Homeschooling outcomes: How do they compare?

Homeschooling has received high marks…when parents provided structured lessons

homeschooling supplies arrayed on a red desktop, including pencils, paper, rubber bands, tape, a ruler, etc.

Are homeschooled students good students? When the topic comes up in conversation, people often cite studies showing that homeschoolers score higher on standardized tests.

For instance, Eric Rudner analyzed the test scores of over 20,000 American homeschooled students and found them to be “exceptionally high—the median scores were typically in the 70th to 80th percentile” (Rudner 1999).

That’s impressive, but we have to keep in mind: This wasn’t a random cross-section of homeschoolers. Participants were recruited from a special subset of the homeschooling population — families who subscribed to a fee-based testing service.

Compared to their peers in the public schools, these kids were more likely to have affluent, well-educated parents. Were the parents also more committed to educating their children? Perhaps.

Then there is the problem of self-selection. Who agrees to participate in a study of this kind?

Parents may be more likely to sign up if they believe their children will test well. About 52% of those approached agreed to participate in Rudner’s study. So we have to wonder about the people who declined. When we compare Rudner’s homeschoolers to the general population, it’s a bit like apples and oranges. The parents of public school kids aren’t a select group of motivated volunteers.

Finally, there were differences in the way the tests were administered. Ideally, we’d want everyone to take the test under the same conditions, under the eye of a trained test administrator. But whereas public school students took their tests in the classroom, many homeschoolers took their tests at home with a parent.

New data: “The Impact of Schooling on Academic Achievement: Evidence from homeschooled and traditionally schooled students”

Recently, Sandra Martin-Chang of Concordia University led a new study that attempts to address these problems.

Martin-Chang and her colleagues sought Canadian participants from both the homeschool and public school populations, recruiting through community announcements, radio ads, and email.

They ended up with 37 homeschool students, and matched these with 37 similar-age public school students living in the same area. Overall, the students had these characteristics:

They ranged in age from 5 to 10 years, and almost all of them lived with married or partnered adults.
Most had mothers with college degrees (65% for homeschoolers, 54% for public school kids), and kids in public school were more likely to have mothers with graduate degrees (11% for homeschoolers, 30% for public school kids).
Homeschool families had lower incomes, presumably because mothers in these families were more likely to have left the workforce.

In addition, the researchers discovered that the homeschooling group fell into two categories.

Most homeschooling parents took a structured approach to education. They “set out clear educational goals for their children and offered structured lessons in the form of either purchased curricula or self-made lesson plans (often some combination of both).”

A minority of homeschooling parents said they rarely or never used premade curricula and structured lesson plans. Some called themselves “unschoolers.” As the authors note, such parents “identified more with the pedagogical view that education is gained via the natural consequences of the child’s day-to-day activities.”

These approaches to homeschooling were very different, so Martin-Chang and colleagues didn’t lump them together. Instead, they decided to study distinct three groups:

Public school students
Structured homeschooling students
Unstructured homeschooling students

Achievement testing, and the results

How did the three groups compare? To find out, researchers administered a 45-minute achievement test in the children’s homes. The questions — which were borrowed from the popular Woodcock-Johnson Test of Achievement — covered seven distinct academic areas, including reading comprehension, science, and mathematics.

Overall, the structured homeschooling group performed much better than the public school group. And the margin was pretty dramatic.

In 5 of 7 test areas, (word identification, phonic decoding, science, social science, humanities) structured homeschoolers were at least one grade level ahead of public schoolers.

They were almost half a year ahead in math, and slightly, but not significantly, advanced in reading comprehension.

But this is a relatively small study. Was the homeschool advantage due to random factors?

It’s unlikely.

Researchers calculated the probabilities of getting these results due to random chance alone. For science and calculation, these probabilities were 1.9% and 2.6%. For word identification, decoding, and social science, the probabilities were all below 0.07%.

Was the homeschool advantage merely the result of socioeconomic privilege? That seems rather unlikely too. Homeschoolers retained their edge even after researchers made statistical adjustments for differences in family income and mother’s education level.

And if the recruitment process selected for homeschoolers with high skill levels, we can say the same about public school students. Both groups — structured homeschoolers and public schoolers — consisted of volunteers. Both tested well above grade level.

So the implications seem clear: Canadian kids receiving structured home schooling are testing very well, and it’s not merely a reflection of their parents’ affluence or educational levels.

But the story may be different for kids who receive unstructured homeschooling.

In every test area, unstructured homeschoolers got lower scores than the structured homeschoolers did.

In 5 of 7 areas, the differences were substantial, ranging from 1.32 grade levels for the math test to 4.2 grade levels for the word identification test. Where the structured homeschoolers performed above grade level, the unstructured homeschoolers performed below it.

The chance that unstructured homeschoolers performed worse due to random factors? Less than 0.07%. And again, the pattern held even after controlling for family income and maternal education. Unstructured homeschoolers also performed worse than the public school kids did, though not by enough margin to rule out chance.

The researchers conclude that “structured homeschooling may offer opportunities for academic performance beyond those typically experienced in public school.”

What are these opportunities?

They seem pretty obvious. Homeschooling typically involves a low teacher-student ratio and highly individualized instruction. It’s private tutoring, which has always been associated with efficient learning.

But Martin-Chang and colleagues are keen to point out the limitations of this research. We need more studies with larger samples. And the researchers would like to investigate the relationship between structure and academic achievement. Might homeschool students benefit from a mixed approach? If so, how much structure is optimal?

I wonder, too, about individual variation. We all know that some kids find it harder to adapt to demands of formal instruction. Are some parents are drawn to unstructured homeschooling because their kids don’t fit the mold? If so, that might explain some of the results here. And it suggests that homeschooling parents — like many classroom teachers — may need to find new ways to reach these students.

References: Homeschooling outcomes

For a concise analysis of the history of research on this topic, check out Eric Isenberg’s article for the Peabody Journal of Education:

Isenberg E. 2007. What have we learned about homeschooling? Peabody Journal of Education, 82: 327–409.