Towards Establishing a Training Program to Support Future CS Teaching-focused Faculty

Computer Science programs have seen high enrollments in recent years, which contributed to widening the capacity gap. One way to address this problem is to hire more teaching-focused faculty at both research and non-doctoral granting institutions. Although this kind of hiring has already been taking place in several institutions, PhD-granting CS departments have not been able to produce enough PhDs to meet the increasing demand, especially for PhD holders with interest in - and capacity for - teaching. In this paper, we describe our experience with the initial phase of building a training program within our (large, land grant, R1) institution, targeting graduate students interested in pursuing an academic teaching-focused career in CS. Through a semester-long set of meetings, conversations, and activities, we worked with participants on improving their teaching skills and applying effective pedagogies in the classroom. At the end of the semester, we surveyed participants about the value of those meetings to them, ideas for improvement, and perspectives for future directions. Most participants rated the meetings positively in terms of content relevance and usefulness, and the opportunity to connect and interact with other participants and invited faculty members. We also discuss the lessons learned and best practices, which can be widely applied by other departments looking to better prepare their graduate students for a CS teaching-focused faculty position.


INTRODUCTION
The enrollments in Computer Science (CS) and computing-related programs at the university level is skyrocketing with more students eager to learn about CS and pursue a CS career after graduation [20].Between 2008 and 2018, CS departments across the United States have experienced a tremendous increase in undergraduate enrollments [9].One reason behind this increasing interest in CS and CS-related degrees is that computing skills are now pervasive and essential across many domains [15].Demand is also increasing for non-major students to learn and acquire CS skills [17].
This situation created a CS capacity crisis [5], because it adds pressure on CS departments at both R1 institutions and non-doctoral colleges and universities.More than 65% of doctoral-granting institutions have already started working on addressing this problem by taking four actions [1]: (1) Involving Undergraduate Teaching Assistants (UTAs) and tutors in departmental teaching activities.(2) Relying on Graduate Teaching Assistants (GTAs) as instructors.
The first and second actions in the above list were directed towards hiring more GTAs and UTAs, and giving them more teaching responsibilities.The literature describes attempts to train TAs to better accomplish their teaching tasks (as discussed in Section 2).
Regarding the fourth action, there is an increasing demand to hire CS teaching-focused faculty, with many universities advertising for this kind of position [18].Teaching faculty positions are different from traditional tenured/tenure-track positions with more focus on the instructional mission of CS departments and the scholarship of teaching and learning.Teaching faculty take more teaching load than their tenured/tenure-track peers, and they are expected to remain active in the Computer Science Education (CSEd) community [3].Teaching faculty candidates do not always need to have a PhD, with many candidates hired with a Master's degree [18].
According to Taulbee report 2022 [20], about 3.1% of CS PhD holders got a teaching-focused faculty position in North America, and there is a 27.3% increase in the number of teaching-faculty hires reported in the 2021-22 academic year as compared to 2020-21.
The increase in hiring teaching-focused faculty doesn't match the surge in CS student enrollments in a way that will help in addressing the CS capacity crisis [5,12].In the academic year 2021-2022, there were 278 computing-related teaching faculty positions available in North America, and only 173 were filled (62.2%).Reasons behind unsuccessful teaching faculty searches include: • Lack of qualified teaching faculty applicants [9,20].
• Many faculty believe that teaching ability is a preordained skill, and so many graduate students are not motivated to improve their teaching skills and apply for teaching faculty positions [15].• The differences in responsibilities, education, and employment conditions between teaching-focused faculty positions and traditional tenured/tenure-track positions are not well understood [5].• Applicants were not well prepared [20].
According to Taulbee 2022 [20] report, there is an expected 2year growth (2024-25 as compared to 2022-23) of about 29.5% in the number of teaching faculty in North American computing departments.Recognizing this high demand for teaching-focused faculty on the one hand, and the historic lack of applicants' preparedness and expertise on the other, we have taken on the task to build a training program to better prepare our graduate students who have a clear interest in applying for teaching-focused faculty positions for the job market after graduation.Furthermore, we aim to improve the reputation of our CS department as a place to become trained as a future teaching faculty member, as this naturally fits with our department's existing strengths in CS Education Research.As a first step towards achieving these goals, we organized a set of meetings within our department involving selected graduate students with previous teaching experience and who are interested in applying for a teaching-focused faculty position after graduation, and invited faculty members.
We received a small grant of $2000 from the Center of Excellence in Teaching and Learning at Virginia Tech to partially support this semester-long program, which helped us organize six meetings during Spring 2023 with selected participants facilitated by invited faculty members (referred to as the "invited faculty" hereafter) to share experiences, engage in group conversations, and to provide advice on how to improve teaching skills and apply effective pedagogies in the classroom.We invited participation from faculty in both the CS and Education departments at Virginia Tech.These meetings were also a good opportunity to learn about the participants' teaching concerns and challenges, and to collect their input to guide our thinking toward our next steps.
This paper is organized as follows.Section 2 discusses the relevant literature.Section 3 describes the planning for the program meetings and the activities performed in each meeting.Section 4 presents how the meetings were evaluated and describes our methodology and results from a survey carried out at the end of Spring 2023.Section 5 discusses the evaluation results and lessons learned.Finally, Section 6 presents conclusions and future directions for the subsequent program activities.

RELATED WORK 2.1 CS TA Training
As mentioned in Section 1, one strategy to address the CS capacity crisis is to rely on more TAs and/or to assign them more teaching responsibilities.Teaching responsibilities include but are not limited to: leading recitation sessions, helping with assignment grading, writing test cases for auto-graded programming assignments, holding office hours, and serving as instructors of record.Based on these roles, TAs have a lot of interaction with students, even more than course instructors in some cases.Accordingly, TAs can impact student retention and broadening participation in CS [11].However, in many cases, TAs are not prepared to carry out these teaching responsibilities [16].
Some research describes the development and evaluation of interventions to improve TAs' teaching skills, so that they can be better prepared for their teaching tasks.These attempts employ training through seminars [7], courses and self-paced online modules [4,6,11,12,16], and practical training without any programs or courses [13].
Most of these training approaches are directed towards pedagogical skills.In [16], a five-module, non-credit course in a blendedlearning format was required for newly hired TAs to train them in conducting tutorials, providing feedback on student's assignments, and conducting oral and coding assignments.In [4], a twelve-week TA training program was developed to fill in the gap between actual and desired TA teaching skills.The main goal was to help CS instructors scale their courses.In [11], a course with three modalities (synchronous, self-paced, and hybrid) was developed for first-time TAs, helping them to navigate the mechanics of their new role.In [7], a seminar-based course was developed for first-time CS TAs.This was a required course with a P/F grading scheme.Seminar participants performed much better on student evaluations than non-participants.
Some research focused on improving TAs' interpersonal skills.For example, the MaGE training curriculum [14] of Mount Holyoke College prepared CS students for inclusive peer mentoring and teaching through active learning.In [12], an adjustable TA training curriculum was developed to help TAs to be more responsible and empathetic in the classroom.In [6], a scalable and modular TA training curriculum was developed that integrated training in pedagogical skills and interpersonal skills.The goal was to foster an inclusive, welcoming, and empathetic classroom.
Most of these attempts were only targeting first-time TAs who did not have previous teaching experience in their departments.Furthermore, the training was mostly department-specific, focusing on helping TAs carrying out their day-to-day teaching tasks with little improvement in their general pedagogical knowledge and teaching preparedness.It can be argued that these attempts might not be taken seriously by trainees, especially as many of them don't have plans for teaching-focused careers.

CS Faculty Training
The graduate school at Virginia Tech offers the "Future Professoriate Certificate"1 to prepare graduate students for faculty positions.Many other institutions offer similar certificates.For example, the "Harvard Bok Higher Education Certificate"2 offered by Harvard's Derek Bok center for teaching and learning aims to explore various approaches to higher education pedagogy, and to investigate the most relevant research on students' learning.These certificates focus on existing practices in higher education, but they are not CS-specific.
Relatively little literature exists related to formal interventions to improve the teaching skills of prospective or newly hired CS faculty.To our knowledge, this body of work did not target teaching-focused faculty specifically.
In [19], a faculty training program at a new CS university in Russia was developed based on the Instructional Skills Workshop (ISW) [10].Faculty participants indicated that their teaching practices were changed to a more learner-centered approach, and that they were motivated to experiment with new teaching tools and techniques as a direct result of participating in this program.In [15], a boot camp workshop on how to be effective in CS teaching was implemented.This workshop targeted research-focused CS faculty at research institutions, especially those who have some anxiety about teaching.The focus of the workshop was on how to succeed as a CS faculty, without spending time on general or inapplicable teaching strategies to CS.In [8], a preparing future faculty program was designed in the department of Electrical and Computer Engineering and Computer Science at the University of Cincinnati with the goal of improving undergraduate and graduate teaching and to help graduate students better prepare for their academic job search.Participants commented positively on the program in terms of their preparedness for academic jobs.
To our knowledge, building capacity for better preparing CS graduate students for a teaching-focused academic career is an area that would benefit from more research work.

PLANNING PROGRAM MEETINGS 3.1 Selecting Meetings Participants
At the beginning of Spring 2023, we emailed 148 GTAs and graduate instructors of record from the CS department at Virginia Tech, introducing them to our program and inviting them to fill out an interest form if they have clear plans of applying for a teachingfocused faculty position after graduation.We received responses from 29 applicants expressing their interest in joining this initial phase of the program.Due to funding and logistic restrictions, we were able to accommodate 23 applicants, prioritizing those who are: • Instructors of record.
• GTAs with clear intentions of applying for a teaching-focused faculty position after graduation as indicated in their interest statement.• GTAs with previous teaching experience.91.3% of the selected participants are PhD students, while the remaining are master's students.60% had been the instructor of record for at least one class before, while 40% have been just TAs.

Selecting Meetings Topics
Participants were asked to select five to ten topics that interest them the most from a list of 18 topics related to CS teaching and learning.Based on the responses, we selected five topics to be covered in the meetings, taking into consideration participants' interests as well as the availability of experienced faculty members to facilitate discussions on these topics.The proportion of participants who selected the topic in the original survey is presented in parenthesis in the list below.It's worth noting that we selected only five topics for six meetings, because the first meeting was dedicated to group introductions, kicking off the program, and explaining its structure and goals.

Meetings Details
We organized six meetings with the selected participants during Spring 2023.All meetings were scheduled on Fridays from 11:30am -1:00pm so that we can allow for a friendly interaction between participants while they are having lunch together with the invited faculty.Lunch cost was covered by the grant we received (see Section 1), so participants didn't need to worry about bringing their own lunch unless they wanted to.Participants were asked to talk freely to each other and discuss their experiences related to the meeting topic during the first half hour of the meeting.For the remaining hour, the invited faculty led a discussion with participants presenting their experience with the topic and asking participants to share their thoughts.Below we provide more details about the activities and discussions in each meeting.

Meeting 1: Introducing the program
The first two authors led this first meeting to motivate and introduce the program to participants.While eating their lunch during the first half hour, participants were asked to divide into groups of 5.They were given a discussion prompt to chat about the moment when they thought they successfully did their job as a teacher.Then each participant was asked to introduce him/herself to the whole group, stating why they are interested in being part of this program, their plans after graduation, and why a teaching-focused faculty position is appealing to them.The two leading faculty then discussed the increasing need for teaching-focused faculty in the US and in our own institution, the need for proper training of our graduates for those positions, and the opportunities we have in our department that can help in achieving this.Meeting 2: Academic Integrity and the Honor Code We invited the director of academic operations in the CS department, who is an experienced teaching-focused faculty member, to lead the second meeting about dealing with students' conflicts and enforcing the honor code.Participants were engaged in active discussions about a wide range of issues related to academic dishonesty and the ways to catch it.Furthermore, few participants talked about cases where they were disrespected by students as a result of reporting an academic dishonesty case.The invited faculty talked about his experience dealing with cases like that, and he emphasized gathering as much evidence as possible in the case of a conflict with a student.Meeting 3: Students Motivation For the third meeting, we invited a professor from the school of Education with a lot of research experience in examining instructional methods that support students' motivation and learning.The invited faculty presented issues that affect students' motivation in a course, such as student-instructor interaction, communicating course objectives early in the course, working in groups, assignment procrastination, lack of communication skills, lack of confidence, and social anxiety.The invited faculty talked about different ways of handling these issues by explaining approaches like mastery-based grading, grading students on group dynamics, having student testimonials about the course, using checkpoints (milestones) through the course to check student understanding, and empowering students by giving them the ability to steer course mechanics.

Meeting 4: Curriculum Design
The first author who is also a teaching-focused faculty in the CS department led the fourth meeting about course design.The main theme of this meeting was defining specific, measurable, and realistic Intended Learning Outcomes (ILOs), and using them to identify methods of instruction and to create matching assessment questions.The invited faculty gave examples about good and bad ILOs from real courses, and how the bad ILOs could be reidentified.To link ILOs with course assessment, the invited faculty introduced Bloom's Taxonomy (BT) as a framework to measure the cognitive effort required for each assessable ILO, and to design matching assessments at the same cognitive level.At the end of the meeting, participants were given 15 minutes to work on a group activity in which they were asked to think of a course they are currently teaching or willing to teach, and identify important topics they think their students should learn, identify the course ILOs, and think of assessment questions to measure the identified ILOs.Meeting 5: TA Management For the fifth meeting, the third author who is also the Associate Department Head for Graduate Studies in the CS department was invited to lead the session.This meeting was mainly about managing TAs from a departmental perspective.Participants were engaged in active discussions about how TAs are assigned to courses and the related funding and skill constraints.Issues like TA workload and evaluation were also discussed.One of the participants suggested having a shared resource for GTAs explaining the courses and the expectations to guide them in filling their course request form.Meeting 6: Diversity, Inclusion, Equity, and Accessibility The sixth session was led by the chair of the Diversity Committee in our CS Department, who is also an experienced teaching-focused faculty member.The invited faculty started the meeting by talking about approaches to improve students' sense of belonging to the class like learning students' names, using in-class examples that reach a wide audience, and avoiding saying words like "obviously", "it's simple" or other means of expressing instructor dominance.In addition, ways to improve students' belief that they can succeed were discussed like reminding students of their previous successes, flexible pace of work, having milestones for their deliverables, observing the success of "similar" students, reducing high stakes assignments, facilitating teamwork, and minimizing the level of anxiety.After that, participants were given about 20 minutes to complete a related in-class activity asking them to share the strategies they use or plan to use to support their students' belief they can succeed, benefits out of it, and anticipated challenges.At the end of this meeting, copies of "Make it stick: The science of successful learning" [2] were provided to participants.The cost for this was covered by the grant described in Section 1.

EVALUATION 4.1 Methodology
After the last meeting we sent an online survey via email to all 23 participants asking them to provide their input regarding the six meetings.The survey consisted of 17 questions, 13 of them were ratings on a scale from 1 to 5 (where 1 means strongly disagree and 5 means strongly agree) about the following: Four survey questions were open-ended, asking participants about the following: • What they liked the most about the meetings.
• Participants least favorite aspect(s) about the meetings.
• Ideas for improvement.
• Participants suggestions for future program activities.We received 15 responses (65% response rate) over a period of 12 days.The average completion time for the survey was six minutes.

Results
Overall, results were positive regarding the value of the meetings to participants.Table 1 summarizes the results from the first 13 questions.
Below, we present the results from the four open-ended questions.
What thing(s) did you like the most about the meetings?All 15 respondents reported that what they liked the most about the meetings was its interactive discussion-based nature.As described in Section 3.3, in all meetings, participants were asked to engage in active open discussions with each other and with the invited faculty as well to share their experiences, challenges, and questions.This gave them the opportunity to better connect with their colleagues and with the invited faculty, which is something desirable towards achieving our main objective of this initial phase.
What was/were your least favorite aspect(s) of the meetings?Five participants complained about the timing and duration of the meetings.Two participants mentioned that their schedule conflicted with the meetings time.Three of them mentioned that having just a 1-hour long meeting is insufficient.There were many topics to discuss, but the time was too short to cover everything.One negative aspect reported was related to the structure of the meetings and the absence of a systematic discussion around a single topic.In the TA management meeting, some participants asked off topic questions unrelated to the current discussion.This made it hard for some participants to follow up and engage with the discussions.
In your opinion, how could the meetings be improved to better serve the needs of participants?A common theme we found in the responses was to expand the duration of the meetings and to cover more topics.Three participants wanted more time for the meetings to better allow more participants to talk.One participant suggested addressing GTAs' depression and mental health.One participant suggested having experienced instructors observe the participants while they are teaching in their classroom and giving them feedback.Another participant asked to have a clear agenda for each meeting posted at least one or two days before the meeting time to allow participants to better prepare their ideas and questions.
Any additional comments or feedback about the meetings and suggestions for future program activities?For future program activities, participants wanted future rounds of these meetings to take into consideration the points mentioned in the previous question.One participant noted wanting a different cohort of invited faculty so as to hear from diverse perspectives.Another participant wanted written notes or a study guide for the material presented.Another participant wanted more topics to be covered, like syllabus requirements, consistent GTA grading using clear rubrics, keeping track of the instructor/student relationship in the case of a conflict, and accessibility related concerns.

DISCUSSION
Section 4.2 showed that the meetings for this initial phase have achieved their goal as a pilot program, and are a good initial step towards a more sustainable apprenticeship program within our department for prospective CS teaching-focused faculty.For most of the survey questions, a significant proportion (80% or more) of participants provided positive feedback (Strongly Agree, or Agree) on the meetings overall.Participants believe that the content covered in the meetings was relevant and helpful to them as current GTAs or instructors of record, which we believe will also have an impact on their future career after graduation as prospective teaching-focused faculty members.The meetings provided a good opportunity for the participants to connect with their colleagues and experienced faculty members.This connection is essential in building capacity within our institution that will engage in future activities towards a well-defined and more sustainable apprenticeship program.
Looking into individual meeting evaluations, most participants provided a positive self-efficacy rating.Most participants (60% or more) feel they are more confident in the knowledge and teaching practices covered in the meetings.It is worth noting that some meetings received lower ratings than others.For example, the meetings for student motivation, and diversity, inclusion, and equity received positive ratings from 85% and 90% of the participants, respectively.While meetings about handling students' conflicts, TA management, and course design received positive ratings of 60%, 65%, and 70%, respectively.While there might be multiple reasons for this rating difference, based on the feedback we received from the open-ended survey questions in Section 4.2, we believe the main reasons are: 1: Short meeting duration: All meetings started at 11:30am and ended at 1:00pm.Half an hour was dedicated for lunch and chatting about the meeting topic, leaving only one hour for the session itself.While this was sufficient for some meetings, for other meetings it was not.For example, participants were highly engaged in the handling students' conflicts meeting.They shared their own cases and asked for feedback from the invited faculty and other participants.However, some participants did not have the chance to talk and share their cases.A similar problem occurred for the course design meeting.The invited faculty introduced topics like ILOs, Bloom's taxonomy, and assessment design with some real examples.Participants did not find enough time to work on the in-class activity at the end of the meeting to apply what they have learned.2: Not following a clear meeting agenda: When we invited the guest speakers to lead the meetings, we asked them to make the sessions as active and as possible, and allow for in-class discussions and activities.For some of the meetings, this contributed to the absence of a clear meeting plan or agenda.This was clear for the TA management meeting.The invited faculty asked the participants if they have questions related to the topic, and they started asking.However, some participants asked off topic questions, and this caused some confusion for some of the participants as was stated in Section 4.2.
Here we describe what we learned from this initial phase based on participants' feedback presented in Section 4.2.We present what we think has worked well, and things to improve in future program activities.We believe these lessons will be helpful for those CS departments looking to building a similar program for CS graduate students interested in a teaching-focused faculty position.
• Give participants the opportunity to chat informally (over lunch if possible) with each other and the invited faculty on the meeting topic before the invited faculty starts the discussion.This semi-facilitated conversations allow participants and the invited faculty to better know each other.Furthermore, sharing experiences related to the meeting topic helps the participants better prepare for the meeting discussions and prepare questions for the invited faculty.• Incorporate activities to encourage participants to take part in meeting discussions and share their experiences, questions, problems, and challenges with other participants and the invited faculty.• Allow participants to share cases and ask for advice from other participants and the invited faculty.• For each meeting, save some time for a meeting activity (or activities) related to the discussed topic.Allow participants to work on this activity in groups, so that they can interact with each other more.• If possible, invite a different faculty member for each meeting so that participants get to know as many faculty as possible and learn from their different experiences.• Ask the invited faculty to define a clear topic and agenda for the meetings and share it with participants at least one day before the meeting time.• Plan for weekly meetings during the semester, and allow for more meeting time so that more topics can be covered, and all participants can have the opportunity to share their thoughts, questions, and cases.Based on our experience, one session a week for 150 minutes seems to be reasonable.• Organizing the sessions over a meal creates a more intimate environment for participants to share their thoughts.Offering a free meal is helpful since it gives participants one less item to worry about in their day.

CONCLUSIONS AND FUTURE WORK
In this paper, we shared our experience with the first step towards creating a formal training program for CS graduate students interested in pursuing a teaching-focused academic career after graduation.During Spring 2023, we organized six meetings including selected participants who are current GTAs and instructors of record in the CS department at Virginia Tech, and invited faculty members from the same institution to lead the meetings.Each meeting focused on a specific topic related to CS teaching and learning.The invited faculty acted as a session lead and a discussion facilitator.At the end of the semester, we surveyed participants asking them about the value of those meetings to them, ideas for improvement, and perspectives for future directions.Most of the participants rated the meetings positively in terms of content relevance and usefulness, and the opportunity to connect and interact with other participants and invited faculty members.A few participants gave negative feedback about the meetings' duration and the not-so-clear agenda for some of the meetings.
To address this, one future direction could be towards designing a semester-long training program with weekly meetings, which would cover a wide variety of topics related to CS teaching and learning.Having weekly meetings should overcome the problem of short meeting duration, as instead of having a limited time meeting for 90 minutes six times a semester, we could have a weekly meeting for 150 minutes.This program can be further extended into a graduate course, which would allow us to cover more topics than we covered in the initial phase and allow us to have a variety of inclass activities and discussions.Also, by having more meeting time, we believe more participants will be engaged in the discussions by sharing their experiences, ideas, and questions.Furthermore, by designing a course like this, clear ILOs should be defined along with matching assessment.This will make it easier to define a clear agenda for each meeting, which overcomes the second problem.

•
Participants' overall evaluation of the meetings.• To what extent the meetings provided a way for participants to know and interact with colleagues.• To what extent participants learned new things from the invited faculty.• To what extent participant's ideas and opinions were heard and valued.• To what extent the content covered was relevant and helpful.• To what extent each meeting makes participants more confident in their knowledge of the content covered.• To what extent participants will consider participating in future program activities.• To what extent participants are more confident in their current role as GTAs or instructors of record as a result of participating in the meetings.

Table 1 :
Results from surveying program participants.The rows show the 13 rating questions, and the columns show the ratings on a Likert scale.Cell values show the proportion of participants who provided the corresponding rating.