Choose your research methods wisely

For many students, the first hurdle on the road to completing a dissertation-choosing a topic-can seem like the largest.

But even after you've completed that step, much hard work remains. That hard work begins with choosing the right research method to answer the questions you've posed, says Mary Heppner, PhD, who teaches a class on dissertation methods to graduate students at the University of Missouri-Columbia.

"With the quality of students that we get, they can obviously do the work of the dissertation," she says. "But sometimes fears get in the way-and a lot of those fears relate to the methods."

But fear not: In this second article in gradPSYCH's six-part guide to completing your dissertation, psychologists who teach dissertation classes and those who've recently completed their own dissertations offer advice on getting through the methods section. They say:

• Let the literature be your guide. A thorough literature review is the best starting point for choosing your methods, Heppner advises, because evaluating previous researchers' efforts can suggest a path to answer your own research question.

"You might find out that people have used certain designs and that they've worked well or that there have been problems," she explains.

For example, she and many of her students study the effectiveness of men's rape-awareness programs. For years, she says, researchers in the area have used between-groups research designs, comparing the outcomes of groups of men in separate treatment conditions-a strategy that has often worked well. But sometimes evidence that certain types of programs are effective with certain subgroups or individuals was obscured when researchers only looked at the mean differences between groups, she says.

"We realized we need to conduct cluster analyses, looking at different groups of men, to get at subgroup differences in attitude change," she says.

Students who are familiar with all of the literature in their area will be more likely to take these kinds of subtleties into account in their own research, she says.

Following the lead of another researcher eased the dissertation process for Ira Saiger, PhD, a developmental psychologist and recent graduate of the Ferkauf Graduate School of Psychology at Yeshiva University. For his dissertation, he compared the accuracy levels of two tests that screen for developmental delays-the Denver II, which takes about five minutes, and the Bayley Scales of Infant Assessment II, which can take more than an hour. Another researcher had already compared the tests' accuracy with 3- to 6-year-olds, so Saiger examined their accuracy among 1- to 3-year-olds.

"I was in constant e-mail contact with the author of the first paper, and I cited her all over the place," Saiger says. "It definitely made my experience easier, because I didn't have to rethink all the methodology-most of it had already been laid out."

• Choose what's most important. Too often, doctoral students feel like they have to include in their dissertation every variable related to their interest that they've ever come across, says Kathleen Malley-Morrison, PhD, a psychology professor at Boston University who teaches dissertation classes.

"But you can't always fit it all into one project," she says.

Instead, Malley-Morrison teaches her students to think logically about which predictors are linked to which outcomes.

For example, a student interested in the effects of social class, stress and other variables on parents' use of corporal punishment could have a multitude of variables to consider.

"I have them draw conceptual models," she says, "with lines from antecedents to mediators to the outcome. As they work with their models, they decide which concepts are really central enough to include."

Working with a conceptually derived model, Malley-Morrison says, helps students keep their research to a manageable size.

"The more constructs you have," she points out, "the larger your sample size needs to be." And sample size, she adds, can constrain students, who often don't have the funding to recruit and run many participants in a study-or who must rely on their school's undergraduate participant pool.

"The hardest part of the process was narrowing and refining the project-and the other students helped so much."

Leila Azarbad University of Virginia

• Tap others' expertise. Your dissertation committee is a valuable resource when you're refining your research methods, says Marcus Patterson, who is completing his dissertation in clinical psychology at Boston University. Patterson, an adjunct at the University of Massachusetts, Boston, co-edited the book "The Portable Mentor: Expert Guide to a Successful Career in Psychology" (Kluwer Academic, 2003). He advises students to make sure that at least one member of their committee is an expert on research methods and statistics.

Fellow students are another good resource, experts say.

For example, Leila Azarbad, a clinical psychology graduate student at the University of Virginia's Curry School of Education, met weekly with a group of students while beginning her dissertation. She studies the physical and psychological factors that influence the accuracy of weight loss expectations among patients who are about to undergo bariatric surgery.

The group began by sharing one-page, general descriptions of their research interests and then eventually moved on to evaluating project outlines.

"The hardest part of the process was narrowing and refining the project," she says, "and the other students helped so much."

• Remember that the dissertation is not a life's work. When choosing research methods, consider how long the project will take to complete, says Saiger. "And however long you think it's going to take, multiply that by three," he adds. Often, he says, the best advice is to keep it simple.

For example, says Patterson, the first question many clinical psychology students face is whether to do a clinical outcomes study. Such studies-which follow a group of participants through a treatment or intervention to determine its effectiveness-are both prestigious and useful for the public good, he says. But they are also very time-consuming, and thus not always the best choice for a student who wants to complete his or her doctorate on time.

Instead, experts say, consider a less time-intensive but equally valuable project-a study that takes place at one point in time, for example, rather than a longitudinal one-on a related subject.

"For your dissertation, you want to do something that's reasonable, and that has a relationship to what you want to study in the future," Patterson says. "Your dissertation should set you up for future research."