This post reviews Linda G. Hanna, “Homeschooling Education: Longitudinal Study of Methods, Materials, and Curricula” in Education and Urban Society 20, no. 10 (2012): 1-23.
Hanna, an education professor at West Chester University, here gives us the results of a study of 225 homeschooling families coming from 25 school districts in Pennsylvania to get representative data on who chooses to homeschool, why they choose it, and how they do it. She surveyed and interviewed these families in 1998 and then went back to them 10 years later for another round of questions. Thus we have here one of the only longitudinal studies of homeschooling ever done.
Hanna begins, I have to say, with a very poor summary of the literature. It’s poor because it’s based on sources published several decades ago. Her summary assumes that a source from the early 1990s that described curriculum or homeschooling magazines then is still applicable today. It also gets a good bit of the history of homeschooling wrong. It reads, basically, like something she probably wrote back in 1996, which was the last time she published anything on homeschooling. But let’s ignore all of that and get to the actual study.
Hanna wanted to get as representative a sample of homeschoolers as she could. She wisely chose Pennsylvania, whose relatively stringent State requirements allowed her to tap into the records of 25 school districts across the state. She chose these districts so as to get a wide array of people–rural, urban, and suburban. In each district ten of the registered homeschooling families were randomly selected and asked if they were willing to participate in the study. Eventually Hanna secured her 250 subjects. Due to attrition and other problems the sample eventually was reduced to 225.
In the fall of 1998 Hanna interviewed every one of these 250 subjects and secured questionnaires from them. Then in 2008 she contacted these families again and gathered data a second time. Thus we have here a beautifully designed longitudinal study of a pretty representative sample of Pennsylvania homeschoolers.
And what did she find? In 1998 the vast majority of the children being homeschooled in her sample were very young, and this was consistent in all regions. Here’s the breakdown by age of the 703 children:
age 6-8: 405 children
age 9-11: 197
age 12-14: 86
age 15-18: 15
The level of parent education varied widely by region. Only 9% of urban and 12% of rural homeschooling mothers had a college degree, while 52% of suburban mothers had one.
In 1998 computer use also varied widely by region, with only 27.5% of urban and 32% of rural families having a computer, while 80 % of suburban (and 90% of suburban-rural) families did.
Fast forward ten years and what did she find? First of all there were only 499 children left. Some had graduated, some had returned to institutional schools, some had moved to other districts not included in the study. Of these 499 remaning children, the spread was much more even:
age 6-8: 123 children
age 9-11: 142
age 12-16: 123
age 15-18: 111
Why was it that by 2008 far more older children were staying with homeschooling? Computers. In 1998 ZERO families, even those who had computers, were using internet materials or online programs for curriculum, and very few were using them for networking with other homeschoolers. In fact, when you look at the data Hanna has on networking in general, in 1998 urban and rural homeschoolers were not very well networked with other homeschoolers (22.5% of urban and only 4% of rural homeschoolers were involved in networks.) Suburban (87.5%) and especially urban-suburban (94%) were.
By 2008, however, almost everybody was using computers for everything. 78% of urban and 72% of rural families now had computers in the home. 80% of urban, 90% of suburban, and a still rather low 30% of rural families were using these computers for curriculum, and networking with other homeschoolers was far more common among urban homeschoolers (80% in 2008) and even among rural homeschoolers (35%). Hanna does note that there remained a significant minority, especially of rural homeschoolers, that continued to distrust the computer as an educative tool and avoided using it for homeschooling even if they owned one.
So more and better curricular options via the computer and more and better networking via the same was helping keep children at home longer.
That’s the main finding from Hanna’s study. Along the way she provides a few more significant morsels that I’d like to mention. First, she found that over the ten year span a significant number of the mothers doing the homeschooling had gone back to college. In 1998 a total of 29.6% of her subjects had a college degree. By 2008 42.2% did. Fully 60% of the mothers had by 2008 acquired at least some post-secondary education. This finding suggests to me a dynamic relationship between many mothers’ homeschooling and their own educational aspirations. When paired with the other demographic data, one gets the picture that as children grow up and either go to school or become more independent working at home with the computer, the mother is freed up to do her own thing a bit more.
A second point Hanna’s study makes has to do with the perennial question of why parents homeschool. What percentage do so for religious reasons? What for pedagogical? Hanna uses Jane Van Galen’s much-discussed term “ideologues” to name the mostly conservative Christians who choose homeschooling so that they can suffuse their children’s lives with their religious beliefs, and “pedagogues” to name homeschoolers not especially concerned with religion but troubled by what they take to be the restrictive forms of education practiced in institutional schools. In 1998 Hanna explained this distinction to her subjects and asked them if they identified themselves as ideologues, pedagogues, both, or neither. Interestingly, though some scholars have taken exception to the actual terms, in both surveys only 5 of the homeschoolers themselves chose “neither.”
Here’s what she found in 1998:
– Urban Urban-Suburban Suburban Suburban-Rural Rural
Ideologue: 20 22 41 18 38
Pedagogue: 16 12 17 6 4
Both: 4 16 21 4 6
Fast forward to 2008 and here is the spread:
– Urban Urban-Suburban Suburban Suburban-Rural Rural
Ideologue: 18 19 24 18 31
Pedagogue: 12 14 21 5 6
Both: 7 12 32 4 7
What do we learn from this? Two things. First, religion has always been and continues to be a huge motivator for homeschooling, even among many of those who also choose it out of a desire to offer more progressive pedagogical practices to their children. Corroborating evidence of this comes from Hanna’s data on the percentages of families who use explicitly religious curriculum. In 1998 57.5% of urban, 55% or suburban, and 96% of rural families used religious curriculum. In 2008 an even larger percentage did so (75% urban, 62.5% suburban, and 90% rural). While nobody really knows what percentage of homeschoolers are conservative Christians, Hanna’s Pennsylvania sample suggests that’s very high, 75% or greater, and that if anything this percentage is growing.
The second, and more tentative thing I notice from this data is that at least in suburbia there seems to be a slight shift away from exclusive ideological conservativism to a more pragmatic pedagogical orientation. These mothers have on average the highest level of educational attainment and no doubt have high aspirations for their own children, who are now getting older. Other studies of explicitly Christian curriculum have found that their memory recall/worksheet emphasis works better for young children than for older kids in more sophisticated subjects. Though it’s only suggested by the data, it seems to me one could make a case that Hanna has uncovered a slight move away from ideological conservativism among suburbanites with more education and older children. They are still using Christian curriculum to be sure (at very high rates), but only as one among several influences.
Finally, Hanna notes, fascinatingly, that for all the effort and controversy surrounding the successful bid by homeschoolers to gain access to public school resources in Pennsylvania (a right they secured in 2005), hardly any of them take advantage of it. In 2008 exactly 0% of the urban, 2% of the rural, and 12.5% of suburban homeschoolers partook of things like public school sports, clubs, music groups, and so forth.
While there are things Hanna did not ask her subjects that I wish she had (especially about their religious and political beliefs), and while the demographic data could have been a bit more complete, this article is a tremendous boon to homeschooling research, providing solid empirical evidence for many important claims. All homeschooling researchers are in her debt for this excellent work.
One of the best ever? Seriously?
Take, for example, the Ideologue/pedagogue tables, above – is there really enough data to reach any true conclusions? Why isn’t the Neither data in that table?
“pretty representative sample of Pennsylvania homeschoolers.” – we have no idea if this is true – her data is simply comprised of about 1% of PA homeschoolers.
“fascinatingly, that for all the effort and controversy surrounding the successful bid by homeschoolers to gain access to public school resources in Pennsylvania (a right they secured in 2005), hardly any of them take advantage of it.” – This is “fascinating”? Too little data (again), and hardly surprising if you’ve been a member of a home schooling community.
I do find it strange she tracked (and you wrote about) how many homeschooling MOTHERS went back to college – when you give no reference data as to the percentage of mothers in PA who did the same thing (i.e., were homeschooling mothers more likely to, less likely to, or similar) – and why does this have ANY bearing on ANYTHING? Why, why didn’t she (and you write about) the number of homeschoolers who received a college degree instead?
You know, her article may have some amazing revelations but I just don’t see any mentioned in your blog post here.
Jack,
1. I left the “neither” data out of the ideologue/pedagogue table to save myself time writing it out. Hanna includes it. You’ll note that I did say right before the table that in both years only 5 people chose “neither.” I could have written out the spread, but it doesn’t really matter. The important point is that hardly any of the subjects chose neither.
2. It’s true that this is not a random sample. But given what usually passes for sampling in homeschooling research (convenience samples almost always), this is a very clever method. That it’s only 1% of the overall population is not important–that’s what samples are by definition. What IS important is that her sample was geographically representative, and within each geographic region the ten families chosen WERE randomly generated. Plus since PA requires all homeschoolers to register (though not all do) you get a more representative sample then you would in a state that doesn’t require registration.
3. Maybe you’re not fascinated by the fact that very few homeschoolers take advantage of the public services now made available to them. But I am for many reasons. First, many, many states have just passed or are now considering similar bills (often called “Tebow laws”). It’s pretty much the hottest thing going on right now in homeschooling politics. Second, the lack of interest in these services suggests to me the reach (at least in PA) of homeschoolers with HSLDA-type sensibilities. HSLDA has always had a standoffish approach to these laws, suspicious as they are of any linkage between the homeschooling movement and public education. Maybe you’re right that insiders would not be surprised by all of this, but insiders tend to know more about the homeschoolers in their immediate circle and in their own ideological corner. Hanna’s data here allows us to generalize in ways that anecdotal impressions do not.
4. Of the four criticisms you make this last one is the best. It’s true that she did not compare the rates of homeschooling mothers returning to school with those of the general population. Thus we don’t know whether homeschooling itself had anything to do with their decisions (as I speculatively asserted) or whether it’s just part of the life cycle of a mother whose children are getting older. Regardless, what her study does show is that for at least a significant group of homeschooling mothers being a homeschooler does not prohibit going back to college. Your point that there are many mothers who DON’T go back is also a good one. A follow-up study could try to dig deeper here. What factors lead some homeschooling mothers to go back to school and others not to? I’d love to know the answer.
Is “15-18” a typo or do the last two age categories overlap by two years?
typo. Fixed it. Thanks.
Dr. Gaither,
I purchased your book and it is excellent. I am considering homeschooling as a potential dissertation topic and was wondering what other homeschooling articles/studies would you recommend, in addition to Dr. Hanna’s?
Thank you for your invaluable contributions to this important body of knowledge.
Craig,
You can find my more recent reviews of homeschooling research at this site: icher.org/blog
Thanks,
Milton