Teaching Responsible Computing Playbook

Broadening Participation and Responsible Computing

Authors: Augustin Chaintreau, Udayan Das, Sorelle Friedler

Responsible computing by definition cannot be done without input from all groups and differing viewpoints (see the related section on discussing justice and equity). More generally, the need for a diverse computing workforce is simply that since computing creates products that affect a diverse user base, to create more effective products (or to make products that do not harm marginalized communities) one needs to have computing professionals who are diverse as well. In this section, we consider this in the narrower context of teaching responsible computing in class. Specifically, to effectively teach responsible computing in a course, a diverse student body is critical in fostering meaningful but difficult discussions in class where discussion will be hampered if a diversity of lived experiences relating to the studied topic are not present in the classroom. Finally, for students to share their lived experiences, the classroom must be welcoming and inclusive.

Of course to have a diverse classroom, the student population itself must be diverse. To achieve the latter, there must be a concerted effort on both recruiting and retaining a diverse student population. This falls under the wider umbrella of broadening participation: we will keep the scope of this section modest and consider how broadening participation interacts specifically with responsible computing. Then to close the loop, we will argue how responsible computing itself can be used as a tool to recruit and retain a more diverse set of students.

The lack of diversity among the computing student population is well documented: e.g. as per the 2019 Taulbee survey, only about 20% of undergraduate computer science students are women, down from the highs of 37.1% in 1983-84. The number of students from other under-represented groups is even lower. The reasons for this small participation from under-represented groups are many and structural (and specific to the US-- e.g. the percentage of women computing students in India is at a much higher 40% in 2018-19 (based on the 2018-2019 AISHE report)) but with the right effort and resources computing departments in places like CMU, Harvey Mudd and University of Washington have either achieved gender parity or have a much higher percentage of women students than the national average. While there is no silver bullet, one common factor in these success stories was to redesign the computing curriculum.

There has been a lot of research done on broadening participation, we focus on one aspect that is most relevant to responsible computing. Women and first-generation students (and likely other underrepresented groups) are more likely to hold communal goals as opposed to individually-focused goals, and so integrating curricular changes that speak to these goals can help retain these groups in computing (Lewis et al, 2019, Diekman and Steinberg, 2013, Allen et al, 2015). While there does not seem to be any research that specifically connects responsible computing ideas to broadening participation or retaining students from underrepresented groups (and if so, consider this enthusiastic encouragement to research and explore this connection) it seems reasonable to hypothesize that given the ways that responsible computing curricula bring community-focused perspectives into the classroom these curricular changes would also help to broaden participation and retain students.

Conversely, there are many techniques we can use to broaden participation in computing and these efforts should be considered part of a responsible computing curriculum, since responsible computing can not be achieved without participation from all groups. Along these lines, we highlight two important issues that can be handled just at the university level: 1) what kinds of barriers are there for entry into your programs, and 2) what kinds of supports are there within the program that will enable students to succeed despite systemic disadvantages. When considering barriers it is important to understand that many gatekeeping requirements for Computing programs tend to filter out historically/traditionally disadvantaged groups. Examples of such barriers are the application process (e.g.: SAT/ACT requirements, GPA requirements, co-curricular requirements etc. each of which may not accurately predict student success but nevertheless disadvantage certain groups of students), non-flexible programs (e.g., full time status requirements that disadvantage students who may not be able to attend college full-time; sometimes such requirements are implicit in that a fee discount only kicks in at a certain number of credits and already financially struggling student may be forced to take on more credits than would be optimal), curriculum factors (noted above), lack of diversity in faculty and student body, and social circumstances (e.g., issues with housing, finances, mental health etc.). The last two factors are crucially where supports are critically important. Students need to have a mechanism of connecting with others in their social/demographic group, which can impact both their success as well as retention. Having support staff that can work with students to address non-academic issues that can impact their performance is critical. When a student is undergoing a crisis where do they go for support? Further, if those supports exist, how do students discover these resources? Faculty can be invaluable in this process as far as identifying students who are struggling, but this may require training. Finally, initial investment needs to be made in fostering student clubs and organizations to facilitate student-to-student contact. Minority students shouldn’t be left in the lurch to find others of their kind.

We circle back to more direct connections to responsible computing curriculum. As mentioned above, while recruiting a diverse student population is important and a necessary condition, getting students from under-represented groups through the door is only part of the job. Retaining such students is just as important-- if such students do not see how computing can help their communities, then they might not relate to computing and could drop out. Again responsible computing embedded in the curriculum can be helpful here-- see the section on service learning for a more detailed discussion on one way to do this.

We wind down this discussion by pointing out that the classrooms need to be more inclusive to facilitate more meaningful discussion of responsible computing topics. This is a longer topic for discussion, which is covered in more details in the difficult conversations section.

Key Questions:

Is there an existing diverse set of students in the university that can be recruited to the course? If the department already has a diverse student body then the answer to this question is essentially yes. However, many computing departments do not have a very diverse student population. In such a case couple of other options: (i) The course could have a section specifically for non-computing students to register (though if the course does not satisfy degree requirements for non-computing students this might not be effective); (ii) have projects that span across the computing course and non-computing courses so that computing students get an opportunity to meaningfully interact with non-computing students.
Are there barriers preventing a broader group of students from applying to the program? Is there an existing plan to recruit a more diverse class of incoming computing students? A longer-term strategy is to recruit a more diverse set of incoming computing students to the department. If students are admitted directly into a computing department when they join the school, there needs to be more coordination between the department and the admission office as well as a concerted effort to create a pipeline to recruit students from under-represented groups from e.g. local area schools. A note of caution-- such efforts need investments in time and resources over multiple years: e.g. establishing a working relationship with a local school takes time to establish (schools have a busy schedule and it requires a lot of advance notice and planning to schedule an outside activity). Consider incorporating responsible computing material in outreach efforts to local schools to give a more accurate picture instead of only highlighting the virtues of computing. If students declare their major after being admitted to the university, then the recruiting effort has to focus on students that are already there in the university. These might be students who are in introductory computing courses but hadn’t previously thought of themselves as computing majors or students who would be interested in taking a computing course if it was explicitly related to their major, e.g., bioinformatics, computational linguistics, or other interdisciplinary courses. Explicit outreach to other departments to highlight such interdisciplinary connections can also be generally useful for responsible computing efforts. Such effort can also be part of broadening participation in computing plans that are now required for some NSF proposals.
Is there an existing plan to recruit a more diverse faculty? It has been well established that students from underrepresented groups respond to faculty who look like them, which means having a more diverse faculty helps recruit a more diverse set of students as well. Again, this needs close coordination between the department and the rest of the university. One option, if looking for research-intensive tenure-track faculty, is to consider recruiting from researchers working on responsible computing, which anecdotally is more diverse than other computing research areas.
Is there a plan to retain students from underrepresented groups? Research has been conducted on changes to make to the curriculum and the department as a whole that can be implemented to help with retaining students (see e.g. Lewis et al, 2019, Diekman and Steinberg, 2013, Allen et al, 2015, Searle and Karai, 2015 for changes to curriculum and Newhall et al, 2014 for making the TA team more diverse). While this has not been rigorously established, there is some evidence to believe that responsible computing material can bring community-focused notions into the classroom, which has been shown to help to broaden participation and retain students. (See section on service learning for more ideas.) Are strong support services (including for social circumstances e.g., issues with housing, finances, mental health etc.) a key component of those plans? Strong support services can also have a positive impact for the whole student body.
Is there a plan to be more inclusive of under-represented students in the creation of the course? When discussing responsible computing topics, the instructor must be mindful of the fact that there will be students in the class, who by their lived experience might have a deeper understanding of the topics than the instructor. If possible, the instructor should consider working with (and paying) such students before the relevant class to help serve as an additional classroom expert. The instructor should consider how they can co-teach these topics with such students serving as content expert TAs or in other ways directly helping to shape the course material and pedagogy.

Checklist

☐ Establish a plan to recruit a diverse set of students already in the university to the course.

☐ Establish a plan to recruit a more diverse class of incoming computing students.

☐ Establish a plan to recruit a diverse faculty.

☐ Establish a plan to retain a diverse student population

☐ Involve students with relevant lived experience to develop responsible computing material in courses.

Examples

Loyola University Chicago

Loyola University Chicago

At Loyola University Chicago’s School of Continuing and Professional Studies (SCPS), Students with almost no background have successfully completed programs and achieved significant success. CS certificate students are particularly noteworthy because they often come with little or no prior CS/Tech background and have gone on to pursue grad school in top 20 CS departments (e.g., University of Illinois, Georgia Tech). A massage therapist with no prior programming background, within 8 months from when they began the program was employed as a software developer, during COVID19 pandemic; a laundry list of requirements would perhaps have been detrimental in not enabling them to make a career change so smoothly and quickly. Does a long list of requirements and prerequisites ensure student success or ensure that certain students will never consider CS as a major?

Checklist walk-through (Here, we show how some of the checklist items in the previous section of this document could be used. The checklist did not exist when this partnership was created, so this walk-through is meant only for illustrative purposes.)

Establish a plan to recruit a more diverse class of incoming computing students.
- Since Fall 2018 all incoming threshold requirements (such as Math assessment) have been removed from the Computer Science Certificate and BA Information Technology programs.
Establish a plan to retain a diverse student population.
- Instruction focuses on revised curriculum -- heavily incorporating active learning, Jesuit ethics framework, and focus on belonging in the profession -- as well as in-course (1-on-1 meetings with instructor and Program Director) and out-of-course supports (SCPS has extremely strong advising and student support beginning from pre-enrollment through to graduation; student support staff are in continuous communication with students to address issues and challenges as they arise). Students are encouraged to reach out for support right from orientation and they do; part of the equation is explaining, and repeating, that there is no shame in seeking support and that that is not an indication of being less-than.

Resources

Canning, Elizabeth A., Katherine Muenks, Dorainne J. Green, and Mary C. Murphy. "STEM faculty who believe ability is fixed have larger racial achievement gaps and inspire less student motivation in their classes." Science advances 5, no. 2 (2019): eaau4734.
Leslie, Sarah-Jane, Andrei Cimpian, Meredith Meyer, and Edward Freeland. "Expectations of brilliance underlie gender distributions across academic disciplines." Science 347, no. 6219 (2015): 262-265.
Lewis, Colleen, Paul Bruno, Jonathan Raygoza, and Julia Wang. "Alignment of goals and perceptions of computing predicts students' sense of belonging in computing." In Proceedings of the 2019 ACM Conference on International Computing Education Research, pp. 11-19. 2019.
Tucker, Jessica, and David Ferguson. "Incorporating ethics and social responsibility in undergraduate engineering education." In International Conference on Engineering Education–ICEE, pp. 3-7. 2007.
General Broadening Participation in Computing resource: BPCnet