Conducting empirical research on homeschooling is complicated by two factors. One is the overall lack of longitudinal data on the topic—especially data not reported by survey or other voluntary means. The other is lack of clarity as to what precisely constitutes “homeschooling” in the first place, especially as educational options have broadened in recent years (virtual schools, micro-schooling, ESAs, etc.). A new report from the Journal of School Choice clearly illustrates the issues homeschooling researchers face and offers a couple of insights for further work.
Researcher Albert Cheng of the University of Arkansas uses five nationally representative data sets to construct a picture of the information available on homeschooling. Three of these focus on school-aged children in the U.S. who were being homeschooled at the time their families were surveyed for the National Household Education Surveys Program (NHES) in 2012, 2016, and 2019. The other two focus on U.S. adults who report ever having been homeschooled. Importantly, the adult surveys include full education histories for the respondents, which allows Cheng to identify times when they were most likely not being homeschooled (i.e., formally attending a public or private school). The observable patterns of students moving into and out of homeschooling prove interesting and important.
First up, findings from the Understanding America Study (UAS), conducted between February 2016 and October 2022 by the University of Southern California’s Center for Economic and Social Research. Among those who reported ever having been homeschooled, the majority either did so for only one to two years total (43 percent), or else for all of their primary and secondary education careers (11 percent). The remainder ranged between three and nine years of homeschooling. Very few adults reported being homeschooled for 10 or 11 years. Adults who were homeschooled for only one year spent about 10 years in traditional public schools. Adults who were homeschooled for two or three years were enrolled in traditional public schools for about eight years. Meanwhile, adults who were homeschooled for four or five years spent nearly an equal amount of time in traditional public schools. Private school attendance was low across the board for UAS homeschoolers, but peaked for students reporting six (2.4 years), four (1.9 years), and seven (1.5 years) years of homeschooling. Interestingly, some of those mid-range homeschoolers also reported from one-half to nearly a full year spent not attending school at all during their K-12 careers.
Next, findings from the 2023 Education Survey fielded by Cardus, a Canada-based research entity. Their most recent U.S. survey sought to “examine the school-sector effect in the lives of a nationally representative sample of high school graduates, aged 24 to 39.” The results are remarkably similar to those of UAS: Thirty-six percent of respondents who were ever homeschooled did so for just one or two years while 17 percent did so for all of their primary and secondary education. The mid-range homeschooling durations were also similar, as were the splits between non-homeschooling time spent in traditional public (a lot) versus private schools (a little).
According to both UAS and Cardus, the grade levels at which students moved into and out of homeschooling were all over the map. For example, nearly one-third of adults who were homeschooled in eighth grade exited homeschooling for ninth grade, which could make some intuitive sense given the greater complexity of high school-level curricula and/or the numerous requirements for adequate college preparation. At the same time, though, 64 percent of ever-homeschooled adults were enrolled in formal schools in eighth grade and only about six percent of them moved to homeschooling for ninth grade. Such dichotomies exist at other grade levels, too, indicating what is likely a wide variety of motivations (unobservable in the data at this level) for families choosing to start or end homeschooling at any given time. The main point is that the choice to homeschool seems to be under evaluation at all times among this population.
Finally, Cheng uses the (admittedly incomplete) data from three administrations of NHES as a sort of robustness check on the other survey results, and finds broad congruence. For example, one out of every five NHES children who were being homeschooled at the time of data collection in 2012 had been homeschooled for less than 20 percent of their years of schooling, while 36 percent of those children reported being homeschooled for all of the educational career they had completed at the time of the survey.
These findings, confounding as they may seem, actually provide interesting food for thought for future research on homeschooling. For instance, only a small percentage of students likely fit the definition of “homeschooled” if that is limited to students who never attended a formal public or private school. Indeed, the findings generally reject easy dichotomous comparison between “homeschooled” and “non-homeschooled” students, given the amount of movement between sectors that’s evident in these data. There is a whole continuum of students—probably the majority of those who would cop to the “homeschooled” label—who have actually experienced varying amounts of each type of education. It’s likely that even more students could opt for the “homeschooled” label in the wake of pandemic innovations and post-pandemic school choice expansion. And since those options muddy the definition of “homeschooling” further, future analysts must hone their meanings and refine their comparisons according to the truer picture of what homeschooling is and isn’t, even before they start examining their data.
SOURCE: Albert Cheng, “The Year-By-Year Primary and Secondary Education Histories of Homeschooled Individuals and the Implications for Empirical Homeschooling Research,” Journal of School Choice (December 2024).
