On this page
Disclosure. Some links on this page are affiliate links. Every Homeschool may earn a small commission if you purchase through them, at no extra cost to you. Editorial picks are not influenced by commissions; see how we make money.
As an Amazon Associate I earn from qualifying purchases.
Introduction
A family weighing home education may ask whether the child will be alright at the end of it. In our assessment, the research can take you partway, and the favorable findings need caveats attached. This page lays out the favorable numbers and the objections to them side by side.
Two facts frame everything that follows. Homeschooled students in the large advocacy-sponsored studies do score well, often very well (Kunzman & Gaither). And the reviewers note that these studies of academic achievement do not employ random sampling nor control for confounding variables (Kunzman & Gaither), which in our assessment means they cannot separate the effect of homeschooling from the kind of family that homeschools. For the broader picture of who homeschools and how many, see the homeschool statistics hub and the companion guide on who homeschools and why.
Summary
- The best-known percentile figures come from volunteer samples. NHERI reports that home-educated students “typically score 15 to 25 percentile points above public-school students” on standardized tests (NHERI), while Kunzman and Gaither note that the five large achievement studies conducted under HSLDA sponsorship relied on samples of homeschoolers recruited for the purpose (Kunzman & Gaither).
- Control for family background and the gap shrinks. A comprehensive 2020 review of the research reports that across many smaller studies homeschooling “does not have that much of an effect on student achievement once family background variables are controlled” (Kunzman & Gaither).
- The 2003 adult study had a 12% response rate. Brian Ray’s 2003 survey of adults who were homeschooled, summarized in HSLDA’s “Homeschooling Grows Up,” drew about 7,300 replies from roughly 58,500 contacts, with 65% of respondents’ families current or former members of a single advocacy organization (CRHE).
- Representative samples paint a less flattering picture. Reviewers say that the Cardus Education Surveys, “probably the most rigorous data set ever created to measure homeschooling’s long-term academic impact,” show homeschoolers “as a whole do not have great educational and economic success if measured by conventional standards like a college degree and a high-paying job” (Kunzman & Gaither).
- Some studies hint that outcomes spread at both ends. One analysis of 182,351 records found homeschooled children age 12 and up were two to three times more likely to report being behind grade level (via Kunzman & Gaither), a result the reviewers present as a hint that needs more research.
The headline numbers
Start with what gets quoted. The National Home Education Research Institute reports that home-educated students “typically score 15 to 25 percentile points above public-school students on standardized academic achievement tests,” and that they “go to college at a similar rate and succeed at college at an equal or higher rate than the general population” (NHERI Research Facts).
The adult-outcome version comes from a 2003 study commissioned by the Home School Legal Defense Association and conducted by Brian Ray, surveying more than 7,300 adults who had been homeschooled, over 5,000 of them for at least seven years (Homeschooling Grows Up). The figures it produced are striking on their face.
| Measure | Home-educated adults | Comparison group (U.S. data) |
|---|---|---|
| Took college-level courses, ages 18 to 24 | 74% | 46% |
| Voted in the past five years, ages 18 to 24 | 76% | 29% |
| Participate in ongoing community service | 71% | 37% |
| Member of any organization (community group, church or synagogue, union, homeschool group, or professional organization) | 88% | 50% |
| “Very happy” with life | 59% | 28% |
In our assessment, taken alone that table looks decisive: 74% of the home-educated adults aged 18 to 24 had taken college-level courses against 46% of the comparison group, 76% had voted in the past five years against 29%, and 71% took part in ongoing community service against 37% (Ray, 2003). The question these figures raise is who answered the survey.
The sampling problem
The Coalition for Responsible Home Education examined how Ray’s respondents were recruited and found a sample stacked toward the most committed end of the movement. Roughly 12% of those contacted responded, about 7,306 out of some 58,500, a rate the group calls abysmal beside the 70 to 80% response rate it cites as the goal for surveys of this type (CRHE). In our assessment, a low response rate matters because the people who answer an advocacy survey may not be a random slice of the population.
CRHE also examines the composition of the sample. About 65% of participants’ families were or had been HSLDA members, even though, by CRHE’s estimate, HSLDA member families made up between 6 and 10% of the national homeschooling population when the study was conducted. The sample was 93% white against 77% of homeschooled students nationally in 2003 (NCES), with just 9 Black respondents out of more than 5,200, and 85% raised in a Protestant religious environment (CRHE). CRHE also reports that 72% of the 7,306 respondents had been home-educated seven years or more and argues that the participants were therefore likely to be long-term ideological homeschoolers (CRHE). In our assessment, what the survey measures is the outcome of a particular subset of homeschoolers, not homeschooling in general.
The academic literature raises the same objection. Reviewing the five large achievement studies conducted under HSLDA sponsorship between 1990 and 2010, Kunzman and Gaither note they “all relied for their data on samples of homeschoolers recruited for the purpose,” with volunteers submitting self-reported scores from tests “typically proctored by the parent in the home,” on the understanding that the results would be used for advocacy (Kunzman & Gaither). In every case, the reviewers report, homeschooled students consistently scored in the 80th percentile or above on nearly every measure. In our assessment, that consistency is what you would expect from a sample built out of confident, engaged families.
Lawrence Rudner’s 1999 analysis of 20,760 students drew its subjects from a single testing service, the Bob Jones University Press Testing and Evaluation Service. The sample came out “far whiter, more religious, more married, better educated, and wealthier than national averages” (Kunzman & Gaither). Rudner himself wrote that the study “does not demonstrate that home schooling is superior to public or private schools” and “should not be cited as evidence that our public schools are failing.” The reviewers call it “perhaps the most misrepresented research in the homeschooling universe” (Kunzman & Gaither).
What controlled studies find
The reviewers report that several smaller-scale studies do control for family background variables. One generalization that emerges from many of them is that “homeschooling actually does not have that much of an effect on student achievement once family background variables are controlled,” a conclusion the reviewers note is implicit even in many HSLDA-funded studies (Kunzman & Gaither).
A second pattern holds across the three decades of research in Kunzman and Gaither’s 2020 review: homeschoolers tend to do better on verbal measures than on mathematics. One well-designed study that controlled for family background (Belfield, 2005) found homeschooled seniors taking the SAT scoring slightly better than predicted on the verbal section and slightly worse on math, and state testing data from Arkansas and from Alaska’s correspondence school program also show lower math results for homeschoolers (Kunzman & Gaither). The reviewers conclude, tentatively, that there may be at least a modest homeschooling effect on academic achievement, one that tends to improve verbal skills and weaken math capacities (Kunzman & Gaither).
Studies of children at the extremes hint that the distribution may be wider at both ends, not only the top. An analysis of National Survey on Drug Use and Health data covering 182,351 records found homeschooled children age 12 and up were “two to three times more likely than their public-school equivalents to report being behind grade level” (via Kunzman & Gaither). The reviewers add that much more research is needed, that the pattern may have nothing to do with homeschooling itself, and that, as the analysis’s author notes, homeschooled children behind grade level “may well be struggling academically before homeschooling commences” (via Kunzman & Gaither).
Kunzman and Gaither describe a demographically matched comparison, pioneered in 2011 by Martin-Chang and colleagues, as real progress: recruit homeschooled and conventionally schooled children with similar backgrounds and test both under the same conditions. That 2011 study found homeschoolers using a more structured approach outperforming their matched public-school peers, while those using an unstructured “unschooling” approach underperformed (Kunzman & Gaither). In our assessment, that suggests the approach may matter more than the label, a nuance the headline percentile claims do not capture.
Adult outcomes and college
On college, the reviewers note that most studies of homeschoolers in college are convenience samples taken from the researcher’s own university. Several found homeschoolers outperforming institutionally schooled peers with similar demographic backgrounds on grade point average, a few found no difference, and surveys of admissions officers consistently find that homeschooled applicants are accepted at roughly the same rates as their conventionally schooled peers (Kunzman & Gaither). Those findings are worth knowing, and the practical side of getting in is covered in the guide to college admissions for homeschoolers.
The catch, in the reviewers’ words, is that studies of homeschoolers in college are “not capturing the post-homeschool experiences of most homeschooled children,” because representative studies find that very few homeschoolers take the SAT or ACT or go on to attend selective four-year colleges and universities (Kunzman & Gaither). The Cardus Education Surveys, run on representative samples of young adults aged 24 to 39 in the United States and Canada, found in every phase that formerly homeschooled respondents “reported lower SAT scores than the privately schooled subjects, attended less selective colleges for less time, and ended up working in jobs for lower pay” (Kunzman & Gaither). The reviewers call this “probably the most rigorous data set ever created to measure homeschooling’s long-term academic impact.”
There is a fair rejoinder built into that finding. As several researchers point out, conventional yardsticks like degree attainment and salary “might not be what is motivating a large percentage of homeschooling families anyway” (Kunzman & Gaither). In our assessment, the evidence tells you about averages on the metrics that were measured, not about whether a given family met its own goals.
Socialization, briefly
The socialization question gets its own treatment in a dedicated guide, but the methodological note is the same one running through this page. Studies on homeschooler socialization have “relied almost entirely on self-report of students or their parents,” and most treat schooling as a simple yes or no, ignoring how many years a child was home-educated or whether the family mixes settings (Kunzman & Gaither).
How to read the evidence
In our assessment, the high-profile favorable numbers come from self-selected samples of homeschooling families, while the representative data the reviewers describe report weaker results on conventional measures and some studies hint at a wider spread at the edges. Kunzman and Gaither, who describe themselves as “neither indiscriminate advocates for homeschooling nor unrestrained critics,” note that the great majority of homeschooling scholarship is qualitative and based on small convenience samples, so that the scholarship in the aggregate is “little more than a series of anecdotes embellished by elegant methodology,” a limitation they say remains, though with marked improvement (Kunzman & Gaither). In our assessment, that is a caution, not a verdict against homeschooling.
In our assessment, the research cannot tell you how your child will turn out, because many smaller studies that control for family background find that homeschooling does not have that much of an effect on achievement. For a household weighing the decision, that points the attention somewhere more productive than percentile charts: the daily structure, the curriculum fit, and whether the plan is sustainable over years. The guides on whether homeschooling fits your family, choosing curriculum, and staying sustainable address those variables. When you are ready to match an approach to your child, the curriculum finder is the place to start.
The Thinking & Leadership page sets out six skills a family can schedule and labels the research on each Strong, Mixed, or Unproven; none of that research measured leadership.