| ADVANCED SEARCH |
| BROWSE ARCHIVES |
MOST REQUESTED  |
| FUN STUFF |
When BYU professor Michael J. Lambert was preparing for a life in academia, he wouldn't have guessed his future professional duties would include foregoing a private hotel room to share the floor with wall-to-wall students. That scenario is Lambert's fate several times a year when he takes students along to conferences. There, they get the opportunity to network, learn about cutting edge research, and present their own papers. And Lambert doesn't mind any personal inconvenience.
"It's fun. It's a good time. It's the best part of being a professor," said Lambert, one of the psychology department's most prolific researchers, and who is well-known for involving large numbers of students in his projects.
Lambert and many professors throughout the university are eschewing the aging dichotomy that pits teaching against research.
Instead they are merging these two essential elements of academic life, creating a synergy that benefits both professor and student. Teaching via research is the paradigm these professors have adopted. They number in the hundreds and can be found throughout the university, though only a few could be interviewed for this article--Lambert and his psychology colleague Dr. Gary Burlingame, Dr. Tony R. Martinez in computer science, and Dr. Martha M. Peacock in art history.
Burlingame oversees two research groups totaling 30 to 50 students during any given semester. Martinez guides a large portion of computer science's 10 to 15 PhD candidates through his laboratory. And Peacock recently shepherded a dozen students through the preparation and mounting of a Rembrandt exhibit at the BYU Museum of Art.
For these professors, collaborating with students yields more and better research. The students, in turn, gain the opportunity to apply textbook principles to real-life problems, rub shoulders with their professors' colleagues around the country, and, in many cases, publish career-launching papers.
Lambert and Burlingame's latest studies are distinctly in the applied domain: assessing psychotherapy. About eight years ago, they and a handful of students began investigating how to create an instrument that would measure the progress of adults undergoing psychotherapy and the ultimate outcome. The result is the 45-item "Outcome Questionnaire," already in wide use and nicknamed the "OQ-45."
Burlingame, Lambert, and Dr. M. Gawain Wells then decided to devise a similar measure for children and adolescents--the Youth Outcome Questionnaire, or Y-OQ, nicknamed "Yoke." The team then went a step further, developing a mechanism for measuring risk factors that are indicative of whether therapy will be difficult with a given child. In April, Burlingame, Wells, and nearly a dozen students presented data at a Seattle conference showing that the higher the score on this measure, called the Prognostic Assessment (PA), the slower the change process.
"If children have low risk, they'll get better faster," said Burlingame. "If they have high risk, they'll get better slower, but they'll still get better, in most cases."
Several major insurers have adopted the OQ and/or Y-OQ to help them determine how many psychotherapy sessions are reasonable given various symptoms. Where before these companies made such decisions based mostly on the bottom-line, Lambert and Burlingame's research is helping them ascertain treatment dosage based on empirical findings. And increasingly insurers are finding they can't cut off severely ill patients after just a few sessions without having to confront the research done at BYU.
"This is affecting tens of millions of covered lives," said Burlingame. "It's no longer one person's opinion against another's. It's, 'Here's what the data says.'"
With their research showing such high relevance and immediate impact, Lambert and Burlingame don't have to recruit students. The word is out about their projects, and students come to them.
"Most of our research is not very theoretical and not very abstract," said Lambert. "It's applied clinical work that should have an impact on clinical practice and patient care. We want to improve patient outcomes, and our research is all directed toward that."
Burlingame's Y-OQ research group usually numbers about 30 students, both graduate and undergraduate, and even a few from other universities. Most students, particularly undergraduates, pay their dues by first doing simple data entry. As their skills increase, they are given a piece of the research pie to call their own. Within a short time, they frequently have a conference paper or a published article. This year honors undergraduate Miriam Carey completed a study that yielded Y-OQ norms for 4- and 5-year-olds.
"She's tickled to death because she participated in a real-life research study, and we're tickled to death because she gave us something that people around the country are already using," said Burlingame.
In the social sciences, enticing students to research can be particularly difficult because so many are intent on clinical work. They want to learn treatment methods, not statistical analysis and study design.
"Everybody thinks research is horrible. Who wants to go take a research class if you're a social science person?" said Burlingame. "But if they're dealing with a real-life problem--such as the little-known effects of psychotherapy with children--then they see these techniques create a product to help people. They can see the benevolent effects of research. And I'm benignly introducing a learning experience."
Burlingame teaches classroom statistics and research methods, but he finds that students in his research groups often learn as much or more than those who actually take the courses. "They learn what I'm trying to teach students out of a textbook, except they'll remember it because they did it. They'll keep it in their heads."
Burlingame's second research group has included up to six or seven doctoral and 10 undergraduate students. Their focus is group therapy, Burlingame's prime research arena for two decades. Research from this group recently generated the surprising conclusion that group therapy is as effective as individual therapy with virtually every type of mental health problem.
Because of these large research groups and several others, psychology student publications soared during the last few years. Two doctoral students, Todd Corelli and Sterling Johnson, recently won national American Psychological Association awards for their dissertations. And while hundreds of clinical psychology students around the country are not getting internships (required to complete their PhDs), BYU students typically get their first or second choice, at least in part because of their research experience.
One of those students is Nick Taylor, who aims to become a group therapist. With Burlingame, Taylor is developing a national survey to find out why group therapy is so rarely used despite its demonstrated effectiveness. He completed his clinical psychology coursework last year at BYU and landed his first-choice internship slot with a community mental health agency in Columbus, Ind., whose primary treatment format is group therapy.
"I feel I did my best learning through research," said Taylor. "Group therapy is so complex that becoming a good group therapist takes years and years of experience. To research group therapy you must become an astute observer so you can identify what's happening in a group that's working and how it's working. That teaches you to be a better practitioner. Participating in research had as much impact on my skills as class training."
Last year, Burlingame co-authored 27 publications with students, a figure he jokingly said would be "sick" if it weren't for the fact that the labor was distributed among some 50-odd people. And besides, he said, working with students is fun, especially during weekly research group meetings, where there's camaraderie and laughter--as well as pizza and donuts.
"I'm just pumped about this," says Burlingame. "It's play. I get to turn people on to research, I get to do it myself, and I get paid!"
Over in the Harris Fine Arts Center, visual arts professor Martha M. Peacock led about a dozen students during the past year in a two-semester effort to research and mount a Rembrandt exhibit at the Museum of Fine Arts. She envisioned the exhibit building on the work of seven Dutch art historians who have been scrutinizing every presumed Rembrandt work in the world and pronouncing each a Rembrandt or not a Rembrandt, a process still incomplete after nearly 30 years. The Dutch project has provoked raging controversy as various experts debate the historians' conclusions and as new, more precise scientific methods arise for analyzing art that is hundreds of years old. In 199596, the Metropolitan Museum of Art in New York City presented an exhibit entitled "Rembrandt, Not Rembrandt," placing side by side works declared Rembrandts and those "de-attributed" to Rembrandt by the Dutch historians. The point of the exhibit, said Peacock, was to ratchet Rembrandt further up the genius scale and expose his imitators as clear inferiors.
Peacock believed the Met's approach was unnecessarily cut and dried and left unchallenged the conventional idea that none of Rembrandt's contemporaries even approached his genius. To devise a more balanced exhibit, Peacock turned to BYU's own collection of Rembrandt etchings (or supposed Rembrandts) and offered them to her students as a singular lesson in art history.
Peacock began the project during fall semester 1996 with a seminar that confronted students firsthand with the problem of discerning attribution. They reviewed the scholarship on Rembrandt and studied how paper, ink, and printing methods changed over the centuries. They were able to hold in their own hands works of art as old as 350 years and take measurements, examine techniques, scrutinize for watermarks, and even analyze ink under a microscope.
Charlotte Poulton, who graduated last year in art history, found it "thrilling" to work with these masterpieces, especially when an uncatalogued work was found in the library and determined to be an original Rembrandt.
"To actually see a work that Rembrandt himself would have printed was very exciting," said Poulton.
In the end, the group found they had some pieces they knew were Rembrandts, some they knew were collaborations, and some whose origins remain up in the air, said Peacock. With access to evidence that several of the artists who contributed to Rembrandt's works did indeed approach his brilliance, if not equal it, the group was able to focus the exhibit.
"What's more significant than looking for the lone, isolated, singular genius is to understand that art during Rembrandt's time was not seen the same way it is today--art was collaboration," said Peacock.
Once the students and Peacock decided as a group which prints to include in the exhibit, they began the laborious process of writing an exhibition catalog--the book that describes each piece of work in an exhibit. Every student was credited in the catalog, giving each a valuable first publication.
During winter semester 1997, Peacock led the effort to mount the exhibit, which they named "Seeking Rembrandt" (held Feb. 20June 14, 1997). Every decision was made jointly with the students and museum staff, from choosing wall paint colors (pale blue) to designing and writing the labels to accompany each piece. Once the exhibit opened, students became docents, taught seminars, and talked to reporters.
"We had the opportunity to see how an exhibit comes together from the ground up," said Jennilyn T. Smith, a master's student in art history from Decatur, Ill. "Having to do research and discuss our findings [in the seminar] brought us up to a higher academic level. If you threw out an idea that couldn't really be supported, Martha would call you on it."
After the exhibit was up, Peacock didn't yet let her students rest. She had them critique the exhibition and articulate what they would do differently with their deepened knowledge.
In the hard sciences, associate professor Tony R. Martinez has been helping doctoral students become top-notch computer science researchers for a decade. Martinez's specialty is neural networks, a subset of artificial intelligence that seeks to simulate with computers the learning process of the human brain. About half of the department's PhD candidates work their entire four to six years in Martinez's Neural Networks and Machine Learning Laboratory. By the time they graduate, they know their field intimately because they have done intensive research themselves, not just read about it in textbooks or journal articles.
Typically students in Martinez's lab amass six to 10 published articles and are exceptionally well-prepared for a career in either academia or industry.
Martinez encourages his students to take risks in their research, not just "put a little mound" on what others have already done. One student, Dan Ventura, took Martinez's advice further than most. He initiated a completely new area of research and is now positioned to be a leader in the cutting-edge field of quantum computational principles. Ventura said it was Martinez's belief in him that gave him the confidence to take such an aggressive tack.
"Tony's attitude right from the get-go is we can do work that's good enough to be published just like anyone else," said Ventura.
Randy Wilson, who recently finished his PhD, also set his sights, high while a member of Martinez's lab team. Wilson seized the challenge of trying to outperform the best instance-based learning models in the world by analyzing their weaknesses. In instance-based learning, computer scientists take numerous examples of a certain type of problem, such as medical diagnosis, and download the examples, or instances, into a computer. They add computer processing, then introduce a new example (set of symptoms), aiming for the computer to make the proper generalization (diagnosis) on its own.
Wilson attacked the problems in existing instance-based learning models, took the best solutions he could come up with, and combined them into a new system.
"His final system, when tested over about 60 real-world applications, was able to outperform the other 16 top machine-learning algorithms by a reasonable amount," said Martinez. "That's an impressive piece of work."
Martinez works closely with his students, but he also stands back far enough to let them learn on their own. "Many times I will give them some basic ideas to run with, but if you're too explicit they never break through that threshold of learning how to do research themselves. Then when they're on their own, they still won't have the confidence necessary to come up with their own ideas. They'll always have that insecurity. Once they've done it, they have confidence, and then it's amazing how creative the mind can be."
Martinez believes faculty should involve students in research as much as possible. "There are a number of reasons to do research. One is to expand the knowledge that we have in the world. Another is to publish and thus give the university notoriety. I believe that both of those things, if not done with students, are a waste of time at BYU because teaching's not involved."
Martinez goes so far as to argue that nearly every paper a faculty member writes ought to be co-authored with a student. He allows that many of his colleagues will disagree, but his passion on the topic is boundless. "I believe very strongly that if we're publishing without students then it's frankly of little or at least of minor benefit overall to the university mission."
Martinez doesn't teach only computer science as he guides his students through research. He also takes the occasional opportunity to teach about spiritual matters.
"Even in a field as technical as computer science there are often times when it's right to add a spiritual dimension, especially when you're talking about intelligence and making machines intelligent," said Martinez. "I believe you're never doing your very best research until the Lord's helping you out in it. You need to be humble and reliant and realize that you can be inspired in everything you put your hand to, which includes your research."
Communicating these values is one of the facets of working so closely with students that Martinez enjoys most.
"When you're working shoulder to shoulder with students, you have a special opportunity to let them understand where you're coming from and what's important to you. That can have a real effect on them that they may not get in any other way. And to me that's perhaps the most important thing we do here."
That's not to say that collaborating with students is always easy. Burlingame, Lambert, and Martinez all remarked that regularly having to train a new crop of students is labor-intensive and draining. Helping students prepare their first paper for publication can be a nightmare, with long tutoring sessions and endless rewrites. On occasion, Lambert realizes a student has joined his research group as a perfunctory exercise for their degree, and his or her lack of passion is "a drag" on the project.
For Peacock, collaborating with students is a professional sacrifice, however personally rewarding it may be. Academic publishing in her field has a deeply entrenched tradition of single authors, and the quality of a paper's writing is valued as highly as any new information or insight.
"Doing collaborative projects with students in the humanities is always a bit suspect," said Peacock. "Rarely do you ever find papers co-authored, let alone with students."
On balance, though, students' contributions to each professors' professional life is worth the drawbacks.
"I'm having a blast with these students," said Lambert. "They're eager, they're enthusiastic, they're young and bright and impressionable and good intentioned. I don't see how people can do research without them."
