Scaling up research opportunities in introductory science courses requires a new way of thinking and working.

Holli Duhon arrived at the University of Texas at Austin as an honors science student—who didn’t really understand the process of science. “In high school, science was straightforward,” she says. “The Nobel Prize was yours to claim if you just followed logical steps.”

In a traditional program, Duhon could have maintained that assumption for years as she worked her way through textbooks and lectures. But instead, she—along with more than 500 other first-year students—joined the three-semester Freshman Research Initiative (FRI), a large-scale program started in 2005 that teaches through experimental research.

After spending her first semester learning basic research methods, she chose a project from 20-plus research “streams,” ranging from biofuels to nanomaterials. Duhon decided on the nucleic acids aptamer stream aimed at examining the interactions of certain nucleic acids with an eye toward drug development. She spent two semesters searching for short DNA sequences called oligonucleotides that could bind a target protein from the bacterium Burkholderia pseudomallei, a pathogen that causes an infectious disease common in Southeast Asia called melioidosis.

The search was like looking for a needle in a haystack, but that needle had the potential to be very valuable. A tightly binding oligonucleotide could lead to a diagnostic test for melioidosis. Meanwhile, Duhon’s fellow aptamer stream students looked for oligonucleotide sequences to bind target proteins linked to Parkinson’s disease, diabetes, and Alzheimer’s—all before wrapping up their sophomore year.

Duhon soon realized that real-world research didn’t look much like the clean, streamlined labs she’d experienced in high school. “There are so many times you try out a protocol thinking it will work out brilliantly, only to find out that something fails on the first step. Or the last step. Or anywhere in between,” she says. “I was not aware that science involved such creativity.”

University of Texas at Austin Freshman Research Initiative student Holli Duhon describes her research.

Duhon didn’t find the oligonucleotide magic bullet, but she developed critical thinking skills, tenacity, and an appreciation for the challenges and joys of science in a way that more traditional courses don’t often allow. More important, she wasn’t one of a privileged few having this type of powerful experience; she was one of hundreds.

The University of Texas (UT) is one of many major research universities—along with dozens of smaller colleges—experimenting with classroom-based research opportunities for undergraduates. Sarah Simmons, director of UT’s FRI, which is partially funded by HHMI, acknowledges that there are challenges to upending the traditional “lecture and lab” model for introductory science courses. Moving toward a research-based approach requires creating appropriate projects, for example, as well as staffing labs for longer hours.

Those issues get even trickier as student numbers climb from dozens to hundreds. But compared with a group of similar UT students, FRI students have better graduation rates—67 percent compared with 53 percent—and a higher likelihood of pursuing advanced degrees in science—32 percent compared with 9 percent.

Page 2 of 5

For Simmons, those numbers indicate that making difficult changes will have a meaningful impact. “The status quo doesn’t reach students early enough,” Simmons says. “We need to invent a new paradigm.”

Think Big

There’s a growing drumbeat to increase research opportunities for undergrads, including pressing recommendations in recent reports released by the American Association for the Advancement of Science and the President’s Council of Advisors on Science and Technology (see Perspectives & Opinions, “Engage to Excel”). Some major organizations already have a jump on these goals: the National Science Foundation supports thousands of students each summer through its Research Experiences for Undergraduates program. HHMI also provides funding for some 4,000 students to do life science research each summer and supports efforts to scale up research opportunities in the classroom.

Graham Hatfull, an early promoter of large-scale classroom research, created—and now helps oversee—a project that has grown from a few high school classrooms to 70 colleges and more than 2,000 students. A decade or so ago, Hatfull, an HHMI professor at the University of Pittsburgh, developed a course for high school and undergraduate students to discover, sequence, and annotate the genomes of bacteriophages, viruses that infect bacteria linked to human diseases such as tuberculosis. It was a small course that was perfectly designed to grow. The processes and tools were simple enough even for novice scientists to understand. The vast, unexplored territory gave students ownership of their projects, and the results were often notable enough to warrant publication. Even better, the work provided rich data for Hatfull’s own bacteriophage research.

Hatfull worked with HHMI to tweak the model for the undergraduate classroom and then introduce it into college curricula around the country through the Institute’s Science Education Alliance (SEA). Since it was first introduced to college classes nationwide in 2008, SEA has had major successes, including two research papers in PLoS One and a paper announcing the genomic sequences of 138 bacteriophages. One of the PLoS One papers had nearly 200 student coauthors, and the genome announcement represented the contributions of more than a thousand students. Hatfull is convinced that the model can be applied to a wide range of projects. “[Faculty] who can identify a research-based platform that can be implemented on the freshman level while advancing their research programs will see a great impact,” he says.

At the University of California, Santa Barbara (UCSB), biology professors Joel Rothman and Rolf Christoffersen, along with academic coordinator Douglas Bush, saw an opportunity for students to gain research experience working on a piece of a larger study by Rothman on the roundworm Caenorhabditis elegans. With HHMI funding, they developed a 3-week module as part of a 10-week sophomore biology course. Students learn to knock down certain genes in the worms using RNA interference, perform a chemotaxis assay to learn what odors the worms are attracted to (or repelled by), and then compare their results with other experimental findings. The research is designed to help identify genes involved in the worm’s chemosensory signaling pathway.

After testing the concept with about 50 honor students last year and making some modest changes, they expanded it to a much larger audience: this year, some 800 students will participate in the C. elegans module. To accommodate all 30 sections that meet each week, the school opened two adjacent lab classrooms with three-hour lab sessions running from morning through evening, five days a week. The labs are taught by TAs, with help from two staff members and part-time undergraduate lab assistants. “We still have bugs to work out,” says Rothman. “Nonetheless, we’ve already made several original research discoveries, and we’re really excited about this. The results are something we plan to publish, not just in educational literature but also in the primary scientific literature.”

Not all—or even most—students who are part of a large-scale research project will pursue science careers, but that’s not the point, says David Asai, director of precollege and undergraduate science education at HHMI. “It’s not just about adding scientists,” he says. “We also need a lot more people who understand science—teachers, lawyers, journalists, and parents.”

Simplify and Succeed

A significant stumbling block to creating a research experience in the classroom is finding projects that are small enough and straightforward enough for novices—but that also move a project ahead in a meaningful way. If a research project is a marathon, the collective work of students may move it forward only a single step or simply show researchers which roads are not worth traveling. But these results can be valuable.

Andy Ellington, a UT biochemistry professor who heads the aptamer stream, chiseled away at the larger scope of his aptamer-based research to find small but critical pieces where students could contribute. While each student’s project is unique, the processes are similar enough that students learn the basic procedures together and go off to do research on their own. “We’re not reinventing the wheel,” he says. “In some ways, they can work together as a group and use one another’s successes and failures to hone their technique for their individual ends.”

Page 3 of 5

About 10 percent of the students in each group produce something that Ellington describes as “interesting” and worth pursuing. Those who don’t might team up with students who have had success, and the best—those who are sharp, enthusiastic, and have the “good hands” that are essential for lab work—often end up working in Ellington’s lab a year or two later. Many of them go on to publish papers on their work. But it all starts with a very simple project.

At UCSB, Rothman, Christoffersen, and Bush learned quickly that they couldn’t work with novice students the way they could with postdocs, who are typically much more proficient at the physical manipulations required for research. A procedure that a postdoc could perform on C. elegans in two minutes might take an undergrad just learning the process 20 times longer. That kind of time lag can confound findings.

Instead of trying to get students to do more, Rothman and his colleagues redesigned the experiments so students were less likely to introduce errors—or injure the animals they were studying. “It’s like asking a student who has never been on a bicycle to enter a race,” says Rothman. “We’ve had to build a bicycle that students can stay on without falling over and killing themselves.” Adapting the teaching module required time and creativity, but students were better able to complete their research successfully.


Research Experiences for Undergrads

See an infographic comparing five undergraduate research programs.
View Graphic >>

For classroom-based experiences, researchers and instructors must navigate one of the most difficult parts of the research process—frequent failure. Most traditional labs are designed so that students who do everything correctly will succeed; in individual research projects, a positive outcome is never guaranteed. Instructors have learned to offer students early research experiences that give them the taste of success before they delve into unknown territory.

At the University of California, Los Angeles (UCLA), for example, classes of 20 to 25 students work in the lab of HHMI professor Utpal Banerjee on a range of projects studying the fruit fly, Drosophila melanogaster. In one project, students learn a clever genetic trick called “lineage tracing.” They fluorescently label a group of cells in early Drosophila development and watch to see what tissues those cells eventually become part of.

The lab work is difficult, says Ira Clark, the academic administrator for the UCLA minor in biomedical research, and while failure is common, he, Banerjee, and instructor John Olson did their best to make sure it wasn’t inevitable. Instead of designing projects in which only a tiny portion of Drosophila lines would yield a positive result, they were able to design one in which positive results were more common. “It is the nature of the project that if you do the experiments on 10 random lines, you are likely to get at least one—and in many cases several—positive results," he says. While some students may still see all negative lines, it's quite rare. “We wanted to give every student that discovery moment,” he says.

As students move forward, however, there is no guarantee of success. Instead, they must find different ways of feeling accomplishment, whether it’s creating new approaches to solving a problem or discovering something unexpected in a failed experiment. Duhon realized early on that she might be well-suited for research. “In most lab classes, you pursue an A and put the experience second,” she says. “But in [the research course] the priority was not the grade. We needed to engage in the experiments and learn what it feels like to personally contribute to legitimate academic discoveries. I learned that I was willing to fail 99 times for one successful moment.”

An Infrastructure for Growth

Many large-scale research programs, including Banerjee’s Drosophila projects and Hatfull’s bacteriophage work, are showing significant progress, but translating those successes to other schools—different sizes, different cultures, different goals—is a tall order. Faculty from successful institutions are sharing individual successes and best practices to create a framework that others can use to adapt existing programs and build their own.

Page 4 of 5

Sarah Elgin, an HHMI professor at Washington University in St. Louis who runs the successful Genomics Education Partnership (GEP) that now includes more than 70 schools, has been refining this process for years. The GEP, which focuses on the “dot chromosome” of Drosophila, so-called because of its small size and condensed genetic material, is designed to help students work with large data sets to transform the genome’s raw data into a more polished sequence through universally accepted annotation and finishing standards.

We needed to engage in the experiments and learn what it feels like to personally contribute to legitimate academic discoveries. I learned that I was willing to fail 99 times for one successful moment.

Holli Duhon

As the program grows, Elgin has found ways to share lessons learned to help others get their courses off the ground. She has run one- to five-day workshops, for example, and has set up a website ( where faculty can share curricula and details of their approach. “It’s got entries from different members about their class sizes, how the course was organized, and the hours they scheduled for the research and guiding their students in critical thinking,” she says. “It’s a place for people to look for help when they begin to think about bringing [research] into the curriculum.”

University of Georgia biochemistry professor Erin Dolan is taking an even broader approach. She is heading up a fledgling national network, called CUREnet, aimed at creating and sharing course-based undergraduate research experiences in biology. Started in 2011, the National Science Foundation–funded network is being initiated through a collaboration of about 25 programs across the nation that are already sharing best practices in teaching through course-based research.

Dolan expects to find common ground among programs. “There will be annual meetings, but we’ll also have a website and social networking functions so that people can discuss what they’ve tried in their classrooms or share software that undergraduates might find useful, for example,” she says.

Research and evaluation of these experiences are providing important data for professors to use to further develop and expand their courses. One of the most comprehensive surveys of broad-based classroom research is the Classroom Undergraduate Research Experience survey (CURE; unrelated to CUREnet). Over several years, Grinnell College psychology professor David Lopatto and colleagues have collected thousands of data points about classroom research and how it compares to more in-depth summer programs and traditional courses. Last year alone, 51 institutions participated in the CURE survey.

The surveys showed where otherwise strong programs needed work, says Lopatto. “When we first started doing surveys, one of the lowest-scoring learning gains [overall] was in learning ethical conduct in the field,” he says. “Program directors told us that they hadn’t been formally teaching ethics or the proper conduct of research and that they would start doing so.” Some individual programs found gaps in writing or discussion and changed their programs to strengthen those components.

But perhaps more important, the surveys showcased some of the powerful benefits of a research-based approach. In the self-reported surveys, students participating in classroom-based research experienced, to a somewhat lesser degree, an almost identical list of benefits as those in summer programs. From understanding the scientific process to the ability to analyze data, students in research-based courses tended to come out far ahead of their peers in traditional classes.

Looking Ahead

Opening up the scientific process to large numbers of undergraduates has shown early success. In a variety of measurable ways—from CURE survey reported skill improvements to FRI student graduation and subsequent graduate school enrollment rates—it has helped boost students’ performance while encouraging a larger percentage of students to pursue science classes and careers.

But it’s time to think even bigger. Grants have helped some schools open up a class or two to research, but that might not be enough, says Asai. “We could challenge big schools to take on [research-based courses] not just for one section, but for 40 sections,” he says. “We could bring this approach to big schools that produce lots of science teachers. There are a lot of exciting ways to think about using these courses.”

The courses don’t just shape semesters or college experiences—they can also shape careers. Holli Duhon had planned to be a doctor, but now she’ll add research to her plans. “I am on my way to completing a B.S. in medical laboratory science, and it’s my goal to incorporate research into my career,” she says. “The aptamer stream provided me with the insight to critically evaluate what my next steps should be.”

Page 5 of 5

Roehrig runs one of them, an induction program for middle and high school STEM teachers in Minnesota. Mentors and new teachers in the Teacher Induction Network can hold video chats using Skype-like technology, share lesson plans via Google Docs, or take part in virtual classroom observation. Rookies videotape themselves teaching, post the video, and then peers and mentors use video annotation technology to comment on the teacher’s interactions with the students. “It’s as good as, if not better than, being in a classroom,” Roehrig says. And it makes effective mentoring possible even when the beginning science teachers work in northern Minnesota, hundreds of miles from her university’s Minneapolis campus.

Scaling Up

With budgets tight everywhere, district training programs have to do more with less. Anita O’Neill supervises professional development programs in science, technology, and engineering for Montgomery County Public Schools in Maryland, a district with 200 schools and about 600 teachers at each grade level. In 2006, she and project manager Mary Doran Brown began building a district-wide cadre of teacher leaders to help train their elementary school peers to teach science better—and they are moving it online.

They recruited prospective teacher leaders from 90 of the district’s 131 elementary schools—not all of them, as they had hoped. Then, with HHMI support, they trained those teachers to help colleagues at their respective schools teach inquiry-based science lessons. At some elementary schools, the teacher leaders got their colleagues to take students to annual “inquiry conferences” at a local college. There, the students presented a science project to their peers and fielded questions from them, just as practicing scientists do at a scientific conference. The program lasted four years until the district’s budget tightened in 2010.

To affordably reach the district’s throng of elementary school teachers, O’Neill’s team enlisted its teacher leaders to help move the training online. At a summer workshop, teacher leaders from elementary and middle schools learned to videotape a lesson, edit the video, and then post it as an example of effective teaching. Ultimately, O’Neill’s team wants an interactive website for all K–12 teachers that allows them to review the district’s science, technology, and engineering curriculum and plan lessons or learn inquiry-based teaching in line with national standards. “Our vision is a professional learning community,” O’Neill says.

To change science and math teaching nationwide, though, there’s really no substitute for investment. And no state has invested as much as Alabama. Thanks to an enthusiastic state superintendent and a powerful booster group that included leaders of the state’s high-tech businesses, Alabama has invested up to $46 million per year in the Alabama Math, Science and Technology Initiative (AMSTI), says Steve Ricks, who directs the program at the state’s education department. AMSTI employs 850 teacher trainers to train up to 8,500 K–12 science and math teachers each year, offering them subject-specific, grade-specific mentoring. Since 1999 they’ve trained half the STEM teachers in the state, Ricks says. Teachers come for two-week workshops for two consecutive summers. AMSTI also employs 300 master science or master math teachers who advise and mentor teachers and even co-teach if the mentees need a hand. AMSTI operates 11 regional 35,000-square-foot warehouses, where workers run forklifts to help sort bins of laboratory materials and equipment designated for math and science teachers.

The state’s investment is paying off in better student performance, according to eight years of external evaluations. For example, Alabama students improved more in math than those in all but one other state, as judged by an internationally recognized test called the National Assessment of Educational Progress. “The state has seen that if you really want students to compete, they need top-notch math and science skills,” says Ricks.

Wendy Bramlett, who used AMSTI to raise her game, is a fan. “My whole way of teaching changed,” she says. “I went from a lecture class to no lecture and all hands on,” she says. Her students’ performance has improved—92 percent scored at the top level on the state science exam last year. And she hears something else she never heard in her first years of teaching. “I have children tell me, ‘Science is my favorite subject.’”

  • 1
  • 2
  • 3
  • 4
  • 5