Paul Kitchen
School of Education
Andrews University
Berrien Springs, MI 49104

Keywords: Homeschooling, home schooling, home education, socialization.

IN THE PAST DECADE, what had begun as a trickle of parents pulling their children out of traditional schools and teaching them at home, has become a steady stream. Perhaps as many as hundreds of thousands of children are being schooled at home (Lines, 1987). Based on percentage projections from Lines, the Home School Legal Defense Association estimates that 474,000 children are presently home schooled (Farris, 1990). Why is this? The reasons for choosing to educate children at home are as varied as the families and children involved (Lines, 1987).
Van Galen (1988) divided the parents who chose to home school into two groups, Ideologues and Pedagogues.1 Unfortunately, she only studied sixteen families to draw this conclusion. For this reason I believe that her classifications are a little too simplistic. Parents choose to home school for many reasons: educational standards, moral standards, value orientation, religious convictions, and philosophical issues. One reason often cited by parents for taking their children out of the school system is the perceived negative socialization (negative peer pressure) that their children are exposed to in conventional school settings (Van Galen, 1988). It is this issue, the socialization effect of different school environments, which will be the focus of this paper.
The academic excellence of home schooled children has been repeatedly demonstrated (Alaska Department of Education 1985, 1986; Hewitt Research Foundation 1985, 1986; Ray, 1991). Most research shows that home schooled children consistently score at or above the 50th percentile on standardized achievement tests, with more than half scoring at the 70th to 80th percentile.
The fact is, most research into home schooling has centered on the academic outcomes (Wright, l988). Yet school is not just academics. One of the main goals of education is to prepare children to function in society. Many critics claim that children schooled at home do not receive the benefit of the socializing influence of conventional schools. Researchers have not spent much time, however, looking at this aspect of the home schoolers’ education versus the norm or even if falling within the “norm” is necessarily good.
Moore, a principal figure in the resurgence of home schooling in North America, and others (Moore, Lorenz, Willey, More, & DuPreez, 1975) have said concerning this issue, “The socialization of children becomes a problem in a society where traditional values are questioned. When adults, and parents in particular, are unsure of their responsibilities, children may be left to socialize each other, to form their values from their own peer group with consequent insecurity and negative self esteem” (p. 65). A kind of Lord of the Flies mentality sets in upon children who are left to socialize themselves, rather than to be socialized by the example of responsible adults. Many parents agree with Moore, and have concluded that this negative self-esteem is a direct result of the overcrowded, artificial school environment (Dobson, 1984).
In Golding’s classic novel (1962), Lord of the Flies, children, by circumstances of war, are stranded on an island with no adults. Slowly but surely, the restraining influences of society are set aside by the children and mob rule becomes the norm. Children quickly learn the importance of identifying with the strong, and conforming to the norm, no matter how abnormal the norm is. It becomes too dangerous to stand alone, or be weak. This is how many view the average school yard. One has only to think back to his own school days, or talk to any average high school student to see the amazing similarity between the Lord of the Flies and the socialization process in many conventional school environments. If this is so, it could be speculated that other components of socialization would possibly be compromised Such as academic success or family relationships.
In spite of this, educators continue to argue that the conventional educational system better than home education in preparing children for life. The purpose of this study was to answer the question, Are home schooled children advantaged or disadvantaged in their social adaptation/self-esteem by being educated at home? (The connection between social adaptation and self-esteem will be discussed in the review of literature.)
The hypothesis of the study was: Children who have been educated at home will have higher self-esteem than those who are educated in more conventional ways. To test this hypothesis the following variables will be taken into consideration:
1. Type of School: home, private (parochial), public.
2. Self-esteem: As measured by the Self‑Esteem Index, (PRO-ED, Inc., 1991).
3. Demographic Survey: Parental income, parental education, and religious affiliation.

Review of Literature

Home schooling is not a new idea. In fact, throughout history, the home has been the predominant place of learning. For example, the Bible states quite clearly that it is the parents’ responsibility to instruct their children (Genesis 18:19, Deuteronomy 6:6‑9, Ephesians 6:4).
Even in America, the majority of children were educated at home until about World War I. Compulsory education laws were introduced in some states around the late l9th century, and by 1918, every state had some form of compulsory school attendance legislation. It is interesting to note that public schools were begun for the purpose of socializing its students. According to Knowles (1989), “…public schooling sought to remove the stamp of individualized ethnic orientations that immigrant family‑related learning environments promoted.” In other words, the goal of public schools was originally to “Americanize” the children.
Socialization has continued to be one of the functions of the educational institution. This is precisely the problem some parents are having with the public school. The perceived negative social pressures on children found in the schools, i# the reason many parents have chosen home schooling (Gustavsen, 1980). This is not limited to negative peer pressure, it encompasses the new social agenda of moral relativism implicit in value‑free education (Dobson & Bauer, 1990, p. 31).
Through my review of literature, I have noticed a trend develop. Almost always, when home schooling is challenged
legally, the debate revolves around the academic quality of the education of home schooled children, that is, teacher certification of the parent. However, the individual who has doubts about the wisdom of home schooling is generally concerned about the socialization issues.
The ground breaking study of socialization in home schooled children was undertaken by Taylor (1986). Taylor compared the self‑concept of 224 home schooled children using the Piers Harris Children’s Self‑Concept Scale (PHSCS). Taylor found that the self‑concept of the home schooled was significantly higher than that of conventionally schooled (p<.001). He stated that, “…Insofar as self‑concept is a reflector of socialization…the findings of this study would suggest that few home‑schooling children are socially deprived” (p. 160‑161). What was lacking in Taylor’s study, however, was that his subjects had a diversity of prior educational backgrounds; not all were solely home educated.
The question arises, however: Is self‑concept a reflector of socialization? Purkey (1970) reported that the way children react to people, tasks and roles is often consistent with their view of self. Later, DeFrancesco and Taylor (1985) conducted a study on self-concept in middle school students where they found, “…what a person believes about himself affects what he does, what he sees and hears, and his capacity to cope with his environment” (p. 99).
The views of these researchers is congruent with Cooley’s (1902) theory of the Looking Glass Self. This theory has been the cornerstone concept in the development of research into socialization. It also plays an important role in the sociological conceptualization of self-concept. Briefly stated, the Looking Glass Self refers to the idea that we look to significant others in our lives in order to understand how they see us, and then in turn we build our self-concept and self-esteem from the reflection of ourselves that we see from the.
Gecas and Swalbe (1983) followed up on this Looking Glass Self theory and strengthened the tie between self-concept and socialization. They contend that more than the passivity of the looking glass self, it is the active involvement in society which molds our self image.
…and recognizing that self evaluations are also formed through the experience of producing effects upon the world, we can grasp another dimension of the relationship between self evaluations and social structure. By focusing upon efficacious action as a source of self esteem, we can see how social-structural conditions can shape possibilities for individuals to act efficaciously and to experience this in a way that enhances feelings of self esteem. (p. 86)
It would appear that the ability to successfully cope with one’s environment is intricately connected with one’s self-esteem. As one interacts with society, the experience builds or tears down the self-esteem, which in turn influences the way one will interact with society in the future.


For the purpose of this study, an availability sample was used. Three groups of children were asked to respond. The groups were home, private, and public schooled children. Attempts were made to keep each group as pure as possible (i.e., that the children in each group should have only received one type of education). To be included in the sample, the child had to be of the same type of school environment in grades 1 through 8. This was to ensure that any difference in self‑esteem could be more clearly correlated with the school environment.
Grades 6 through 8 were selected for a variety of reasons. First, by the 8th grade many home school parents choose to put their children back into a traditional school. This is probably a result of the more complex academic needs of the child. Thus, it becomes more difficult as the children get older, to find willing participants. Second, 6th through 8th grade is 11 through 14 years of age. This range covers prepubescence and pubescence. These children are dealing with the changes of adolescence, and for this        reason it becomes more important that a positive self-esteem is present. The Self Esteem Index, a standardized measure, was used to measure the children’s level of self-esteem. The SEI has been well correlated with other scales such as the Piers-Harris Children’s Self Concept Scale, Revised (PRO-ED, Inc., 1991). The self reported reliability coefficients were acceptable for this type of instrument (at greater than .80).
The subjects were chosen in a variety of ways. The home school group was chosen through personal contacts and assistance from home school organizations. Two private schools, and one local public school provided mailing lists of children that had only been in that particular type of environment.
Approximately 50 packets were mailed to each group. Each packet contained a cover letter, informed consent forms, a demographic survey, an instruction sheet, Self‑Esteem Index Response booklet, and a self‑addressed stamped envelope.
Each subject was instructed to read all the information included before filling out the response booklet. The parents were asked to read the questions in the response booklet before the child filled it out. After the child completed the form, the parents were instructed not to read the answers and to sign the demographic form stating that they had not read the answers. This wag to ensure that the child would answer the questions as honestly as possible.
In order to minimize bias between the groups, the tests were
all administered at home, rather than the home educated at home
and the conventionally educated in their environments. It was felt that the possibility of a parent reading the responses was still high enough to factor in, so at least this way all children would be responding under the same conditions.
As the responses were returned, they were scored and evaluated according to the SEI examiners manual (PRO‑ED Inc., l991).


Due to the small number of responses from the public school group (Table 1), the private and public school results were combined to form one category called Conventional schooled children.
–This table is missing in this web page presentation; you may find the PDF format elsewhere on this web site with the table included–

Table 1. Response rates from different categories.

The SEI is broken down into four subscales: Perception of Personal Security, Perception of Peer Popularity, Perception of Academic Competence, and Perception of Familial Acceptance. The results, depicted as percentages, of those who scored above average in each category are shown in Figure 1.
The statistical procedures used to evaluate the following data were Chi Square and Gamma. Chi Square (P2) is a nonparametric statistical method used to determine the relationships between variables. All the p values are based on Chi Square tabulations unless otherwise stated. One degree of freedom (df) was used for all calculations. Gamma (() is a nonparametric correlation coefficient which is used to indicate the strength of the correlation between variables. The range of gamma is 0 to 1. Since the data are, for the most part, ordinal level data, these procedures are appropriate for statistical analysis. In three categories, Personal Security, Academic Competence and Familial Acceptance, the home schooled group had higher percentages of children that scored above average as compared to the conventionally schooled children. The conventionally schooled children had 9% more children score higher on the Peer Popularity scale than the home schoolers. The overall Self Esteem Quotient (SEQ) showed much the same pattern as the subscales.
Notable statistics that were drawn from the SEI scores included:
1. Academic Competence proved to be significant (p<.05), and had a gamma of .50 (Table 2). This is consistent with the belief that people gain self-esteem from doing their work well. If we
–This figure is missing in this web page presentation; you may find the PDF format elsewhere on this web site with the figure included–

Figure 1. Percentage of students scoring above the 75th percentile in each subscale. (Note: The SEQ is a compilation of all the subcategories combined.)

see school or learning as the job of children, it would be expected that it will greatly effect their self-esteem. Thus, if a child’s perception is that he or she is doing well in his or her school work, they would naturally be more confident.
2. There was a ‑0.23 correlation in the Peer Popularity subscale showing an inverse relationship between self-esteem and peer popularity. This indicates that with a rise in Peer popularity there is a negative effect on overall self-esteem. It is only a moderate correlation, but certainly one that can’t be ignored.
In addition to the SEI, the demographic survey was studied. The results of three factors were compiled, they were: Parental education, religious affiliation, and income. Parental education and religious affiliation had apparently no relationship to the child’s self-esteem. But there was a ‑.50 correlation and near significance with p<.06, between parental income and self‑esteem in

–This table is missing in this web page presentation; you may find the PDF format elsewhere on this web site with the table included–

Table 2. Relationship between self-esteem and type of school environment.

the home educated group. This shows an inverse relationship. This same relationship was not found in the conventionally schooled group. Any explanation for this finding has to explain both why there would be an inverse relationship, and also why it only appears in the home school population.
The home school population, by necessity, would require one parent at home, thus limiting the earning potential of most families. The implication of this result is that having a parent at home improves the self-esteem of children. So why wasn’t the same relationship found in conventionally schooled children? Even if a parent is at home, for most of the day, the conventionally schooled child isn’t. As a result, conventionally schooled children have to look outside the family for esteem building influences. This explanation could support why home schooled children scored higher on Perception of Family Acceptance and why conventionally schooled children scored higher on Perception of Peer Popularity.


The major weakness of this research is that the public schooled group only had a return rate of 16%. This was in spite of the fact that the mail‑outs to this group contained a cover letter by the school principal encouraging them to participate. A return rate of only 16% can hardly be considered a representative sample. As a result of this, further analysis was done. The public school group was dropped from the data and the differences between the home school and the private school populations were considered. In almost every case, the strength of correlation between factors increased (Table 3). However, the reduction of an already small sample made it impossible to achieve statistical significance.
However, using the binomial formula, the probability of the number of children scoring in the top 25% was computed. To use one example, the probability that the number of home school children who scored above average on the SEQ would do so by chance alone was p=0.0005; and for the private school group, the probability was p=0.06. This is not a traditional method for computing statistical significance, since it looks only at the results within groups and not at the differences between them, therefore this is presented merely as food for thought.
–This table is missing in this web page presentation; you may find the PDF format elsewhere on this web site with the table included–

Table 3. Relationship between self-esteem and type of school environment.

Implications for Further Research

Because the sample for this study was an availability sample, it is not possible to generalize these findings too far from the sample itself. However, the fact that home schoolers scored higher on almost every subscale suggests that a large, pure2 sample of home schooled versus conventionally schooled children might result in statistical significance that would lead to rejection of the null hypothesis.
A focused examination into the Lord of the Flies mentality and its effects on self-esteem could produce some interesting results, as could the negative correlation between parental income and home schooled children’s self-esteem. This is an issue which begs for subsequent investigation. In almost every measure, the home schooled children scored higher on the Self‑Esteem Index, which gives every indication that my original hypothesis is supportable. However, this limited sample size renders it unable to demonstrate statistical significance, except in the academic competence sub‑scale. As to the issue of whether a home schooled child is being adequately socialized, it appears as reflected by the overall SEQ scores, that the home schooled child is most likely doing well, and in some ways may even be advantaged.


1. Ideologues are those who object to conventional schools for ideological reasons. They often have specific beliefs, values, and skills which they want their children to learn.
Pedagogues are those who criticize conventional schools not because of what they teach, or don’t teach, but rather, pedagogues feel that the conventional system is inept at teaching children.

2. Pure is defined here to be children that have been in only one type of school environment.


Alaska Department of Education. (1985). SRA Survey of Basic Skills, Alaska Statewide Assessment, Spring of 1985. Juneau, AK: Author.
Alaska Department of Education. (1986). Results from 1981 CAT (For CCS). Juneau, AK: Author.
Cooley, C. H. (1964 [1902]). Human nature and the social order. New York, NY: Schocken Books.
DeFrancesco, J. J., & Taylor, J. (1985). Dimensions of self-concept in primary and middle school learning disabled and non-disabled students. Child Study Journal, 15 (2), 99-104.
Dobson, J. C. (1984, August/September). Groups can damage young children’s self‑concept. The Teaching Home, 2(4), 11.
Dobson, J. C., & Bauer, G. L. (1990). Children at Risk: The battle for the hearts and minds of our kids. Dallas, TX: Word.
Farris, Michael. (December, l990). Home School Court Report, 8. Paeonian Springs, VA: Home School Legal Defense Association.
Gecas, V., & Swalbe, M. L. (1983). Beyond the Looking Glass Self: Social structure and efficacy based self-esteem. Social Psychology Quarterly, 46(2), 77-88.
Golding, William. (1962). The Lord of the Flies. New York, NY: Coward‑McCann.
Gustavsen, G. A. (1980). Selected characteristics of home schools and parents who operate them. Unpublished Doctoral Dissertation, Andrews University, Berrien Springs, MI.
Hewitt Research Foundation. (1985). North Dakota trial results pending. Parent Educator and Family Report, 3(2), 5.
Hewitt Research Foundation. (1986). Study of homeschoolers taken to court. Parent Educator and Family Report, 4(1), 2.
Knowles, J. Gary. (1989). Cooperating with home school parents: A new agenda for public schools? Urban Education, 23(4), 392-410.
Lines, Patricia. (1987, March). Overview of home instruction. Phi Delta Kappan, 68, 510‑517.
Moore, R. S., Lorenz, M., Willey, T., More, D., & DuPreez, I. (1975). Influences on learning in early childhood: A literature review. Berrien Springs, MI: Hewitt Research Center.
PRO‑ED, Inc. (1991). Self Esteem Index: Examiners manual. Austin, TX: Author.
Purkey, W. (1970). Self concept and school achievement. Englewood Cliffs, NJ: Prentice Hall.
Ray, Brian D. (1991). A nationwide study of home education: Family characteristics, legal matters, and student achievement. Seattle, WA: National Home Education Research Institute. (currently in Salem, OR)
Taylor, John W., IV. (1986). Self concept in home-schooling children. Unpublished doctoral dissertation, Andrews University, Berrien Springs, MI.
Van Galen, Jane A. (1988, November). Ideology, curriculum and pedagogy in home education. Education and Urban Society, 21(1), 52‑68.
Wright, Cheryl. (1988, November). Home school research: Critique and suggestions for the future. Education and Urban Society, 21(1), 96‑113.

Author’s Notes

My thanks go to the principals of the schools involved in this research, for their cooperation in providing subjects. I also want to express my appreciation for my wife who did so much of the scoring. My thanks also go to Dr. McBride for his guidance and help in choosing statistical approaches.

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply