Self-Concept in Home-Schooling Children

Prepared by John Wesley Taylor V, Ph.D.

The home-schooling movement appears to be in an upward trend in America. It has arisen to the notice of parents, educators, legislatures, and judicial systems. Legitimate questions have been raised regarding the home school. Many of the answers, however, have been based upon hearsay, intuition, folklore, or anecdotal accounts.
    This national study sought to provide an empirical base upon which to formulate viable decisions regarding home-schooling children. It sought to address the issue of socialization and self-concept in home-schooling children--perhaps the most-frequently asked question regarding the home school.

    The study considered home schoolers in grades four through twelve. Instruments included a demographic form and the Piers-Harris Children's Self-Concept Scale. The randomized sample, drawn from the mailing lists of the two largest national home-schooling agencies--Holt Associates, Inc. and Hewitt Research Foundation, yielded 224 qualified participants.

 Conclusions include the following:

  1. The self-concept of the home-schooling children was significantly higher (p<.001) than that of the conventionally schooled population on the global scale and all six subscales of the PHSCS. On the global scale, half of the home schoolers scored at or above the 91st percentile. This condition may be due to higher achievement and mastery levels, independent study characteristics, or one-on-one tutoring situations in the home-school environment. It could also be due, perhaps, to higher levels of parental interest and communication, peer independence, a sense of responsibility, and lowered anxiety levels.
  2. Insofar as self-concept is a reflector of socialization, it would appear that few home-schooling children are socially deprived. Critics of the home school should not urge self-concept and socialization rationales. These factors apparently favor home schoolers over the conventionally schooled population.
  3. The self-concept of the home-schooling children decreases significantly (p<.01) as age and grade level rise. This, however, is not likely due to increasing number of years of home schooling, as this factor had a significantly positive effect (p<.001) when a part of the best predictive model for self-concept. It could be due to a higher age and grade level at that time when a child entered the home-schooling environment from that of the conventional school.
  4. The factors of gender, number of siblings, locale of residence, prior conventional schooling, educational level of home-school operators, and geographical region were not significantly related to the self-concept of home-schooling children neither When considered in isolation nor as a part of the best predictive model of self-concept.
  5. While not significantly related to self-concept when in isolation, the factors of the number of years of home schooling and the beginning school age did become, when in the presence of certain other demographic variables, significant predictors (p<.001) in the best predictive model of self-concept in home-schooling children.
  6. Higher socioeconomic status and an increase in the total number of home-schooling children in a family, within the limits of one to seven children examined by this study, is significantly related (p<.05) to a more positive level of self-concept in home schoolers.
  7. The best predictive model of self-concept in home-schooling children (p<.001) is related to lower grade-equivalence, higher years of home schooling, higher socioeconomic status, higher number of home-schooling children in the family, and higher beginning school age. The model is statistically stable and accounts for over 12 percent of the variance in the self-concept.
  8. Home schoolers are apparently concentrated towards lower grade levels and tend to commence formal instruction at a somewhat later age than the national average.
  9. There is an approximately balanced distribution of home schoolers in terms of gender and geographical region.
  10. It appears that home-schooling families frequently have more children than the national average and usually have more than a single child in the home school.
  11. Very few of the children are in their first year of home schooling and most have previously attended a conventional school.
  12. The educational level and socioeconomic status attained by home-school operators seems to be considerably higher and their locale more rural than that of the comparable general population.


Inasmuch as the literature indicates a strong, consistent relationship between self-concept and academic achievement, I would suggest that a study evaluating the academic achievement of home-schooling children could find them to be significantly above average. I would also recommend that the conventional schools and the home schools seek to cooperate with each other so as to maximize the contributions of each.

P.S.: A special thank-you to each of you who helped to make this study become a reality. I believe that the findings will be significant for many individuals nationwide.

Complete copies of the study (approx. 330 pages) in book form are available from University Microfilms International, 300 N. Zeeb Rd., Ann Arbor, MI 48106. If you would like to request me to speak regarding the home school--from the perspectives of research, education, and personal experience as a home schooler for 12 years, please direct your communication to John Wesley Taylor V, Dept. of Curriculum and Instruction, Andrews University, Berrien Springs, MI 49104. Phone: (616) 471-3514.



Then we find 3.6% in the first 25 percentiles, 6.7% in the second 25 percentiles, 12.0% in the third, and 77.7% in the highest 25 percentiles. (It must be remembered that these figures address percentiles in which the average child who attends the conventional school places at the 50th percentile.) Stated another way, while half of the conventionally schooled children scored at or below the 50th percentile, only 10.3% of the home-schooling children did so. What does this mean? The self-concept of home-schooling children is significantly higher (and very much so statistically) than that of children attending the conventional school. This of course has important implications in the areas of academic achievement and socialization, to mention only two. These two areas have been found to parallel self-concept very closely. Regarding socialization, it appears that very few home-schooling children are socially deprived. Critics who speak out against home schooling on the basis of social deprivation are actually addressing an area which favors home schoolers. Apparently, the research data indicates that it is the conventionally schooled child who is actually deprived.