institution, as reported by U.S. Department of Education Title II data.
As a whole, the Wisconsin initial teacher preparation pipeline has produced an average of only three new CS teachers per year over the past nine years, and that average drops to two per year in the past five. What may be worse than the already dire situation shown in Figure 2 is that nearly half of these institutions have ceased to produce any new CS teachers. The combination of low student demand and changing criterion for initial teacher licensing program renewals by the state have resulted in many extant CS education programs closing their doors, leaving ever fewer pathways for pre-service teachers to earn an initial educator’s license in CS. This stands in stark contrast to the level of demand for CS teachers, as we shall see below both in the number of unlicensed instructors teaching CS and the salary premium that districts are paying for CS teachers. In this particular case, market pressures caused by the demand for more CS teachers have not translated either to more CS education programs, or more pre-service teachers choosing CS as their field of study. A general shortage of pre-service teachers in all areas, but especially STEM fields, plus the stark salary gap between career educators and CS professionals combine to increase the likelihood of continuing anemic new CS initial teacher production for the immediate future.
For in-service teachers already licensed in another academic content area at the appropriate grade levels, there are currently two pathways in Wisconsin which allow a teacher to add a CS 1405 teaching license:
For simplicity, we refer to these pathways as add-on CS licenses.
While Content Test pathways have existed for other academic areas for many years, WI added a Content Test for CS only in 2017. Since that time, the option to add-on a CS license by passing the CS Praxis exam has proven to be the most important source of new CS teachers. To pursue the Content Test pathway, many teachers have made connections through the CSTA Wisconsin-Dairyland chapter, the local professional association for WI CS teachers. The chapter’s Google Group often shares resources and information on study groups that help prepare educators for the content area Praxis test on CS. For the district sponsored pathway, teachers who hold a three-year ‘License with Stipulations’ must submit a portfolio of evidence to demonstrate their proficiency in the new license area. Information on how to assemble the portfolio and the rubrics that will be used to evaluate the portfolio can be found in the LWS3 Handbook.
A new and infrequently exercised pathway to licensure is the Professional Teaching Permit pathway, which allows a college graduate with a bachelor’s degree in CS or related field, plus five years of industry experience in that field, to seek a permit in an acknowledged shortage area like CS. The candidate must complete 100 hours of formal instruction in education and pass the subject-specific content test. We are aware of only one current active CS teacher who has entered the profession through this pathway since its inception.
Our goal is to examine how computer science is playing out through state data. The aspirational objective is to understand how to get more teachers engaged in certification. To achieve this, we combined K-12 CS student enrollment data with CS licensure data for each high school in the past five academic years. As shown in Figure 2, we are grouping this information into four categories to examine:
Figure 3: Count of High schools broken down by a combination of certified CS teachers plus DPI-reported CS enrollment by year. As we approach the convergence of -C + E (Gray) with +C + E (Blue), we reach the tipping point where increasing CS demand outstrips the flat or declining supply of CS teachers.
The number of schools with certified CS teachers and students enrolled in CS courses (blue line) has steadily increased by ~20% (from 145 to 172) between 2016 and 2020. With the recent enrollment data, we notice a slight drop of 12 high schools with +C + E. This can occur due to several factors, such as teachers might come to retirement age or that school no longer offering a CS course and is slowly on the verge of falling into -C – E. At the same time, the number of schools with certified CS teachers and without students enrolled in CS courses (orange line) has experienced a mirror-image decline by 46% (from 105 to 56). Some portion of this orange line decline represents inactive CS teachers leaving the classroom, whether due to retirement, death, leaving the field, etc. The other portion of the orange line decline represents inactive CS teachers being called back into service, switching to the blue line as demand from students and other stakeholders opens new CS courses in school buildings that until recently offered no CS.
Of note for the future, both the blue and orange trend lines appear to be reaching their respective floor and ceiling. We’re reaching to the end of the supply of existing certified inactive CS teachers (orange line) who can be easily reactivated by scheduling a course and enrolling students. Existing certified CS teachers (blue line) have reached their saturation point after this multi-year surge in student interest in CS. There is a limit on how many such teachers existed in the first place, and a certain base number of certified CS teachers will likely always appear in the gray line due to work in an administrative role, demands on their teaching load from other fields they teach, or working in a smaller school where CS is offered less than every year.
When a certified CS teacher from blue or orange lines permanently leaves the CS classroom, it remains exceedingly difficult for schools to hire a new certified CS teacher as a replacement. Given the high expense and difficulty for teachers in another field to retrain and add on a CS license, many schools will have few alternatives but to rely on a teacher without the certification or put their CS program on hiatus. The final major point to be made from Figure 2 is the significant number of schools represented by the gray line, in which there is DPI-reported CS enrollment in courses, but the teacher lacks a certification in CS. Some of this may be due to course subject ambiguity. Many high school courses exist that combine some CS content with other topics, like digital literacy, business, or technical education topics. Courses that are primarily digital literacy, for example, can be taught by any teacher who is certified at grade level, regardless of the subject area. But given the constraints on acquiring new CS teachers or getting existing teachers into a CS program, it is apparent that many of the schools represented by the gray trend line have CS courses being taught by teachers who lack the formal CS subject certificate required under state law.
On a particular note, the gray line has increased by an alarming rate of 28% in 2020-21 AY. This may occur due to new schools starting to offer CS courses for the first time and it is assigned to a teacher who does not possess a certification to teach CS courses. Another explanation for the rise in the gray line could be that schools that had CS enrollment and a certified teacher might have retired, moved, etc. making them fall into the gray line. As we approach the convergence of -C + E (Gray) with +C + E (Blue), we reach the tipping point where increasing CS demand outstrips the flat or declining supply of active certified CS teachers in the state. Regardless, this is a systemic failure to see that the gray line is on the path to intersect with the blue line. The trend makes it clear that the intersection is going to occur or has already intersected in the current AY. In each of the past four years, more than 40% of regular public high schools offering CS courses in the state -- and more than 20% of all regular public high schools last year -- are offering CS with a teacher who is not properly credentialed in the teaching and learning of CS, and who is also not enrolled in an appropriate program to add-on a CS license. These high schools should be the target for the state and our research to support and provide enough resources for new teachers teaching a CS course and guide them to participate in the nationwide certification exam called the PRAXIS. Specifically, high schools that recently had CS enrollment require attention to maintain the enrollment rate for the future.
[1] We count each separate academic year in which an individual student measurably encounters CS curricula as a student-year.
[2] The CS license was 405 under the old license numbering system, and is now 1405 under the new system.