SOCIALIZATION OF HOME SCHOOLED CHILDREN: A SELF-CONCEPT STUDY

Steven W. Kelley
Doris Hancock Elementary School
1661 Lindell Road
Las Vegas, Nevada 89102

Keywords: Homeschooling, home schooling, home education, socialization
Every child will undergo social experiences as the price of admission to civilized society (McNeil, 1968). Consequently, socialization is the home schooling concern most frequently mentioned by parents, educators, legislative assemblies, and judicial systemsCinferring that home schooled children need to be around other children in order to be socialized (Ray, 1988; State vs. Riddle, 1981; Taylor, 1986).
Taylor (1986) attempted to address the issue of socialization by analyzing the relationship between home schooling and the self-concept of children in grades four through twelve. Taylor concluded that insofar as self-concept is a reflector of socialization, few home schooled children are socially deprived.

The Problem
A number of researchers have pointed out weaknesses in Taylor’s study. One weakness, for example, is that of self-selection. Taylor’s sample was randomly drawn from mailing lists of parents who subscribe to home school newsletters. Wright (1988) maintained that “the results only reflect findings from particular groups and are not generalizable to large populations of home school families” (p. 98). It should be noted, however, that “self-selection” or sample bias is common in home school research studies. Those subjects selected for study are usually extracted by the researcher from a group or list of names. Therefore, researchers commonly use mailing lists of families who subscribe to home school newsletters (e.g., Taylor, 1986) and lists of families who are members of home school organizations, churches, or who are registered with some type of state agency (e.g., Gustavsen, 1981; Hedin, 1990; Wartes, 1988).
Another weakness in Taylor’s study (1986) was the administration of the self-concept instrument. Hedin (1990) warned that caution should be exercised in “interpreting this particular study because administration was carried out by the parents of the children. This could cause interference, particularly in the direction of more desirable responses” (p. 68).
Although Taylor’s study (1986) provided some empirical evidence that home schooled children are socialized at least as well as children who attend conventional schools, research on the social development of home schooled children is scarce (Lines, 1987). Additional research, however, might provide information upon which parents, educators, legislators and the courts can base decisions regarding home schooling. With the need for further research in mind, I examined the self-concept of home schooled children in 14 suburban Los Angeles cities (Kelley, 1991). The study focused on the self-concept of home schooled children in grades two through ten.
The purposes of the study were to:
1. Address the “bias” issue with regard to parents administering the self-concept scale in Taylor’s 1986 study.
2. Examine the relationship between home schooling and the self-concept of home schooled children in suburban Los Angeles in grades two through ten.
3. Examine the differences between self-concept in home schooled children in suburban Los Angeles and the norms of self-concept in conventionally schooled children in grades four through ten. Grades two and three, however, could not be compared to conventionally schooled children because normative data for these grades are not available.
4. Examine the relationship between the self-concept of home schooled children in suburban Los Angeles and various demographic factors. These factors included the following independent variables: age, grade-equivalence, gender, number of siblings, number of years in home school, socioeconomic status, prior attendance at a public school, number of children home schooled in the family, and operator educational level.

Review of the Literature

The nature of this study prompted three questions to be addressed in the review of the literature: (a) What is socialization? (b) Are home schooled children socially deprived? and (c) Is self-concept a reflector of socialization?

Socialization
It is difficult to succinctly define the term “socialization.” Brim and Wheeler (1966) explained socialization as the “process by which persons acquire the knowledge, skills, and dispositions that make them more or less able members of their society” (p. 3). McNeil (1968) defined socialization as the process by which the child learns the ways of society and how best to function as part of it. McNeil maintained that socialization is a child learning how he or she ought to feel and behave and what he or she ought to value. With respect to children, Elkind (1988) noted that socialization occurs largely in the home because the family is a school of human relations in which “children learn how to live within society” (p. 121).
A number of individuals have contended that children should be socialized in public schools. Brim and Wheeler (1966), for example, maintained that there are certain social relationships whose primary function in society is to train society’s members. In the teacher-pupil relationship, the main purpose is to train the child so that he or she may become a socially suitable member of society. Brim and Wheeler noted that in such a relationship the role of a child is that of a person being socialized. “Thus socialization processes are not restricted to those that occur within the intimate environs of the family or other closely knit networks. Increasingly, they are a function of large-scale bureaucratic organizations” (p. 53). Brim and Wheeler contended that schools are “familiar examples of what might be called developmental socialization systems, where the formal purpose is the training, education, or more generally the further socialization of individuals passing through” (p.68).
Both parents and educators are concerned that children not only need to be around other children to be socialized, but children must also be in school to learn the ways of society and how best to function as part of it. McNeil (1968), of the University of Michigan, for example, believed that the school is a formalized mechanism for introducing the child into the society. McNeil, stated that “The role of the school as a major agent of socialization cannot be overemphasized. It is a ten or twelve year rehearsal in how the child is expected to behave when he grows up” (p. 38). McNeil concluded that in the social development of the child, no experience is more crucial to form the shape of his or her adult life than the school.
While some individuals argue that the social development of children should be experienced in the public schools, home school advocates maintain that if society wants children to become outgoing, altruistic adults, then children must “avoid their regular mixing with their peers on a group basis as in preschool or kindergarten or primary school until they are at least eight or ten” (Moore & Moore, 1981, p. 38). In a report to the Free Congress Research and Education Foundation in Washington, DC, Moore (1984) explained socialization in these words:
Children are not only better taught at home than at school, but also better socialized by parental example and sharing than by other little children….Contrary to common beliefs, little children are not best socialized by other kids; …socialization is not neutral. It tends to be either positive or negative.
Positive or altruistic and principled sociability is firmly linked with the family–with the quantity and quality of self-worth. This is in turn dependent largely on the track of values and experiences provided by the family at least until the child can reason consistently. In other words, the child who works and eats and plays and has his rest and is read to daily, more with his parents than with his peers, senses that he is part of the family corporation–needed, wanted, and depended upon. He is the one who has a sense of self-worth. And when he does enter school, preferably not before 8 to 10, he usually becomes a social leader. He knows where he is going, is independent in values and skills.
Negative, me-first sociability is born from more peer group association and fewer meaningful parental contacts and responsibility influence generally brings an indifference to family values which defy parents’ correction. The child does not yet consistently understand the “why” of parental demands when his peers replace his parents as his models because he is with them more. (pp. 242-243)

According to Moore and Moore (1982), studies have shown that negative socialization brings loss of self-worth and parental respect.

Social Deprivation
The concern most often raised by home school critics is that home school children may become socially deprived (State vs. Riddle, 1981; and Taylor, 1986). Moreover, public educators argue that home schooling necessarily limits students’ exposure to their diverse society (Roach, 1988). Wilson (1988), however, noted that home schooled children have ample opportunities for socialization. It is normal for home school children to be involved in organized classes, youth groups, sports organizations, and religious education activities. Often home schooled children participate in organized field trips and study groups with others learning at home.
Literature also indicated that home schooled children tend to develop social skills and leadership abilities superior to their conventionally schooled peers. Common and MacMullen (1986), for example, noted that, “Substantiated findings have pointed to higher levels of leadership, motivation, and social-emotional development” among home schooled children (p. 6).
Montgomery (1989) studied home schooled youth ages 10 to 21. She focused on the extent to which home schooled youth were experiencing conditions that encourage leadership. Montgomery found that home schooled youth are actively involved in social and group activities. She concluded that home schooling may foster leadership at least as well as conventional schooling.
Parents also maintain that leadership qualities are a benefit of home schooling. Charvoz (1988), for example, interviewed a number of home school practitioners. According to Charvoz, in an interview with Deborah Recksiek, Deborah noted that her children often take positions of leadership among their peers. Deborah believed that the peer leadership in her children was an indication that self-esteem enhances socialization.
Taylor’s self-concept study provided empirical evidence that there is a significant difference in the self-concept of home schooled children as compared to conventionally schooled children. Taylor measured the self-concept of 224 home schooled children and found that (a) 77.7% scored in the top quartile on the PHCSCS, with one-half of the home schooled children above the 91st percentile; (b) the longer children are taught in the home, the higher their self-concept; and (c) the child’s self-concept is unrelated to the parents’ educational levels.
In a subsequent study, Hedin (1990) studied self-concept in home schooled, private schooled, and public schooled children in grades four through six who attend Texas Baptist churches. The mean total scores for the 37 home schooled children in her sample was 61.75. Hedin’s study showed home schooled children scoring higher on the PHCSCS than the mean total scores of 51.84 for the conventionally schooled sample. Hedin’s study provided more empirical evidence that home schooled children are not socially deprived insofar as self-concept is a reflector of socialization.

Self-Concept
Self-concept is often equated with the terms self-image, self-worth, and self-esteem (Moore & Moore, 1988; Schimmels, 1989). Some researchers, however, have differentiated between self-concept and self-esteem suggesting that self-esteem is perhaps more egocentric (Moore & Moore, 1988). Beane, Lipka, and Ludewig (1980), for example, referred to self-concept as “the description we hold of ourselves based on the roles we play and personal attributes we believe we possess,” whereas self-esteem refers to the “level of satisfaction we attach to that description or parts of it” (p. 84).
Hilgard, Atkinson, and Atkinson (1979) explained self-concept as “the composite of ideas, feelings, and attitudes people have about themselves” (p. 605) while Woolfolk (1987) defined self-concept as “how people view themselves physically, emotionally, socially, and academically; all self-perceptions, taken together” (p. 588). Woolfolk also noted that self-concept could be considered to be “our attempt to explain ourselves to ourselves, to build a scheme (in Piaget’s terms) that organizes our impressions, feelings, and attitudes about ourselves” (p. 104). Purkey and Novak (1984) found socialization to be a key ingredient in the development of self-concept and described self-concept as a “continuously active system of subjective beliefs about personal existence” (p. 29). Combs (1982) viewed self-concept as the phenomenal self which he described as “our experienced self, the person one believes he or she is” (p. 19). Combs also described the phenomenal self-concept in terms of what an individual perceives others to think of him or her, such as “People don’t like me much,” and what an individual would like to be, such as “I wish I was a better father to my children.”
According to Combs (1982), people who see themselves in positive ways, as liked, needed, and persons of integrity, tend to behave accordingly and, as a consequence, get along well in our society. Combs believed that self-concepts are acquired from life experiences, particularly from interactions with significant others. He maintained that self-concepts are learned beliefs about self and the “most important ways in which people learn their self-concepts is through some sort of personal experience” (p. 163). Based on a synthesis of the definitions stated above, it would appear that a person’s self-concept largely reflects his or her socialization.
Purkey (1970) noted that many of the problems which people experience in life are “closely connected with the ways they see themselves and the world in which they live” (p. 2). Purkey and Novak (1984) contended that no one is born with a self-concept. In fact, the development of self-concept is a lifelong process, and the life experiences of the developing person continually change the self-concept. Moreover, “the ingredients of self-concepts are primarily social, obtained through countless interactions with persons, places, policies, and programs” (p. 27). McDonald (1980) would agree with these findings. According to McDonald, while a multitude of theoretical assumptions exist, the view that self-concept is learned through interaction with physical and social surroundings appears consistently in the literature.
The importance of positive self-concept is certainly noted among home school proponents. Kink (1983), for example, studied home school parents in northern California and found that positive self-concept was the home school benefit most frequently mentioned. Often home schooling parents are concerned that the pressures of conventional schooling are likely to have an adverse effect on the self-concept. Consequently, a number of studies indicate that conventional schooling tends to have a negative effect on the self-concept. Felker (1974), for example, studied children from a sample of suburban, middle-class families and found a “steady downward trend of self-concept as the child meets the pressures of the early school years” (p. 63). This trend continued until grade four, when self-concepts began to improve slightly. At grade seven, however, the self-concept of girls experienced another downward trend. Felker also noted that children who had low self-concept scores in their early school years showed no improvement at grade four but continued to have negative self-concepts. Moreover, Goodlad (1984) found that students’ self-perceptions dropped off with upward progression through the school years. About 27% of those children in his elementary school sample were not satisfied with how they were doing in school. The percentage climbed to 34% at the junior high level and to 43% at the senior high level.
While research indicated that there has been a steady decline in the self-concept of children who attend conventional schools, no reports were found that would suggest low self-concept measures in home schooled children. It seems rather ironic that “while those wary of home-schooling worry about the children’s socialization, the parents worry about the quality of socialization that takes place in the traditional schools” (Rakestraw & Rakestraw, 1990, p. 74).

Methodology

The study was designed to investigate self-concept in home schooled children. The principal variable measured was self-concept. The study was empirical in nature, directed toward measuring the self-concept of home schooled children and determining the relationship between self-concept and nine independent variables. These variables included age, grade-equivalence, gender, number of siblings, number of years in home school, socioeconomic status, prior attendance at a conventional school, number of children home schooling in the family, and operator educational level (cf., Taylor, 1986).
Two primary groups were considered for this study. The first group contained those children who are home schooled in suburban Los Angeles. The second group, identified as the normative group, included a sample of children who were conventionally schooled.
In addition to the two primary groups, two specialized groups were examined within the home school group. The first group included those children in grade-equivalents 2 and 3. A second group was examined which included home schooled and conventionally schooled children in grades 4 through 10 (Those families with home schooled children in grades 11 and 12 did not respond).
The PHCSCS is intended for use with children and youth, ages 8 to 18 years and was standardized for use above the third grade level because of doubts about the stability of self-concept in younger children below age seven (Piers, 1984). The PHCSCS, however, has been used experimentally with younger age groups (Kugle & Clements, 1981; Kugle, Clements & Powell, 1983). Since the PHCSCS was standardized for use above the third grade level, the self-concept measures of those home schooled children who were in the second or third grade were grouped together and were not used to compare differences between the self-concept of home schooled and conventionally schooled children.

Limitations

The study was limited to 67 home schooled children in 14 suburban Los Angeles cities. The study focused on self-concept of these children and was not intended to identify a cause of any observed difference. It should be noted that the norms for the PHCSCS presented a limitation in that the normative group was based on 1966 data from one Pennsylvania school district. Nevertheless, studies indicate that the normative group is somewhat generalizable to more diverse populations. Piers (1984), for example, listed a number of studies that used the PHCSCS to examine self-concept in “normal” school children. The total mean score of the accumulated samples (total N=3,692) was 55.2 with a standard deviation of 12.6, compared to the PHCSCS standardization (N=1,183) with a mean of 51.8 and a standard deviation of 13.8 (see Beck, Roblee & Hanson, 1982; Black, 1974; Kugle, 1980; Piers, 1977; Silverman & Zigmond, 1983; Wing, 1966; Yauman, 1980). Although subsequent studies have suggested that the PHCSCS may be generalized to more diverse populations in the United States, Piers (1984) warned that “comparability of the normative or other psychometric data for other populations should not be assumed” (p. 5).
In addition to the limitations stated above, the following assumptions were made concerning this study: (a) self-concept is a construct measurable by the PHCSCS, (b) self-concept represents the perceptions of one’s self and what this individual perceives others to think of him or her, (c) home schooled children are honest in their self-evaluations when answering the questions contained in the PHCSCS, and (d) response to the test questions contained in the PHCSCS accurately reflect children’s perceived self-concept.

The Instrument

Selection of the Instrument
Selection of an appropriate self-concept instrument requires thorough investigation. Wylie (1974), a foremost authority in the assessment of self-concept measures, observed that few measures have been adequately conceived. With respect to measurement, Wylie (1989) found that “most of the purported self-concept indices have been used only once or a few times, precluding evaluation of their adequacy and interpretation of the results of studies based on them” (p. 1). However, Wylie conducted several psychometric evaluations and has selected the PHCSCS to be one of ten “most meritorious candidates for current use and for further research and development” (p. i). According to Amato (1984), the PHCSCS is one of the most widely used measures of children’s self-concept (Amato, 1984).

Description of the Instrument
The Piers-Harris Children’s Self-Concept Scale entitled, “The Way I Feel About Myself,” was developed by Ellen V. Piers and Dale B. Harris in 1963. The PHCSCS consists of 80 first person declarative statements. Double negatives and ambiguous words such as “many, often, or rarely” are avoided (Piers, 1984).

Reliability of the Instrument
Measures of self-concept are conducted by means of self-report which is essentially the technique most commonly used in self-concept measures. However, because behavior is not directly observed, but instead inferred through verbal reports, distortion and inaccuracy are inevitable, raising questions of reliability and validity. Combs (1962) warned that the self-concept is ” what the individual believes he is. The self-report on the other hand, is what the subject is ready, willing, able, or can be tricked to say he is. Clearly, these concepts are by no means the same” (p. 52). Nevertheless, self reports are accepted as valuable sources of information regarding the individual (Purkey, 1970).
The internal consistency of the PHCSCS ranges from .88 to .93. Test-retest intervals range from 14 days to 1 year, and values of r range from .42 to .96. The median r is equal to .75 (Piers, 1984). The internal consistency is derived from 10 values, each based on a different sample. Total N = 1,047, ages 6-14, grades 3-10. The test-retest coefficients are derived from 19 test-retest reliabilities, each on a different sample. Total N=1,577, ages 6-16, grades 3-8 (Wylie, 1989).

Validity of the Instrument
Validity addresses the question of whether or not an instrument actually measures what it purports to measure. Consequently, the validity of the PHCSCS, as in any instrument intended to measure psychological quality, is subject to scrutiny.
Content validity was built into the scale by using statements from 2,893 children on the topics “What I Like About Myself’ and “What I Dislike About Myself,” (Wylie, 1989). These statements were defined as the universe to be measured and were used by the authors to design the PHCSCS. Vovkell (1983) noted that “a self-concept test would lack content validity if all the items focused on school situations, ignoring the impact of home, church, and other outside-the-school factors on the self-concept” (p. 57). The PHCSCS does consider factors outside the school environment, however, statements such as “My classmates make fun of me” or “I would rather work alone than in a group” do not necessarily apply to the home school population. Hence, the content validity of the PHCSCS may be somewhat weakened insofar as the extent to which the instrument accurately measures the self-concept of home schooled children.
Concerning criterion-related validity, Wells and Marwell (cited in Taylor, 1986) have warned:
There is no single behavior or class of behaviors which can exhaustively or adequately indicate the property of self-esteem and which can be used to criterion-validate measures of self-esteem. For these reasons, the idea of criterion validity, in its traditional sense is not appropriate to self-esteem measurement. (p. 108)

Confirmation of construct validity came about during the initial standardization. During this time, the PHCSCS was administered to 88 institutionalized retarded females whose mean age was 16.8 years. The girls scored significantly lower than the norms of the same chronological or mental age. This was interpreted as meaning that the self-concept is a construct whose entity is measurable by the PHCSCS. This study showed that the PHCSCS discriminated between high and low self-concept in terms of underlying psychological constructs (Piers, 1984).
Construct validity was also noted in a longitudinal study of self-concept in a group of Cleveland minority children. Brody (1984) concluded that the consistency of the self-concept scores, as measured by the PHCSCS, lends support to the idea that self-concept is a “stable, central core of personality” (p. 10).

Data Collection and Analysis

The PHCSCS was administered by the researcher. The statements were read aloud to children in the second and third grade group as they followed along. The children responded “yes” or “no” to each statement. The PHCSCS, accompanied with a demographic profile of each home schooled child, provided the data to be examined. Examination of the data involved computer analysis using inferential statistical procedures.
A parametric statistical test designed to determine if the mean scores of two groups are significantly different is the t-test. Therefore, t-tests were employed to determine differences between the home schooled sample and the conventionally schooled sample. Relationships between the principle variable, self-concept, and the nine demographic variables were examined using multiple regression analysis. According to Borg (1987), “the advantage of multiple regression analysis is that it permits the researcher to explore simultaneously the relationships of several independent (predictor) variables to a dependent (criterion) variable” (p. 147).

Findings

Characteristics of the Sample
Demographic variables were statistically examined and yielded the following sample characteristics: The average age of the sample was 9.71 years and the average grade-level was 4.56. An approximately equal distribution in terms of gender was indicated.
The average number of siblings for the responding child was 2.81 which was higher than the average United States family (U.S. Bureau of Census, 1989). Only 3.1% of the home schooling families had but one child. The average home schooled child had completed 3.41 years of home schooling.
The most frequently selected earnings category for home schooling families was $40,000 or greater, which is higher than the average United States family earnings of $27,326 (U.S. Bureau of Census, 1989).
The average number of home schooled children in the family was 2.54, with over 90% of the families having more than one child in the home school. The average home school operator had completed 14.21 years of education which is higher than the United States median educational level of 12.7 (U.S. Bureau of Census, 1989).

PHCSCS Item Response
Sample responses were directed toward a positive self-concept tabulation on 77 of the 80 items on the PHCSCS. Item 24, “I am good at music;” item 63, “I am a leader in sports;” and item 77, “I am different than other people” tabulated toward lower self-concept. Of the 80 items on the PHCSCS, only 10 items (numbers 3, 24, 26, 51, 55, 57, 63, 69, 71, and 77) indicated a less favorable response then the normative group. These items were statements that dealt with boy/girl popularity and questions in the context of a conventional school setting.

PHCSCS Global Scale: Overall Sample and Subgroups
The mean raw score for the global scale of the PHCSCS was 64.5 for the overall sample (N=67), 65.4 for those home schooled children in the second and third grade group (n=22), and 64.2 for those home schooled children in the fourth through tenth grade group (n=45). The standard deviation was 10.38 for the overall sample, 7.85 for those home schooled children in the second and third grade group, and 9.75 for those home schooled children in the fourth through tenth grade group. The self-concept scores for the overall sample ranged from 35 to 79.

PHCSCS Global and Subscales: Grades Four Through Ten
Approximately 50% of the respondents were at or above the 80th percentile on the PHCSCS. Only 16.4% of the respondents scored below the 50th percentile. The PHCSCS global scale and five of the subscales indicated higher levels of self-concept in the home schooled sample. The Popularity subscale, however, showed slightly lower self-concept when compared to the normative group. The Popularity subscale included questions in the context of a conventional school setting that may account, in part, for the lower self-concept evidenced by the findings.

Tests of Hypotheses
The null hypotheses in this study were of two categories. Those hypotheses in the first category were tested at the .05 level of significance utilizing t-tests. These hypotheses involved the PHCSCS norms and those home schooled children in suburban Los Angeles confined to the sample. The second category involved demographic data. These hypotheses dealt exclusively with relationships within the home schooling sample and were tested at the .05 level of significance utilizing multiple linear regression.
Testing of the hypotheses in the first category showed the home educated group scoring more positive in self-concept measures with significant differences beyond the .001 probability level between (a) the level of self-concept in the home schooled sample and the self-concept norms for the global scale of the PHCSCS, (b) the self-concept of those children who have been
home schooled for two or more years and the norms of self-concept in the conventionally schooled sample, and (c) the self-concept of those home schooled children who have never attended a conventional school and the self-concept norms of the conventionally schooled sample. Although differences in the PHCSCS subscales were not part of the hypotheses testing, significant differences were found in four of the six subscales (see Table 1).
Table 1. Comparison of global and subscale scores: Grades four through ten.

The hypotheses for the second category were examined by multiple regression analyses of the demographic variables utilizing SAS and SPSS programs on an IBM XT computer. Backward elimination and stepwise regression procedures were utilized to statistically determine the best predictive model.
The regression model with all nine variables failed to yield a significant correlation coefficient at the .05 probability level (” = .05). A second analysis, however, determined three variables to have correlation coefficients significant at the .10 probability level (” = .10). The three best predictors of high self-concept in the second analysis were the number of children in the home school (F_{1, 65} = .0644), the home school operator’s educational level (F_{2, 64} = .0767), and the socioeconomic status of the home schooled child (F_{3, 63} = .0943).

Discussion

This study focused on the relationship between home schooling and the self-concept of home schooled children in suburban Los Angeles in grades two through ten. To this date, it is the only known study that has examined self-concept in home schooled children in grades two and three.
Outside of the context of home schooling, literature relevant to self-concept indicated that a person’s self-concept largely reflects his or her socialization. People who see themselves in positive ways as liked, needed, and persons of integrity, tend to behave accordingly, and as a consequence, get along well in our society. Many of the problems which children experience in life were found to be closely connected with the way they see themselves and the world in which they live. It was noted that no one is born with a self-concept. Rather, the development of the self-concept is a lifelong process. This process is primarily social and involves countless interactions with people.
Approximately 50% of the home schooled children in this study were at or above the 80th percentile on the PHCSCS global scale. Only 16.4% of the home schooled children scored below the 50th percentile. More than one-half of the home schooled children were at or above the 80th percentile on the Behavioral, Intellectual and School Status, Anxiety, and Happiness and Satisfaction subscales. More than one-third of the home schooled children were above the 80th percentile on the Physical Appearance and Attributes subscale. However, the distribution of scores on the Popularity subscale were approximately equal with 50.6% below the 50th percentile and 49.4% above the 50th percentile. Questions in the context of a conventional school setting may have been a factor in less favorable scores among home schooled children in this subscale.
The reason for significantly higher levels of self-concept in home educated children is largely unknown. Speculation could lead to many considerations. Hedin (1990), for example, suggested that other factors, rather than educational setting, may influence the development of self-concept. According to Hedin, “these factors include such areas as parental self-concept, [and] relationships with parents…” (p. 71). Swayze (1980) also noted that a child’s self-concept is based primarily on relationships with his or her immediate family.
More specifically, Coopersmith (1967) reported three family conditions which lead a child to value him or herself as a person of worth. These are (a) parental warmth, whereby the child senses the love and support of his or her family and feels that they see him or her as a person of value; (b) respectful treatment, whereby the child’s views are recognized and where he or she has a democratic position in the family; and (c) clearly defined limits, whereby the child knows, through his or her parents’ love and expectations for success, that they care what happens to him or her.
A regression analysis indicated that high self-concept was related most strongly to three predictors: The number of children home schooled in the family; the educational level of the home school operator; and the socioeconomic status of the home schooling family. These predictors, although approaching significance, were not significant at the .05 probability level. The sample size of 67 could have been a major limitation in this area. An ideal sample size would have required a minimum of 94 subjects (14 subjects for the first variable and 10 subjects for each variable thereafter).

Conclusions

The self-concept of home schooling children in suburban Los Angeles was significantly higher (p<.OOl) than the norms of conventionally schooled children on the global scale of the PHCSCS. A low anxiety level could be a contributing factor, since more than 50% of the home school sample were above the 90th percentile on the Anxiety subscale of the PHCSCS. More contact with significant others, parental love, support, and involvement, peer independence, and a sense of responsibility and self-worth may be other contributing factors.
The findings in this study are similar to John Taylor’s (1986) findings. Taylor also found the self-concept of home schooled children to be significantly higher (p<.OOl) than the norms of conventionally schooled children. Some individuals, however, have criticized Taylor’s study because the PHCSCS was administered by the parents. In this study, the PHCSCS was administered by the researcher. It would appear that it makes no difference who administers the PHCSCS. Self-concept is a construct defined as an individual’s perception of himself or herself. Thus, these findings lend support to Taylor’s study.
The demographic factors in this study were not significantly related to the self-concept of home schooled children in suburban Los Angeles. The sample in this study was, perhaps, too small for significant predictors to be entered into the regression model. Had the study tested the hypotheses with ” = .10, the number of children in the home school, the educational level of the home school operator, and the socioeconomic status of the home school family would have been significant predictors. A larger sample may have yielded predictors that were significant beyond the .05 probability level.
The home schooled children in this sample scored higher than the conventionally schooled children on 89% of the PHCSCS items. A higher percentage of home schooled children, however, perceive themselves as “different from other people,” “not good in music,” and “usually not a leader in games and sports.” Nevertheless, it would appear that insofar as self-concept is a reflector of socialization, few home schooled children are socially deprived. Those who criticize home schooling should not do so on the grounds of self-concept or socialization. These factors tend to favor home schooled children over conventionally schooled children.
On the average, home school families in suburban Los Angeles have more children than the national average. In terms of family income, the socioeconomic status of home schooling families in suburban Los Angeles is higher than the average United States family. Most home schooling families in suburban Los Angeles have more than one child in the home school and the average educational level of the home school operator is higher than the educational level of the general United States population.

Recommendations

The PHCSCS was developed in the 1960s and standardization was based on a single Pennsylvania school district. Therefore, normative data for other populations should not necessarily be assumed. Further research is needed to determine whether home schooled children in suburban Los Angeles would score significantly different than children who attend conventional schools in suburban Los Angeles. Perhaps standardization of the PHCSCS should be updated based on a nationwide sampling.
Further research is needed to determine the effect of social factors, such as parental involvement or peer independence, on children’s self-concept. Further research is also needed to examine the differences in anxiety levels between home schooled and conventionally schooled children. What are the contributing factors to anxiety and how might these factors be dealt with in order to foster a positive and healthy self-concept development?

References

Amato, Paul R. (1981). The Piers-Harris Children’s Self Concept Scale: An evaluation of its use on an Australian population. Melbourne, Australia: Institute of Family Studies. (ERIC Document Reproduction Service No. ED 255 307)
Beane, James A., Lipka, Richard P., & Ludewig, Joan W. (1980). Synthesis of research on self-concept. Educational Leadership, 38, 84-88.
Beck, Mitchell A., Roblee, Kathleen, & Hanson, John. (1982). Special education/regular Education: A comparison of self-concept. Education, 102(3), 277-279.
Black, F. William. (1974). Self-concept as related to achievement and age in L.D. children. Child Development, 4, 1137-1140.
Borg, Walter R. (1987). Applying educational research. White Plains, NY: Longman.
Brim, 0rville G., & Wheeler, Stanton. (1966).Socialization after childhood. New York: John Wiley & Sons.
Brody, Grace F. (1984). A longitudinal study of self-concept, attitudes toward school, and vocational aspirations in a group of East Cleveland minority children. Cleveland, OH: Case Western Reserve University, Armington Research Program. (ERIC Document Reproduction Service No. ED 255 311)
Charvoz, Adrienne. (1988). Reactions to the home school research: Dialogues with practitioners. Education and Urban Society, 21(1), 85-95.
Combs, Arthur W. (1962). The self in chaos. A review of Wylie, R. C., The self-concept: A critical survey of pertinent research literature. Contemporary Psychology, 7, 53-54.
Combs, Arthur W. (1982). A personal approach to teaching. Boston: Allyn and Bacon.
Common, Ron W., & MacMullen, Marilyn. (1986). Home schooling: A growing movement. Education Canada, 26(2), 4-7.
Coopersmith, Stanley. (1967). The antecedents of self-esteem. San Francisco: W. H. Freeman.
Elkind, David. (1988). The hurried child. New York: Addison-Wesley.
Felker, Donald W. (1974). Building positive self-concepts. Minneapolis, MN: Burgess.
Goodlad, John I. (1984). A place called school. New York: McGraw-Hill.
Gustavsen, Gunnar A. (1981). Selected characteristics of home schools and parents who operate them. Ed.D. dissertation, Andrews University, Berrien Springs, MI.
Hedin, Norma S. (1990). A study of the self-concept of older children in selected Texas churches who attend home schools as compared to older children who attend Christian schools and public schools. Doctoral dissertation, Southwestern Baptist Theological Seminary, Fort Worth, TX.
Hilgard, Ernest R., Atkinson, Rita L., & Atkinson, Richard C. (1979). Introduction to psychology (7th ed.). New York: Harcourt Brace Jovanovich.
Kelley, Steven W. (1991). Self-concept in home-schooled children in suburban Los Angeles. M.A. thesis, Eastern New Mexico University, Portales, NM.
Kink, Linda L. (1983). An analysis of Sacramento area parents practicing home schooling. M. A. thesis, California State University, Sacramento, CA.
Kugle, Cherry L. (1980). Self-esteem and academic behavior among second graders. Unpublished doctoral dissertation, University of Texas, Austin, TX.
Kugle, Cherry L. & Clements, Richard 0. (1981). Self-esteem and academic behavior among elementary school children. Austin: Research and Development Center for Teacher Education: University of Texas.
Kugle, Cherry L., Clements, Richard 0. & Powell, Phillip M. (1983). Level and stability of self-esteem in relation to academic behavior of second graders. Journal of Personality and Social Psychology, 4, 201-207.
Lines, Patricia M. (1987). An overview of home instruction. Phi Delta Kappan, 69, 510-517.
McDonald, Kay A. (1980). Enhancing a child’s positive self-concept. In T. D. Yawkey (Ed.), The self-concept of the young child (p. 51-62). Provo, UT: Brigham Young University Press.
McNeil, Elton B. (1968). The concept of human development. Belmont, CA: Wadsworth.
Montgomery, Linda. (1989). The effect of home schooling on the leadership skills of home-schooled students. Home School Researcher, 5(1), 1-10.
Moore, Raymond S., & Moore, Dorothy N. (1982). Home-spun schools. Waco, TX: Word Books.
Moore, Raymond S. (1984). The cradle and the gap. In C. Marshner (Ed.), A blueprint for education reform (p. 231-250). Washington, DC: U. S. Department of Education, Washington. (ERIC Document Reproduction Service No. ED 350 776)
Moore, Raymond S., & Moore, Dorothy N. (1981). Home grown kids. Waco, TX: Word Books.
Moore, Raymond S., & Moore, Dorothy N. (1988). Home school burnout: What it is, what causes it, and how to overcome it. Brentwood TN: Wolgemuth & Hyatt.
Piers, Ellen V. (1977). Children’s self-esteem, level of esteem certainty, and responsibility for success and failure. Journal of Genetic Psychology, 130, 295-304.
Piers, Ellen V. (1984). Piers-Harris Children’s Self-Concept Scale: Revised manual 1984. Los Angeles: Western Psychological Services.
Purkey, William W. (1970). Self-concept and school achievement. Englewood Cliffs, NJ: Prentice-Hall.
Purkey, William W., & Novak, John M. (1984). Inviting school success. Belmont, CA: Wadsworth
Rakestraw, Jennie F. & Rakestraw, Donald A. (1990) Home schooling: A question of quality, an issue of rights. The Educational Forum, 55, 67-77.
Ray, Brian D. (1988). Home schools: A synthesis of research on characteristics and learner outcomes, Education and Urban Society, 21(1), 16-31.
Roach, Virginia. (1988). Home schooling in an era of educational reform. School Business Affairs, 54(11), 10-14.
Schimmels, Cliff. (1989). Parents as partners with public education. In K. Sidney (Ed.), The Blackboard Fumble (p. 139-149). Wheaton, IL: Victor Books.
Silverman, Rita, & Zigmond, Naomi. (1983). Self-concept in learning disabled adolescents. Journal of Learning Disabilities, 16, 478-490.
State v. Riddle, 285 S. E. 2d 367, 1981.
Swayze, Marian C. (1980). Self-concept development in young children. In T. D. Yawkey (Ed.), The self-concept of the young Child (p. 33-48). Provo, UT: Brigham Young University Press.
Taylor, John W., 5th. (1986). Self-concept in home-schooling children. Doctoral dissertation, Andrews University, Berrien Springs, MI.
U. S. Bureau of the Census (1989). Statistical abstract of the United States: 1990. Washington, DC: U. S. Government Printing Office.
Vockell, Edward L. (1983). Educational research. New York: MacMillan.
Wartes, Jon. (1988). The Washington Home School Project: Quantitative measures for informing policy decisions. Education and Urban Society, 21(1), 42-51.
Wing, Sarah W. (1966). A study of children whose reported self-concept differs from classmates’ evaluation of them. Unpublished doctoral dissertation, University of Oregon, Eugene.
Wilson, Stephanie. (1988). Can we clear the air about home schooling? Instructor, 19(1), 11
Wright, Cheryl. (1988). Home school research: Critique and suggestions. Education and Urban Society, 21(1), 96-115.
Woolfolk, Anita E. (1987). Educational psychology. Englewood Cliffs, NJ: Prentice-Hall.
Wylie, Ruth C. (1974). The self-concept: A review of methodological considerations and measuring instruments. Lincoln, NE: University of Nebraska Press.
Wylie, Ruth C. (1989). Measures of self-concept. Lincoln, NE: University of Nebraska Press.

Yauman, Beth E. (1980). Special education placement and the self-concepts of elementary school age children. Learning Disability Quarterly, 3, 30-34.