What characterizes home schoolers? A national study

What characterizes home schoolers? A national study

Theodore C Wagenaar

The number of children and youth being schooled at home is increasing yearly but there is little literature available on the size and nature of family involved. One of the reasons is that home schoolers resist being involved in such research. It is estimated that more than 300,000 children are now involved, and the number is growing. Home schooled children are very similar to others.

More parents are home schooling their children than ever before. Accurate estimates of the number of home schooled children are hard to obtain. Lines (1991) estimates that between 250,000 and 360,000 children were taught at home in the 1990-1991 school year; Ray (1992a) estimates that 375,000 children were home schooled in 1992 and 450,000 to 800,000 in the 1993-1994 school year. Meighan (1995) puts the figure at over a million. Ray (1992b) also suggests that as many as two percent of the school-age population may be home schooled by the year 2000.

A large literature exists on home schooling. Many sources address the practicalities (e.g., Ballmann 1995). Some address the legalities involved (e.g., Zirkel,1994). Others examine the philosophical and pedagogical backgrounds of home schooling (e.g., Knowles, Muchmore, and Spaulding, 1994). Some address testing problems unique to home schoolers (e.g., Mueller, 1989). The responsibilities of school boards concerns others (e.g., American School Board Journal, 1992). Still others address home schooling in other countries (e.g, Henson, 1996).

Research on home schooling is perhaps the most important area in this literature: what do we know about home schoolers, particularly as systematically compared to others? The state of research on home schoolers, particularly as systematically compared to others? The state of research on home schooling is sporadic. Ray (1988) notes the unavailability of reliable data on such characteristics as parental income of home schoolers. Regarding such variables as family size, Ray (1988) observed that the average home school family “probably” includes more children than the average family does, but offers no data. Writing in his introduction to a special journal issue on home schooling, Knowles (1988) notes that the research literature is small. Almost all studies involve small convenience samples of home schoolers willing to share data. Mayberry et al. (1995) discuss various reasons why home schoolers resist participating in research, especially survey research.

Perhaps the most definitive study is that by Mayberry et al. (1995). They combined various home school registration lists provided by school districts, home school organizations, and networking groups and sent questionnaires to 6,064 families. But they covered only three states and obtained only a 25 percent response rate. We need more data on a national basis that involve random samples to better understand the nature of home schoolers. National and randomly gathered data are especially important given the strongly self selective nature of most home school samples. This study helps fill that gap.


The data come from the National Household Education Survey of 1993, a study sponsored by the National Center for Education Statistics. Data were gathered in 1993 via a random digit dial telephone survey of households, including 4,423 parents of preschool children, 2,126 parents of kindergarten children, 4,277 parents of primary school children, and 62 parents of home school children. The overall response rate was 73.6 percent. Blacks and Hispanics were oversampled; weighting was employed in this analysis to adjust for oversampling and other sampling variations. The home schooled children included only the ages of five to seven. Hence, I compare them only with the 6,117 children five to seven years old in the study. Further details on the study design can be found in National Center for Education Statistics (1994).

Except for the basic demographic variables, multiple classification analysis was used to examine each variable while controlling for socioeconomic status (SES). SES was measured by dividing the 11-category household income item by two and adding it to the highest parental educational level (five categories); SES was divided by two to give the two items a relatively similar metric. Tests of statistical significance are not reported because even small differences are statistically significant with large samples.

The size of the home school group is small (62). At best, the results can only be considered suggestive rather than definitive. Hence, small differences will be considered as non-differences Also, since the focus of the National Household Education Survey was the preschool and primary school experiences of children, home schoolers can only be compared to others on a limited number of variables. The disadvantages of small subgroup size and limited comparison variables are counterbalanced by the use of a large national random sample that includes home schoolers, perhaps the first study of its kind.


The 62 home schooled children in the sample represent approximately 99,000 children home schooled nationwide in 1992, the year the data were collected. Since only children ages five to seven were included, the actual number of home schooled children nationwide is considerably greater. Ray’s (1992) estimate of 375,000 is probably the best so far for 1992.

The results support previous research that home schoolers have more education and higher incomes than do others. Almost 39 percent of home schoolers have a college or graduate/professional degree, compared to 25 percent of the others. About two-thirds of home schoolers have an income over $25,000 compared to SS percent of the others. Eighty percent of home schoolers are white (non-Hispanic) compared to 71 percent of non-home schoolers. This 20 percent minority composition is considerably greater than that reported previously. Even the best study to date (Mayberry et al 1995) shows well under five percent minority composition (due perhaps to the three states selected for that study). The results show relatively high minority involvement in home schooling.

Regional differences are small: home schoolers are somewhat more likely to live in the West (28% vs. 22%) and somewhat less likely to live in the Midwest ( 17% vs. 24%). Negligible differences exist for the percentage of blacks and Hispanics in the respondent’s ZIP code, and the only noticeable (but still small) difference in the percentage of residents in the respondent’s ZIP code below the poverty level is for the highest level (20% or more): 2 percent of home schoolers and 9 percent of others are in this category.

The average age of home schoolers is very similar to other parents’ (33.6 vs. 33.8). There is no noticeable difference in the sex of the child. The remaining differences have been controlled for socioeconomic status because they are often related to SES.

Home school families are larger, as Ray (1988) suggests. The total number of household members in home school families is 5.0, vs. 4.5 for others, primarily because home schooled children have more siblings (1.9 vs. 1.5). Because they have more siblings, home schooled children are less likely to be first- or only-bears (42% vs. 47%).

Turning next to some developmental variables, there is no difference in whether or not the child’s birth weight was over 5.5 pounds. There are negligible differences regarding the presence of a developmental delay, a learning disability, or a serious emotional disturbance. There is no difference in the percentage of children in very good or excellent health. However, home schooled children are less likely to have had a speech impairment (3% vs. 9%); parents of such children are more likely to seek out the professional support services offered by school districts.

Home schooled children are more likely to have married mothers (86% vs. 72%), but there is no substantive difference regarding if the mother was married when the child was born. Home schooled children are just as likely to have fathers in the paid labor force as other children, but are much less likely to have mothers in the paid labor force (39% vs. 58%).

Finally, I turn to some differences in reading ability and in what parents do with their children. Interestingly, home schooled children are somewhat less likely to be able to read story books on their own (53% vs. 61%), but among those who can, home schoolers started reading at an earlier age (5.4 years of age vs. 5.9). Home schooled children are more likely to be read to three times or more a week (85% vs. 73%). Home schooled children are considerably more likely to experience someone in the family doing the following activities with them three or more times a week: tell a story (62% vs. 35%), teach letters, words, or numbers (74% vs. 64%), teach songs or music (56% vs. 29%), do arts and crafts (43% vs. 30%), play with toys or games indoors (98% vs. 69%), play games or sports outdoors (44% vs. 29%), take child on errands (73% vs. 63%), and involve child in household chores (83% vs. 73%). A “community enrichment” index was created by adding the number of the following activities done with the child in the past month: visit a library, attend a play or concert or other live show, visit an art gallery or museum or historical site, visit a zoo or aquarium, attend an event sponsored by a community or religious group. Home schooled children experienced 2.2 of these activities while other children experienced 1.6 enrichment activities.


Three conclusions emerge from this study. First, home schoolers and home schooled children are very similar to others on many characteristics: regional differences, percent poverty in neighborhood, percent minorities in neighborhood, parental age, birth weight, developmental delays, learning disabilities, emotional disturbances, overall health of child, mother’s marital status at child’s birth, and father’s labor force participation. In many respects, home schoolers and their children do not differ dramatically from the larger population.

Second, they do differ on some demographic variables. They have much higher levels of education and income. The education difference may reflect the minimum educational criteria established by local or state education officials for parents who wish to home school their children. Both the income and education differences may reflect the often documented greater concern for educational quality among higher socioeconomic status parents. As Knowles (1988, p. 12) observes, “what makes the home school community different is its intense concern for the educational welfare of children,” something also characteristic of higher social class parents. Hence, sound research on home schoolers must control for socioeconomic status.

The results also disconfirm a previously held conclusion that minority involvement is low. Although somewhat lower than white’s, a fifth of the home schooled children were members of a minority group. Families with home schooled children are larger. Mothers of home schooled children are much more likely to be married and less likely to be in the paid labor force. These findings corroborate many previous findings, such as Mayberry et al. (1995).

Third, the data suggest that home schoolers are, in fact, doing educational and developmental activities with their children. Home schooled children are more likely to be read to and consistently experience more educational and developmental/enrichment activities both inside and outside the home than do other children. Presumably, the other children are making up for some of the difference in preschool or primary school. Good studies on the relative achievement differences of home schooled and other children are difficult to obtain.

In short, home schoolers are in many respects representative of the total population. On issues of social class and educational activities they remain quite different.


American School Board Journal. (1992). Understanding home schooling. 179(9), 26-27+.

Ballman, R.E. (1995). The how and why of home

schooling. New York: Crossway Books.

Henson, C. (1996). Do children have to go to school? Child Education, 79(Mar), 68.

Knowles, G.J. (1988). Introduction: The context of home schooling in the United States. Education and Urban Society, 21, 5-15.

Knowles, G.J., Muchmore, J., & Spaulding, H.W. (1994). Home education as an alternative to institutionalized education. The Educational Forum, SS (Spring), 238-243.

Lines, PA. (1991). Estimating the home schooled population (Working paper )R 91-537). Washington: Office of Educational Research and Improvement, U.S. Department of Education.

Mayberry, M.J., Knowles, G, Ray, B., & Marlow, S. (1995). Home schooling: Parents as educators. Thousand Oaks, CA: Corwin Press.

Meighan, R. (1995). Home-based education effectiveness research and some of its implications. Educational Review, 47(3), 275-287.

Mueller, R.J. (1989). Testing problems of homeschooled children. The Clearinghouse, 62(Mar), 325-326.

National Center for Education Statistics. (1994). National household education survey of 1993: School readiness data file user’s manual. Washington: U.S. Department of Education.

Ray, B.D. (1992a). Marching to the beat of their own drum: A profile of home education research. Salem, OR: National Home Education Research Institute.

. (1992b). On the propriety of education professionals instructing home educators. Paper presented at the annual meeting of the American Educational Research Association, San Francisco.

. (1988). Home schools: A synthesis of research on characteristics and learner outcomes. Education and Urban Society, 21(1), 16-31.

Zirkel, P.A. (1994). Home sweet … school. Phi Delta Kappan, 76 (Dec), 332-333.

THEODORE C. WAGENAAR Miami University Oxford, OH 45056

Copyright Project Innovation Spring 1997

Provided by ProQuest Information and Learning Company. All rights Reserved