Mentoring undergraduate research: Creating productive experiences & sharing student work

Talk for the ASA Mentoring Undergraduate Research Panel on February 25, 2022 on creating productive undergraduate research experiences and sharing student work.

Click here to view the slides.

Introduction

Hello! My name is Maria Tackett, and I’m an Assistant Professor of the Practice at Duke University. I’m excited to be on the panel and talk a little about creating productive experiences and sharing student work.

Before sharing some tips on these topics, I wanted to give a brief background on the type of undergraduate experiences I’ve mentored.

• Independent research. These are generally projects with individual student that place during a semester or academic year. The students register for an independent study course, so they receive course credit and a grade at the end of each semester. In these experiences the topic and direction are largely driven by the student. Students have an idea of the research questions they want to pursue, and I mentor them through the research process.

• Duke ShinyEd. This project focused on the development of interactive web apps covering various statistics concepts. The project was modeled after the BOAST program led by Dennis Pearl at Penn State. There were two iterations of this project, a six week experience with four students in Summer 2020 and a semester-long experience with three students in Spring 2021. Given the pace of the work in summer 2020, the team also included a project manager who had recently graduated from the statistics major.

• Mental Health & Justice System in Durham County. This is ongoing research with Nicole Schramm-Sapyta from Duke’s Institute for Brain Sciences exploring the intersection between severe mental illness, health care utilization, and interactions with the criminal justice system in Durham, NC. This team functions more like a “traditional” science lab; it is a larger team of students (7 this academic year) and a graduate student who serves as a project manager. Students commit to this project for one calendar year. They receive a stipend for a 10-week summer experience, then receive course credit to participate in the project during the academic year. Students have the option to continue the next academic year (they are only required to participate in the summer experience their first year), so 3 students on the current team has been working on the project for two years.

I hope these help give an idea of the variety of project topics and program structures, while also broaden the type of projects that come to mind when we think of these undergraduate experiences. These experiences can include traditional statistics or applied research, but they can also include projects where students develop something such as an interactive web app.

Creating a productive experience

I’ll share a few tips based on what I’ve learned about creating productive experiences. Given most of my mentoring experience has been during the pandemic, these tips come from what I’ve learned mentoring virtual, hybrid, and in-person experiences.

• Get started before the project starts! There may be some administrative tasks to complete before starting the experience, such as acquiring the data or getting IRB approval. I’ve found these tend to take longer than anticipated, so it’s best to get them done beforehand (as much as is feasible). I’ve also found it helpful to provide students short coding tutorials to complete before the experience starts to ensure all students can perform basic coding tasks. The goal is to prepare so students are ready to start working from the very beginning.

• Collaborate with students to develop long- and short-term goals. For a semester long-project, we identify what we want to deliver by the end of the semester (e.g., an interactive app, statistical model, publication draft, etc.). Then, we identify intermediate milestones to meet every 3 - 4 weeks. Developing these goals helps students see the big picture for the research while also breaking the work into more quickly achievable “wins”. Unlike a course syllabus, it is important to work with students on developing the goals to give them ownership of the work.

• Meet regularly and leave each meeting with clear tasks. I generally meet with students weekly for semester- or year-long projects, and multiple times a week for shorter experiences. During the meetings students share updates, and the last 5 - 10 minutes are used to identify concrete tasks for the students to complete by the next meeting. Periodically, we also check in on the long and short-term goals to assess the progress.

• Encourage students to regularly document and reflect on the research experience. I encourage students to keep a research journal to write down their ideas and thoughts about the project and experience doing research. For shorter experiences, students do a daily checkin where they post their goals for the day at the beginning of the workday, and post a summary of what they did that day and any questions or roadblocks at the end of each day. This was particularly helpful during the Duke ShinyEd in Summer 2020 when students worked remotely in different time zones. For team projects, using a collaborative tool like Google doc for meeting agendas and note-taking is a nice way to document the experience in a place that’s available to all team members.

Creating a productive experience for teams

• Create opportunities for students to work together. Creating opportunities for collaboration encourages students to establish a workflow in which they are more comfortable asking each other questions, troubleshooting problems, and sharing ideas. This takes some of the load off the faculty mentor and helps students develop important collaboration skills.

• Use software for quick team communication between meetings. Tools such as Slack or Microsoft Teams makes it easier for the team to communicate outside of meetings. This platform for communication may be particularly helpful if a student has a question or runs into a roadblock. The more casual nature of these platforms (as opposed to email) makes it easier for students to post their questions quickly rather than waiting until the next team meeting.

• Add a project manager to the team. A project manager reduce some of the load of the faculty mentor while still providing structure and support to the undergraduate researchers. This role also helps the student working as the project manager develop valuable mentoring and collaboration skills. Having a project manager has been important for experiences with larger teams or short timelines, like the 6-week summer experience. Project managers on teams I’ve mentored have held regular “student only” meetings, that just include them and the undergraduate researchers. These meetings have been used to review computing skills and statistical concepts, to troubleshoot problems, and as general working sessions to complete the tasks assigned at the full team meetings. The project manager has also helped with administrative tasks, such as scheduling. I’ve had graduate students work as project managers for experiences I’ve led; however, many parts of the project manager role could be done by an upper-level undergraduate student with previous research experience.

• Build diverse teams. This means diversity in broad terms, both in identity and background but also in terms of the previous statistics and computing experience. More interesting discussion and ideas are generated when students come from a variety of perspectives. Collaborating with others approaching problems in a unique way pushes students to think out the box, which fosters creativity and richer learning experiences.

Sharing student work

Next, I’ll share some ideas on sharing student work.

• Create a public-facing product. This can include a website or interactive web app to make the work easily accessible to others. Blog posts are also a nice way for students to share their work and reflections on the research experience.

• Share updates with stakeholders through written reports and presentations. At this stage students are often doing applied work, so there are likely subject matter experts who stakeholders in the project. These updates are a nice way to keep the stakeholders engaged in the work and give students practice communicating statistical results to non-statistical business and research partners.

• Share updates with stakeholders through written reports and presentations. At this stage students are often doing applied work, so there are likely subject matter experts who stakeholders in the project. These updates are a nice way to keep the stakeholders engaged in the work and give students practice communicating statistical results to a general audience.

• Co-author publications. Work with students to publish their work in applied journals or appropriate statistics journals.

• Submit work to undergraduate research competitions. The Undergraduate Statistics Project Competition sponsored by CAUSE and the ASA is a competition for undergraduate class projects and research in statistics. There are two rounds per year - one in June for spring semester and year-long projects and one in December for projects completed during the summer and fall semester.

• Present work at conferences. Some conferences, such as the Electronic Undergraduate Statistics Research Conference and StatFest, are specifically for undergraduate students. These conferences also generally have career and graduate school panels, and are relatively low cost to make them more accessible to a lot of students. There are other conferences, such as the Symposium on Statistics and Data Science and Women in Statistics and Data Science , that have poster and lightning sessions that can be a nice venue for student work.

I hope some of these tips are helpful for those interested in mentoring independent undergraduate research and research teams. Thank you to Donna and Kelly for the invitation to be on this panel!