Research: Unexplored Territory

By Jeff Archer — December 08, 1999 17 min read
Home schooling is growing, but many researchers shy away from the topic.

Imagine this call for papers: “Wanted: Studies examining fast-growing educational innovation. Researchers must be willing to contend with near-impossible sampling problems, risk being labeled ideologues, and have their work trashed by supporters and opponents of the approach. Funding possibilities—and the likelihood of publication: minimal.”

Welcome to the world of home schooling research.

Anyone who’s studied some aspect of education knows that research in the social sciences is rife with pitfalls. Conducting controlled experiments is difficult, and sometimes impossible, and any study is likely to prompt not just scientific scrutiny, but political reaction.

There are few areas of education research, though, where those troubles are as prevalent as they are in home education. Technical challenges abound, most of them related to the fact that many home schoolers remain “underground.” At the same time, a debate persists in the home schooling community about whether its practitioners should even cooperate with researchers. And since home schooling still is often viewed as an affront to public education, many people are even quicker to question motives than methodologies.

“Everyone says: ‘Are you for it or against it?’” says Maralee Mayberry, a sociology professor at the University of Nevada, Las Vegas, and the lead author of the book Home Schooling: Parents as Educators. Though her now 8-year- old research simply sought to describe who home schools and why, she says, “I am still called a home schooling advocate.”

When Lawrence M. Rudner published one of the most rigorous analyses of home schoolers’ achievement data earlier this year, it was the first time his research ever generated hate mail, he says. His work—which cast home schooling in a favorable light—also prompted the publication of two critical responses. One actually ran in a home schooling magazine, which called his study “embarrassing and dangerous.”

“Usually when I say nice things about people, they don’t say bad things back,” says Rudner, who directs the Educational Resources Information Center Clearinghouse on Assessment and Evaluation. Based at the University of Maryland College Park, the federally financed center publishes a bibliographic database of education research.

Add to these quirks the claim—made by many researchers—that the interest of scholarly journals and potential funders diminishes the further a topic moves away from public education. The result is a dearth of high-quality studies on a form of education that, by most accounts, serves at least 1 million U.S. children.

But as the home schooling phenomenon leaves the educational fringes and gains in popularity as an alternative to traditional schools, the need for information becomes increasingly pressing. Exactly how big is the movement? What are those students learning? Is the quality of their education better, or worse, than that of students in public or private schools? How much does it cost? How will home schoolers do when they get to college?

For now, researchers have few definitive answers.

“It’s dozens of isolated, one-time studies,” Gregory J. Cizek, an education professor at the University of Toledo, says of the current state of research. “It’s a very uncoordinated research agenda, by people who often don’t know each other or who have never met each other, and it’s the first or last thing they’ve ever done on home schooling.”

‘Simply Descriptive’

Cizek’s assertion is based on more than a hunch. In the early 1990s, he analyzed three years of articles published in the Home School Researcher, the only such peer-reviewed journal devoted exclusively to home education.

The publication is edited by Brian D. Ray, who is also an exception to the rule in that he heads the only organization dedicated solely to the study of home schooling—the National Home Education Research Institute, based in Salem, Ore. Ray has home schooled his own children and has testified in support of home schoolers involved in litigation, but the institute he leads is independent. The bulk of its funding comes from the sale of publications, donations, and research contracts.

Cizek examined the research behind the articles and surveyed their authors. He was surprised to discover that nearly half the writers had never engaged in home schooling themselves. The finding called into question the claim that home education research is the “sole domain of its advocates,” he wrote.

He also found, however, that very little of the work could be considered of an experimental nature by scientific standards, or even quasi-experimental. Much of the research also was based on very small samples. His conclusion: Home schooling had not been subjected to much rigorous study.

“We would call the research on home schooling very applied,” he says. “Which means it is simply descriptive: It says what the state of affairs is. Or it’s located in a particular context that makes the results not strongly generalizable.”

Doubtless, one of the reasons is that no one can draw a truly representative sample of home schoolers because no one knows exactly who they are. Nationwide surveys show the numbers of students enrolled in public schools, charter schools, Roman Catholic schools, and other private schools, but no similar census exists for home education.

By their nature an independent lot, home schoolers lack any unifying national organization. Moreover, some states require little official documentation from home schooling parents. Even where filing requirements are more extensive, some parents refuse to cooperate.

“You can’t generate a list—even state by state—of who is home schooling, and if you could, the question is: Would they talk to you?” says Susan A. McDowell, the managing editor of the Peabody Journal of Education, a scholarly publication based at Vanderbilt University’s college of education in Nashville, Tenn. The journal plans a double issue on home schooling next spring.


The best anyone has been able to do on the number of home schoolers is make an educated guess. Few researchers have thought about the issue as much as Patricia M. Lines, who recently left the U.S. Department of Education’s office of educational research and improvement.

A lawyer by training, Lines began examining home education in the 1980s when she was the director of the legal center of the Denver-based Education Commission on the States. Based largely on surveys of companies that sell curricular products to home schooling parents, she guessed that some 50,000 children were being educated at home.

When Lines mentioned that figure in an Education Week Commentary in 1985 (“States Should Help, Not Hinder, Parents’ Home- Schooling Efforts,” May 15, 1985), one reader wrote in a published response that such estimates were likely “subject to the inflationary effects of wishful thinking, or at least artful advocacy.”

That was when Lines, who sent her own son to public school, realized how touchy the subject was. Any time a researcher sets out to explore the politically charged subject, she says, “someone’s going to question your motives.”

But she continued to design more sophisticated methods to estimate the population—all of which have shown the number skyrocketing from where she guessed it was 15 years ago. Before leaving the Education Department this year, she published a paper describing three different approaches.

Each used information she collected from state education officials about the number of documented home schoolers in their states and their rules on whether home schooling parents had to file official paperwork.

To guess at how many “nonfilers” there were in each state, she depended on the results of surveys done by Ray of the National Home Education Research Institute. Using home schooling organizations and other groups, Ray has polled parents about whether they file any documentation with state or local education agencies.

Using the data, Line’s approaches yielded estimates of between 691,000 and 750,000 children for the 1995-96 school year. Based largely on his own contact with home schooling groups and state education officials, Ray estimated that the population was closer to 1.2 million that year.

“In terms of a planned overall critical approach, she puts more into it than I do,” says Ray. But, he adds, “that doesn’t prove to me that she’s right.”

Better Than Average

Ray periodically ratchets up his estimates based on further surveys and contact with home schooling groups. This year, for instance, he found that many such groups said the news of the Columbine High School shootings and other such incidents had piqued the interest of more parents. Currently, he says, between 1.2 and 1.8 million children are educated at home. Lines says it’s likely that the figure has passed the 1 million mark.

Soon, researchers may have yet another source. The Education Department is, for the first time, analyzing home schooling data based on national household surveys, including some from the U.S. Census Bureau. Although the data may shed more light on the size of the population, the sample isn’t large enough to provide many demographic breakdowns showing who is most likely to be a home schooling parent, officials say.

The department expects to release its report within a few weeks. When it does come out, says Rudner of the ERIC Clearinghouse, “there will be a whole bunch of arguments about the methodology.”

Given that no one knows exactly how big the home schooling population is and who is in it, each quantitative study has had to contend with significant hurdles.

Ray has carried out two nationwide studies of achievement among home schoolers. In both, he depended on home education organizations to survey their members, who were asked to fill out questionnaires and return information about their children’s test results.

Researcher Lawrence M. Rudner helps his son, Andrew, with schoolwork. Though he is not a home schooling parent, the ERIC Clearinghouse director has published an analysis of achievement among home school students. (Photo by Benjamin Tice Smith.)

His 1997 study—the larger of the two—included data on students who had taken several different nationally normed tests. Ray found that those students, taken together, had scored at the 84th percentile. He also found statistically significant, but what he termed “relatively small,” variances in home schoolers’ achievement depending on their parents’ income and education levels.

As Ray pointed out in his published report, the study had a number of limitations, however. He couldn’t know exactly how representative his sample was. Further, the fact that parents knew what their children’s scores were before deciding to participate opens up the possibility that those with better results were more likely to respond.

And he could not compare his sample of home schoolers with public school students who had families with similar income levels and family backgrounds.

“We who do research on home schooling have to be very careful not to attribute cause and effect,” Ray says. “I don’t think you’ll ever see cause and effect with home schooling, because we can never have a true experiment. You can’t take 1,000 children and randomly divide them up into different groups.”

‘The Wrong Question’

Rudner got around some of those problems when he carried out his own nationwide study of student achievement this year. Both his study and those by Ray received funding from the Purcellville, Va.-based Home School Legal Defense Association, which defends home schooling parents in legal disputes with state and local officials.

He used data supplied by the Bob Jones University in Greenville, S.C., which sells test-taking services using the Iowa Tests of Basic Skills to home school parents. Rudner examined a larger number of test-takers than Ray, and unlike earlier research, the parents in his study did not know their children’s scores before deciding to participate.

His results, published in the online journal Education Policy Analysis Archives, showed that students in his sample scored, on average, between the 82nd and 92nd percentiles.

But those findings say nothing about why those students scored high, Rudner says. Those in his sample tended to be of a higher socioeconomic status, a characteristic considered by many experts to be the strongest predictor of achievement.

“Everyone wants to ask the question: Is home schooling better than public schooling?” Rudner says. “It’s the wrong question, and it will never be addressed.”

The right question, he believes, is: “Does it work for those who try it and make the commitment? And my study says yes.”

Rudner’s study had its own limitations, though. A response to his report—also published in Education Policy Analysis Archives—praised his analysis but suggested that some parents might be reluctant to hire Bob Jones University’s services for religious or ideological reasons. The university has a fundamentalist Christian orientation and has drawn criticism for its opposition to interracial relationships, the response noted.

Though the ITBS isn’t tailored to the university’s deeply conservative orientation, it’s possible that nonwhites and non-Christians may be less inclined to purchase the school’s services.

Rudner found that nearly 94 percent of his sample considered themselves Christian, and only about 1 percent were either African-American or Hispanic. Rudner says he turned to Bob Jones University for a simple reason: “There’s no one else who has that kind of data.”

He stands by his analysis, but agrees that he should have been more explicit about the possible implications of using data from the university: “I made some disclaimers, but clearly not enough.”

Another possible limitation was that many parents who use Bob Jones’ services administer the tests themselves. Rudner doubts that many parents “cheated,” but researcher Jon Wartes, himself a home schooler, says he’s found that students tend to do better when their parents are the ones giving them an exam.

Now the coordinator of programs serving home schoolers in the 24,000-student Lake Washington public schools in Washington state, Wartes organized an independent and voluntary effort called the Washington Home School Research Project, which carried out studies between 1986 and 1992.

“There is no perfect study,” Rudner says. “They all have limitations, but what consumers need to do is to be much more critical of the research, and they’re not.”

The limitations of each of the studies, however, does not discount their general suggestion that those who are home-schooled tend to do well. It’s just impossible to say how well, or to compare the results meaningfully with those for other students, Lines says.

“If you take the Rudner study and all the other studies, you could do a meta-analysis, and you would find them all coming out with home schooling students testing above the norm,” Lines says.

“At some point I’m willing to step out of the role of researcher and say that home schooling does no harm academically.” But, she adds, “I would not be willing to say the same kids would do differently in schools.”

Reluctant Subjects

Perhaps the most surprising criticism of Rudner’s work came from a group of home education advocates. Last summer, Home Education Magazine published a scathing response to the study that ran for several pages.

Not only did the critique question the sample Rudner used, but it also argued that his work—and similar studies—posed a threat to the rights of home schoolers. Even though the study suggested that home education works, the critique’s authors said that attempts to evaluate and describe the home schooling population could undermine their independence and paint an inaccurate picture of who is in the movement.

“Anyway you slice the American pie, you’re going to find home schoolers,” says Mark Hegener, the magazine’s founder. With a circulation of about 10,000, his magazine is read by everyone from religious conservatives to New Age parents, he says.

Indeed, despite the view that the typical home schooler is a conservative Christian, Mayberry of the University of Nevada says a central finding of her study was the great variety of ideologies and pedagogical beliefs among home schoolers. “The popular myth out there was that this was kind of a monolithic group motivated by religious attitudes,” she says, “and I think one of the things our book was able to do was to shatter that myth.”

Most surveys have shown that the vast majority of home schoolers consider themselves Christian—as do most Americans—and that religion often figures prominently in their decisions to educate their children at home. But, for now, it’s impossible to say for exactly how many the religious motivation is true.

More significantly, Home Education Magazine‘s response to Rudner’s work argued that such studies jeopardize the very reason why many home schoolers have opted to educate their children themselves: They don’t buy in to the values of conventional education and what they see as its one-size-fits-all measures of success.

“When we start doing research, we start relying on credentials, on experts, and start moving away from parents and families and their own values,” Hegener says.

Such a perspective presents an obvious challenge for researchers. Not only must they deal with the increased likelihood of self-selection bias, but they also face the prospect that potential subjects will actively resist any kind of external evaluation.

“I’ve never seen anything like this in my life,” Rudner says of the publication’s advice for home schoolers to oppose research. “But we know there are groups out there with different viewpoints, and they are entitled to them, no matter how bizarre they may seem to us.”

Another Approach

Others believe the research to date has had the wrong focus. Most of the major studies have looked at achievement, using traditional standardized tests as a yardstick. A few studies have also examined the socialization of home schoolers, addressing a common concern that children educated outside school may not develop well socially.

Both areas of research involve value-laden assumptions, either of what it means to be educated or of what it means to be socially mature.

The University of Toledo’s Cizek suggests that researchers try examining home schoolers more on their own terms. If it appears that many parents educate their children at home for religious reasons, then a valid question is how the experience works to buttress certain values while militating against others, he argues.

Sociologists have, for example, examined the link between having attended a Catholic school and the likelihood of practicing the faith as an adult.

“There haven’t been many studies about persistence,” Cizek says. “What’s the typical experience of a home schooling family? It’d be nice to know something about the post-home-schooling experience.”

Now that colleges and universities are reporting larger numbers of home schooled students on their campuses, another question would be how well they do in a traditional higher education environment after having been educated away from conventional K-12 schools.

Others believe home schooling provides a unique setting for studying how students learn in different environments.

In the mid-1980s, University of Cincinnati anthropologist Mary Anne Pitman spent six weeks observing a group of home schooling families to discover how children acquired information outside formal educational structures. She found that the students learned quite well from spontaneous events.

Home schooling gives researchers “an opportunity to get access to a variety of processes, but most of the research still looks at outcomes,” says Pitman, whose research was published in a book she co-edited, Home Schooling: Political, Historical, and Pedagogical Perspectives.

Case studies may also yield findings that could be of use in other forms of education. One such examination suggested a link between the ability of parents to engage their children in learning and the fact that they tend to stand or sit next to their children, as opposed to schoolteachers, who usually stand in front of their students.

Home schooling also could provide lessons on alternative forms of assessment, such as portfolios of student work.

“I do share the feeling that at some point the researchers who are trying to take the schools’ measurements and apply them to home schooling are comparing apples and oranges,” says Patrick Farenga, the president of Holt Associates, which provides information on home education and alternative schools. “But at the same time, I don’t think there should be no research, just a different kind of research.”

As the perception of home schooling as a legitimate educational alternative grows, it’s possible that more rigorous and more varied types of studies may be carried out. Ray says he’s seen an increase in the number of dissertations on the topic in recent years, though that number is still relatively small compared with other areas of education research.

The question is whether that increase will be followed by growing interest among grantmakers and prestigious journals.

“Back in 1989, I’d say I knew 98 percent of the people who were doing dissertations on home schooling, either through letters or phone calls,” Ray says. “Now, there’s no way I could know them all.”

The Research section is underwritten by a grant from the Spencer Foundation.

A version of this article appeared in the December 08, 1999 edition of Education Week as Research: Unexplored Territory