The dark bar smells of decades of cigarette smoke. The Mexicans gathered at the table are desperate, poor and about to risk all they have to cross the border into the United States. Ted Conover tells this story in his book “Coyotes: A Journey Through the Secret World of America’s Illegal Aliens.” To research the book, Conover posed as an immigrant for a year, crossing the U.S.-Mexican border four times and traveling with migrant workers through various states, picking lemons and oranges alongside them.

Conover used a technique known as immersion, a practice that goes back to George Plimpton’s “Paper Lion” and Henry David Thoreau’s “Walden.” The writer submerges himself in his subject, allowing his quest for the story to take over his life. Plimpton, for example, spent a season in training camp with the Detroit Lions. Some writers spend years researching their subjects in this painstaking way. But the intimate details and thorough understanding that result make immersion worth the effort. “I believe in immersion in the events of a story,” writes Tracy Kidder, Pulitzer Prize-winning author of “The Soul of a New Machine,” “House,” “Hometown” and many other books. “I take it on faith that the truth lies in the events somewhere and that immersion in those real events will yield glimpses of that truth.”

Successful immersion often demands that the writer’s presence be as unobtrusive as possible, and some writers go to great lengths to blend in. To understand the lives of American men, Los Angeles Times columnist Norah Vincent worked out to add 15 pounds of bulk to her shoulders, wore a cupless sports bra, glued on facial stubble made of wool crepe paste and sported a flattop, rectangular glasses and rugby shirts. According to a review in The New York Times, the resulting book, “Self-Made Man,” “transcends its premise altogether, offering not an undercover woman’s take on male experience but simply a fascinating, fly-on-the-wall look at various unglamorous male milieus that are well off the radar of most journalists and book authors.”

Such drastic action is not always necessary, however; writers such as Kidder and John McPhee manage to minimize their presence during immersion without resorting to disguises. McPhee, in particular, is known for disappearing into the background, not only while reporting but also in the resulting stories, employing the first-person pronoun only when reporting scenes that take place in situations (on a canoe trip, in a small airplane) which would seem odd if written without an acknowledgment of the author’s presence.

Of course, as Conover’s experience illustrates, being completely invisible is often impossible. Conover has blond hair, blue eyes and an Amherst education, and he admits that his presence probably changed what happened when the Mexicans were caught on the border and how they spoke to each other in his presence. Conover himself was part of the story, and so when writing, he decided the third person would not do. He stuck out in that bar as no fly would, and thus, he had to use the personal pronoun I. It was the honest thing to do.

Although Conover may not have had much of a choice, writers can keep their influence to a minimum during immersion by limiting their interaction with subjects. Lee Gutkind, in “The Art of Creative Nonfiction,” suggests that for many writers, “the tendency is to find a place for yourself and to help out in order to make yourself be and feel useful. This approach does not usually work well. … For a writer, sitting, watching and taking the occasional note is a key and vital activity. If you are helping the people you are observing on a regular basis, then you are not writing. If you are perceived as part of the team, then you are not perceived as a writer, a misconception that may lead to misunderstandings down the line.”

How can a writer know when the immersion process is over? Gutkind advises that you hang in with your first round of immersion until you can think of no more questions to ask. After writing your first draft, you’ll see where the gaps lie in your research, which will force you to go back into immersion for more material. “I usually will go back three or four times until my essay or book seems complete,” says Gutkind.