For a couple days earlier this week (Sunday night, Monday, and Tuesday morning), I attended the first Research Experience for Teachers (RET) Conference. The focus of the conference was on “leveraging our collective impact” across what has been done in RET programs over the years, going forward. I found the conference interesting and insightful, far more so than expected. I took away lessons on how I should view my role in academia and a few ways in which I can do my job more effectively as well as ideas on how RET programs could be improved and how they fit into education in general.

A bit of background…

The RET programs are funded by the National Science Foundation (NSF) to give K-12 teachers and community college lecturers an authentic research experience. During the program, teachers spend six weeks in a university setting of some sort expanding their skill set and understanding of engineering. They also work to develop a module (generally a small set of lessons) which they can take back to their classrooms. Generally, the teachers are pulled from Math and Science, but it isn’t clear that is required. Broadly, the goal of the program is to improve science, engineering, and math education. Specifically, in the context of my experience, I will be discussing the RET programs which are run out of the Directorate of Engineering at the NSF. These programs were the topic of the conference and the type which I was a part of. My understanding is that there are RET programs run out of other segments within NSF as well. There are currently about 40 RET programs being run, each taking about 10 teachers a year. There are two ways in which an RET program can come about. Every Engineering Research Center (ERC) which is funded by the NSF is required to contain an educational component, including an RET program. There are also dedicated RET sites, where an RET is generated as a standalone grant funded package.

For the past two summers and into the school years, I have participated in the GRASP RET Program (NSF #1542301) run out of the General Robotics, Automation, Sensing, and Perception lab (GRASP lab) at the University of Pennsylvania. As part of the program I have mentored three middle school teachers. My first summer I mentored a pre-service middle school science teacher. My second summer I mentored a middle school math teacher and a middle school science teacher.

The GRASP RET program runs on the premise of giving the teachers whom it brings in a challenging and authentic research experience in an engineering lab. Teachers are matched with PhD students who are affiliated with the GRASP lab and act as guides throughout the research process. Teachers begin by doing a search of the literature within the general space of their mentor’s work, using the same tools and techniques used by academic researchers, starting with general review papers and digging deeper into narrower technical papers. This allows the teachers to understand the research space in which they are working and begin to find topics that interest them. As they continue, they are asked to identify a gap in the literature, a question which has been left unanswered in prior work. This then leads them into the process of clarifying a research question and hypothesis around that question. To help test the hypothesis, the teachers develop a methodology for testing, fabricate the systems and produce the hardware and code necessary to carry out their methodology, and begin testing. They then compile a research paper containing their finding from the literature, approach, results, and a conclusion is the same spirit as any academic research paper, make a poster, and give a 10-minute podium talk to everyone involved in the RET and local education leaders.

The process is brutal for the teachers. When they begin, few can understand much of the language or concepts presented in the literature which they are reading or understand how to use the engineering tools needed to complete their projects. Most teachers break down at some point during the process. As a mentor, it has been my role to guide them in the right direction, encourage them along, and when they are about to give up, catch them and move them forward.

Through the process they gain significantly more content knowledge than they entered with. However, I would claim that is auxiliary to the main benefit to their students. Because of the rigor, they develop an empathy for their students who struggle with seemingly simple tasks. Many of the teachers haven’t struggled academically in a long time, if ever. This struggle opens their eyes to what it feels like to be totally lost. In a different thread, at the end of the program, when the teachers look back, they find they have learned more than they thought possible and realize that the reason for that is that they struggled through a complete experience. I think that some teachers take that as a pedagogical lesson to suggest that learning through experience can be a powerful tool. And finally, the teachers get to interact with a variety of engineers throughout the program. That allows them to go back to the classroom with a greater appreciation of what engineering is and how engineers approach problems. My hope is that that knowledge is transferred to their students so that their students can have engineering as a potential goal. Everyone knows what a doctor or lawyer is. It seems that only children of engineers know what an engineer is. If this does in fact happen, I would expect a slight nudge to get more under-represented minorities (URMs), who may not have as many engineers in their lives, into engineering.

To help drive home some of the learnings taken from RET, the mentors visit the classrooms of the teachers three times over the course of the year following the program. This hopefully provides a chance for students to see some type of engineer and to do some interesting activities. Beyond whatever special value I may bring, I have found that just having a second person in the room can help reach otherwise unreachable students (I could easily write on the absurdity of leaving one teacher in a classroom of 30+ students).

The first summer I did the RET, I presented a set of well-defined and scoped projects to my teacher, like what might be received in an industrial/professional setting. I then tried to help the teacher through every step of the process, ensuring that she was able to succeed. The process was hard on her, but not as hard as it could have been. I also tried to involve myself in her curriculum development process. The amount that I took on with her led me to spend 15-25 hours a week on the program, which was not sustainable. At the end of the summer, she seemed to have had a great experience, she has remained engaged with our lab, but I’m not sure that she took away as much as she could have.

The second year, I took on two teachers, you can read more about them in this Penn Engineering profile. I was much more hands off. Instead of constant interaction I tried to deliver just in time critical help in a focused way. For the first five weeks this worked quite well. The final week I found it necessary to be much more involved to help drive the teachers through final implementation and review their writing and presentations. I think at the end they took away more than my first teacher although at the cost of not being as close personally to me. This approach was also a much more sustainable system for me.

Of course, I have no data to take away from the two years I was involved to say which approach was better and the personalities of the teachers varied widely. On whole, I think that allowing the teachers to struggle is critical to driving classroom change. And allowing them to direct their own research allows a significant amount of learning and provides motivation to make it through the challenging rigor of the program.

The conference

For the first year, an NSF sponsored conference brought together PIs, education directors, graduate student mentors, education assessment experts, funding/advocacy organization representatives, and RET teacher participants. The conference was hosted by North Carolina State University. I had no clue what to expect for the conference and figured that maybe 45 people would show up. I found the conference to be highly productive with 100 people showing up and participating in interactive collaborative sessions. I was able to learn about a wide variety of programs, hear the experiences of those involved, and discuss where RET as a concept should be heading.

Through my discussions with people of other programs, I found that our program is unique. I think that the people who created the GRASP RET program deserve a lot of credit for coming up with a model that I believe is the best I have seen so far. But there is a lot to improve, and we found that those things that need improvement in our program have been challenges for many others as well.

What are RETs trying to achieve?

I think the biggest open question around RET is: what are the goals? When asking people what their programs achieve, the responses were all over the place. Some programs focus on teaching concepts directly that can be taken back to the classroom. Others focus on curating experiences that can lead to classroom activities. Few seem to step back as far as us and focus on the way that teachers think about education. At the end of the day we all want to improve student outcomes, but even that is nebulous. No one wants to put the focus on student testing. And I don’t even want to focus on sending students to college. Generally, I don’t think that K-12 is the time to get into hard engineering, although many engineering skills can be applied across the spectrum of subjects, so I don’t think we should measure the number of engineering-based lessons either. The best answer that I can come up with is to say that we should try to 1) increase the number of students who are familiar with the field of engineering 2) improve student perceptions that their teachers understand them when struggling 3) improve student perceptions that their teachers are actual experts and 4) improve students abilities to solve novel open-ended problems.

There were significant discussions on how to assess the outcomes of RET programs. It is sufficient to say that the problem is hard. I do think that evidence-based teaching is valuable, and this is a space that needs more exploration broadly, beyond RET.

The education ecosystem

One of the first activities which we worked on was trying to understand what the education ecosystem looks like. There were a multitude of different takes on the problem that highlighted a few things that carried through the rest of the conference, I think the most important of which was that we need to keep students as the center of our goals. Everything we are doing in this space should be trying to improve their state. I also had a few takeaways beyond that. The RET program is designed to bring teachers into research and engineering, but more is needed pre-service. Schools of education are currently too isolated from the places in the university which focus on the material that teachers teach. It was also clear that we need more industry involvement in these efforts. The RETs as they stand now primarily give exposure to academic engineers with slight industry mentoring. There is a lot of value to that, in my experience the teachers have an easier time seeing their students as grad students. The jump from their kids to the grownup kids that we are isn’t too far. But to fully understand what engineering is, we need to acknowledge that most engineers do not reside in academia. So, we need to expose the teachers to engineers in industry. There were a couple of programs that are using industry cost share to expand on their NSF funding and who are doing site tours at industrial facilities. I think both of those are fantastic and we should all try to follow their lead. I have encouraged our program to look at having an industry day as a part of the program where we bring in people from industry to give quick 10-15 minute talks to the teachers on what they do. I think there is an opportunity to film those talks to give teachers a repository of job descriptions to provide to their students. At the least this would allow teachers to have a broader sense of what engineers do. Finally, I think that we need to understand that often a student’s performance is driven by the community around them, at the national, state, local, neighborhood, friend, and familial levels. When the community around a student promotes the value of learning, students put forth an increased effort and have the resources outside of the classroom to excel. This suggests that we need to work more to engage communities.

Challenges in running an RET

There are many challenges in running an RET which we seemed to conclude could be partly alleviated by sharing resources between centers. From the beginning when a team starts to design a program, they are largely on their own for making design decisions. The request for proposal for the program has a significant number of suggestions for how a program should be run and some programs publish a few details of their approach in research papers, but these only scratch the surface of what is needed. To try to help new programs start, I believe that it would be beneficial to have a database with fully detailed goals and methods of RET programs. And as the programs progress, materials and measured outcomes. There was broad agreement that some sort of RET conference should be occurring regularly and that a condition of receiving grant money should be sending a small team to the conference for one of the three funded years.

Broadening participation

A major focus at the conference was on broadening participation which in simpler terms has the goal of brining more under-represented minorities (Hispanics, people of color, low income students, first generation students, etc.) and women into engineering. I think this is critical for several reasons (that will have to be another post at some point). During a panel discussion on promoting diversity, I asked what I should be doing as a white male when I visit K-12 classes to enable students who look differently than me to see themselves in me and so be able to see who I am as a possibility for their futures. The first suggestion was to try to bring people I work with who represent a diversity of backgrounds into the classroom to show the different ways that engineers look. This is something that I have for some reason not considered, but I think is a great idea. I suspect that if a person of color or woman accompanies me to see these classrooms, that is more impactful than if two people of color or two women went in. Showing diversity in a field sends a message that it is accepting of everyone by everyone and I think that is powerful. It was also pointed out that people who are minorities in an area will often “count” when viewing media. So, if I have a presentation where there are many researchers’ hands shown and they are all white, people will notice that. This is just the sort of thing I don’t notice but will have to keep an eye out for. It also makes me think that I should do more in my lab presentations to highlight the people in our lab. My PI has done a great job bringing in a diverse lab, and we should do more to highlight not just the research, but also the people.

There was also a lot of discussion on what needs to be done at the university level to raise, hire, and support under-represented groups. I think we need to start by admitting that the pipeline to provide candidates is not good enough yet, we need to have more diverse candidate pools at every level. It is also important that the community around and within the university is welcoming to everyone. It was pointed out that a potential employer that goes out of their way to show how a candidate fits into the community will be more successful. There was also some discussion on how many of the support structures for URMs have turned out to be good for all students. We need to remember that an entire cohort of the best people who think the same way are often not as productive/valuable as a cohort of slightly less experienced people who bring new ways to think. We should be willing to take risks on the people who haven’t yet had a chance to prove themselves but have interesting perspectives.

As I have already said, I think that part of broadening participation in engineering will come from broadening familiarity with engineering. I think the RET system is in a good position to make a lot of marketing material for exposing students and teachers to engineering. It would take minimal additional resources to make videos, flyers, etc. that tell the stories of grad students, professors, industry mentors, and teachers involved in RET programs. A database of 2-minute videos on the different stories of participants would allow students across the country to get more exposure and easily learn about more paths in their education. They could find the people with whom they empathize and see themselves as an engineer.

Scaling up the RET program

There was a lot of discussion on how to increase the reach of the RET program. Some suggestions were floated to make shorter sessions, online sessions, disseminate curriculum, create training programs from teachers who participate, etc. I think that these ideas generally miss the point of what the RET does. I have mentioned that I believe the rigor of research, progression through real novel research, and exposure to grad students and lab personnel is what makes the program. Without that, we are just doing another professional development session. One of the limits to scaling up the RET in the way that it looks today is the availability of quality PhD students to serve as mentors. In 2018, the NSF made 2,000 graduate research fellowship program award offers. I see no reason why the NSF should not directly tap its GRFP pool to provide mentors.

What is our role in academia?

This goes more generally into a question of what our role is in academia. I heard time and time again that we sit on three pillars: research, teaching, and service. And many made clear that working in K-12 education is a critical part of the teaching. I even heard a dean of an engineering program say that she does not promote faculty who don’t prioritize teaching down the educational pipeline. But I look around the place where I work at Penn, and that is not what I hear or see. I see grad students who are pushed to do research. I would guess that about a third of us are passionate about working to educate those at lower educational stages, but we don’t get rewarded for that. I constantly hear PhD students who say that their PI expects too much research work to allow them to focus at all on teaching or service. Of course, there is still significant outreach in the form of science fairs, tours, summer camps, etc. sponsored by the university. We even run multiple RET and research experience for undergrad programs. But at the personal grad student level, it seems like only a portion of the students carrying the load and an outsize focus on activities that fall on the outreach side instead of the education side of a narrow line.

I am lucky enough to have a PI who prioritizes both teaching and service. I don’t see her being rewarded for that. When time and resources go into bringing in students from community colleges and high schools to increase exposure, instead of focusing on perfecting a funding application; there is no reward, there is no grant. When I take on seven students over a summer and focus on teaching them, helping them to see what research is, I don’t get publications out, I don’t make progress towards my thesis. At the beginning of this semester, I had begun to lose faith in the importance of those things. If the system tells you to focus on research to the exclusion of all else, it is easy to give in. Hearing so many academics at the conference who see importance in K-12 education, has helped me to reflect further on their importance.

I looked at some of the codes that more or less represent the laws governing the NSF and found that the educational mandate is clearly there. For example, 42 U.S.C. § 1862n instructs the NSF to run programs whereby higher ed partners with other groups to “improve elementary and secondary mathematics and science instruction” and 42 U.S.C. § 1862s–5 talks about the NSF’s responsibility to expand the STEM professional pipeline and especially to exploit the untapped talent in URM groups. The request for proposals for NSF engineering research centers includes a section on workforce development that says a goal is “To broaden pathways to engineering for underrepresented students”. Yet, I don’t hear about this focus at my level.

I believe it is the responsibility of the NSF to move the community in a more teaching and service oriented direction. If the NSF wants academics more involved in teaching, make it a stronger part of our grants. Don’t let us get away with nebulous promises to provide mentorship and do impactful research. When students receive GRFP awards, make them work for them. Make it clear to grad students that when we go for our young career awards or our first R01, if we haven’t shown a commitment to teaching and service, we won’t get those awards. When funding is tied to service and teaching, the universities will follow.

Proving benefit to mentors

I think that beyond telling mentors why this important to society and part of their role in academia, it is also valuable to show how it helps them in the shorter term. I don’t think it is well studied or shared, but I heard anecdotally about how mentors improve their communication skills about their research. That they learn to articulate concepts to non-experts in a way they could not before. I have found that my teaching has improved and some of my perspectives on it have evolved; others echoed this. I have also gotten much more exposure to the K-12 education system, which has allowed me to start to refine my thoughts on how education should operate in this country.

The curriculum development component of RET

Loose the curriculum development portion… From talking to a lot of people, I concluded that it is rare that the curriculum module can be implemented as hoped. It seems to take a lot of resources that could go to scaling the RET program, and as I have said, I don’t think it is a critical component. Some teachers did say having access to experts in curriculum development with tons of existing curriculum resources was valuable. But I think that belongs in a different program.

What do teachers/students get out of this

I heard from a number of teachers who talked about what they and their students have gotten from the program. Their stories were powerful, the change they claimed profound. Teachers who taught only one subject being comfortable with any new material. Teachers who moved from textbook based learning to student led learning. And more. We have seen similar outcomes in our program. I think it is important to communicate these things to mentors, teachers, policy makers, and the public, to highlight the importance of the program.


In closing, I have enjoyed my experiences with RET and hope that my work has had a real impact. There is a lot of work to be done to improve the education ecosystem/pipeline and I hope that I can be a part of that. To enable that, I think that academia needs to remember to focus on education and I believe that large funding organizations can make that happen. I thank the people at NSF for overseeing these programs, the people in GRASP for running the program I was a part of, NC State for hosting the conference, and the American public for their continued support of science and math education and research.