Scott White, Elizabeth Williford, John Brower, Terance Collins,

 Roman Merry, and Maryam Washington

Department of Psychology, Belhaven College, Jackson MS, swhite@belhaven.edu

Home School Researcher, Volume 17, No. 4, 2007, p. 1-7

Abstract

Home schooled students’ ability to successfully adjust to college life is one important criterion to demonstrate a positive outcome of home schooling. The present study compared Christian college freshmen who had previously been home schooled with a matched sample of traditionally schooled Christian freshmen on the College Adjustment Scale. The mean scores of the two groups were compared across the nine CAS scales designed to measure emotional, behavioral, social and academic problems typically presenting to university counseling centers. The home schooled students scored significantly lower on the anxiety subscale, while no difference was found between the two groups on the remaining scales. In addition, it was noted that there was a general trend characterized by home schooled students reporting fewer symptoms of emotional distress and social problems, and achieving higher first semester GPAs. The results suggest that home schooled college freshmen successfully adjust to the social and academic and social environment of a Christian college with a diverse student population. The limitations of this study in addressing key questions regarding the effect of home schooling upon socialization are discussed.

 Keyword descriptors: home schooling, homeschooling, home education, college, college adjustment, adults, socialization, academics

During the past two decades home schooling has become an educational option considered by a growing number of parents in the United States. This growth has been fueled partly by growing criticism of public education, and as a result of less restrictive state laws governing home schooling (Basham, 2001). With the rapid growth in home schooling, questions have been raised in the public forum regarding possible negative effects of home schooling upon children. National publications such as Time, Newsweek, The National Post, and Wall Street Journal and major television broadcasts have addressed the home schooling movement. Early questions and criticisms focused upon concerns that home schooled students might not receive adequate educational instruction within a home school setting. However, research examining home schooled students’ academic achievements have consistently found that they score higher than the national norms on standard achievement tests (Ray, 1994; 1997; 2002). As this body of research has grown, much of the criticism about academic performance of home schoolers has been muted.

Home Schooling and Socialization

More recently, the focus of questions and criticisms regarding home schooling have been directed at concerns about socialization. Home schooling parents with strong ideological convictions often dismiss these concerns based upon anecdotal evidence gathered from their own personal observations or the viewpoints of trusted authorities. However, many of the growing number of parents considering home schooling express concerns about beginning to home school without clear evidence that it will not socially stunt or harm their children’s social and emotional development (Kitchen, 1991). These same concerns have been raised by psychologists and educators who rely heavily upon empirical evidence. A feature article in the APA Monitor (Murray, 1996) covering the home school movement left the reader with concerns about the lack of data addressing the effects of home schooling upon the social and emotional development of children. Wright (1988), an educator, posed the question, “…how will these children function in our diverse multicultural society when they are raised in a setting with monolithic views and beliefs?” (p.111).

The relatively few empirical studies that have examined home schoolers’ socialization have most often focused upon the impact of home schooling on elementary and middle school-aged children’s self-esteem or self-concept (Hedin, 1991; Kelly, 1991; Kitchen, 1991; Shyers, 1992; Taylor, 1986; Tillman, 1995). These studies have generally found home schooled children to have equal or better self-esteem than traditionally schooled students depending upon the norms used for comparison. However, Francis and Keith (2000) have argued that the constructs of self-esteem and self-concept are insufficient measures of social behavior. Methodological concerns have also limited the interpretation of these studies (Francis & Keith, 2000; Kelly, 1991; Miller, 2000). The data showing that self-esteem does not appear to be negatively impacted by home schooling in this age group is important, but a much broader research program is needed to address larger concerns about socialization (Francis & Keith, 2000; 2004).

Some researchers have branched out from this line of research into studying other aspects of socialization such as developmental maturity (Smedley, 1992), leadership skills (Montgomery, 1989), the qualitative examination of adults who were previously home schooled (Knowles & Muchmore, 1995) and social reticence and social anxiety (Brower, McNeil, Park & White, 2003). Recently, Brian Ray (2004) completed an extensive study of the social, civic, religious, educational and life-style practices of 5,254 previously home schooled individuals who had completed their secondary education.

A significant hurdle in examining the impact of home schooling on socialization is that “researchers continue to struggle to measure a very difficult construct within a very unique population” (Francis & Keith, 2000, p. 8). Socialization may not be defined in the same manner by home schooling parents as it is defined by concerned critics (Miller, 2000). Therefore, the dialogue and debate regarding the meaning of socialization may derail many attempts to address the concerns of both critics and parents. After reviewing definitions of socialization in the literature, Miller (2000) recommends a populist view of socialization that he defined as “the ability the child possesses to interact and adapt to social contexts in a successful manner” (p. 8). Miller’s recommendation to focus upon socialization as successful adaptation to social contexts is consistent with recommendations of other home school researchers. In a survey of home schooling researchers, Cizek and Ray (1995) found that researchers recommended that future research should focus upon real-life outcomes of home schooled children. One specific real-life outcome identified by some of the researchers was the need to examine home schoolers’ transition from home schooling to higher education (Cizek and Ray, 1995).

Socialization and College Adjustment

Various estimates suggest that approximately 50% of home schooled children eventually attend college (Galloway & Sutton, 1995). In Ray’s (2004) major study, 75% of this relatively young sample of previously home schooled adults, average age being 21, had attained some college courses. A full 49% of those surveyed were full-time college students. In addition, Ray (2004) found that approximately half of home educating parents had earned a bachelor’s degree or higher. Therefore, for many home schooled students, success in college is one of several final outcomes that demonstrate whether home schooling was successful. Furthermore, a home schooler’s ability to adjust to this important institution in society would be evidence that the child has been adequately socialized. In contrast, if a home schooled student has significant difficulty handling the stressors of the new environment outside of the home, is unable to make friends from diverse backgrounds, or cannot compete academically in the classroom, one could argue that they have not adequately socialized to live in the larger society. While all students entering college face significant adjustments, home schooled students are likely to face more adjustments related to social relationships and educational methodology. A qualitative study of home schooled students adjusting to public high schools found that home schooled students experienced many unique stressors adapting to traditional education (Romanowski, 2002).

Typically the information cited regarding home schoolers and college attendance is related to whether colleges are willing to accept home schoolers (Klicka, 2004). The empirical research available on the performance and transition of home schooled students after arriving at college is limited (Galloway & Sutton, 1995; Jones & Gloeckner, 2004; Lattibeaudiere, 2000; Ray, 2004; Sutton & Galloway, 2000). Perhaps one reason for the lack of research may be the lack of access that most researchers have to this population. For example, a recent investigation of home schooled students attending state universities in Mississippi found that only 11 out of 5,462 incoming freshmen had been home schooled (Kanengiser, 2003).

Because a high proportion of home schooled parents tend to be white, middle-class, conservative Christians (National Center for Education Statistics, 2002; Ray, 2004) many home schooled students may choose to attend Christian colleges rather than state universities. Sutton and Galloway (2000) studied home schooled students at a Christian college across five domains of learning outcome (achievement, leadership, professional aptitude, social behavior, and physical activity). They concluded that previously home schooled students received essentially equivalent educations to college students who had received public or Christian high school educations. Empirical studies of home schooled college students have typically focused upon the cognitive or academic performance of home schooled students at college (e.g. Galloway & Sutton, 1995; Jones & Gloeckner, 2004; Klicka, 2004) and have not focused on socialization. Only one study, an unpublished dissertation, has sought to particularly focus upon home schoolers’ social adjustment to college (Lattibeaudiere, 2000). This qualitative study examined 13 students at two Christian colleges and 12 students at two state universities. On the one measure used in the study, the Student Adaptation to College Questionnaire, the home school sample scored at the 76^th percentile on the social adjustment subscale and the 84^th percentile on the emotional adjustment subscale. This compares quite favorably with the normative sample’s 50^th percentile norm. However, besides this one qualitative study, accounts of successful social and emotional transitions to college are generally anecdotal reports provided by admissions officers (Klicka, 2004).

In light of the limited research on socialization and home schooling, this study’s purpose is to expand the breadth of existing socialization research with home schooled individuals to include emotional and social adjustment, as well as academic adjustment to college. The goal of this study is to compare the adjustment difficulties of home schooled vs. traditionally schooled students attending a Christian liberal arts college. Because the debate regarding the impact of home schooling upon socialization is largely ideological, there would appear to be contradictory hypotheses regarding whether traditionally schooled or home schooled students might better adjust to college life. Advocates of traditional schooling might suggest that due to the greater prior exposure to peer culture and traditional education, that traditionally schooled college students would experience fewer problems adjusting to college. Advocates of home schooling might suggest that due to the warm, stable and accepting environment of most home school homes, the home schooled students might possess greater maturity and resilience and thus have fewer problems adjusting to college. Therefore, due to dearth of empirical evidence supporting either hypothesis, a nondirectional hypothesis was chosen. It was anticipated that a difference would exist between home schooled and traditionally schooled college students, but the nature and direction of the difference was not specified.

Difficulty adjusting to a new situation should produce reports of emotional, social, or academic difficulties. In contrast, the absence of reports of these difficulties might be construed as the result of successful adjustment. The measure chosen to assess the students’ difficulties adjusting to college was the College Adjustment Scale (Anton & Reed, 1990). The College Adjustment Scale allows for the measurement of the experience and expression of adjustment problems among the college students. It is hoped that this study will expand the knowledge available about how well home schooled students can adapt and cope with entering college, a key social environment and societal organization outside of their homes.

Method

Participants

The participants included 36 first-time freshmen college students attending their first semester at a small, private, Christian liberal arts college in Jackson, Mississippi. The study was conducted with two samples, matched for age, race, gender, and having been raised in a Christian home. These two groups were: (a) previously home schooled students (n=18) and (b) previously traditionally schooled students from both private and public high schools (n=18).

For the home schooled sample, thirty one freshmen students who had graduated from a home school were identified through a college database. Researchers attempted to contact by phone each of the home schooled freshmen. Students who could be reached and agreed to participate were then asked to complete a demographic form along with the College Adjustment Scale. Demographic sheet information was used to select those participants meeting the pre-determined criteria for the study. To be placed in the home schooled group, students had to have been home schooled a minimum of 80% of the time between their 7^th and 12^th grades, been within the age range of a traditional freshman without significant life experience after high school (16-20 years of age), and identify that they had been raised in a Christian home. The final home school sample that met these criteria consisted of 13 female and 5 male subjects.

Subjects for the traditionally schooled sample were obtained by administering the demographic sheet and instrument in three freshman English classes (2 advanced and one regular class). A randomly collected set of demographic sheets from these classes was then compared with the characteristics of the home schooled subjects. The 18 home schooled subjects were matched with the traditionally schooled sample across the following characteristics: age, race, gender and affirmation that they were “raised in a Christian family/home that actively encouraged Christian values and lifestyle”. The first traditionally schooled subject that matched the characteristics of each home schooled subject was placed in the traditionally schooled sample.

The geographical regions of the students were determined from college data banks. Of the home schoolers, 9 students were from the southeast, 4 from the southwest, 3 from the north central, and 1 from the northeast regions of the U.S. Among the traditional students, 15 were from the southeast, 1 from the southwest, 1 from the north central, and 1 from the Midwest. The traditional students included 7 private and 11 publicly educated students. To further assess the similarity of the two groups, the college registrar’s office provided group mean achievement scores for each sample as reported prior to college admittance. For the home schooled group, the average ACT score (n= 14) was 25 and the average SAT score (n= 10) was 1237. For the traditionally schooled students, the average ACT score (n=15) was 24.3 and the average SAT (n=5) score was 1108.

Measure

The College Adjustment Scales (CAS) is a tool specifically designed to identify and evaluate “the experience and expression of adjustment problems in college students throughout the college years” (Anton & Reed, 1990, p. 13). The CAS was designed as a screening instrument for common developmental and psychological problems to be used by professionals at college and university counseling centers. CAS scales were based on an analysis of  problems most often presented at college and university counseling services. Nine scales were selected by the test developers based upon their internal consistency reliability coefficients. The CAS manual reported internal consistency reliability coefficients for each CAS scale ranged from .80 to .92, with a mean of .86. The convergent and discriminant validity of the CAS was examined in four separate studies, all of which supported the validity of the CAS. A fifth study compared the standardization group to students who were receiving counseling  services for personal, academic and career concerns and found significant differences between the groups. The CAS scales are: anxiety (AN), depression (DP), suicidal ideation (SI), substance abuse (SA), self-esteem problems (SE), interpersonal problems (IP), family problems (FP), academic problems (AP) and career problems (CP). These scales are not independent and relationships between the scales need to be considered by professionals interpreting the results. All subjects received a 108-question CAS booklet (12 questions on each scale), a separate answer sheet based on a 4-point Likert scale, a demographics sheet, and a consent form.

Procedure

All participants of the study were requested to complete a consent form, the CAS and a demographics sheet. Attempts were made to call 31 students identified in the school database as having graduated from a home school. A call script was used asking students if they would be willing to meet to complete a questionnaire as part of a study for the callers’ honors psychology class. Students were not informed that they had been identified due to their status as a home school graduate. Students who agreed were met, no more than two at a time, in the school library and given instructions on how to complete all three items. Once completed, all materials were returned to the administrator. All students in three freshman English classes were asked to participate in the study using scripted instructions regarding the test. Upon completion of the tests and materials, each student returned his or her packet to the administrator positioned at the front of the class. Five of the home schooled sample completed the instrument in the English classes prior to being reached by phone to request their willingness to participate in the study. Traditional students’ CAS answer sheets were placed in a random order based on collection from the classroom. Tests of individuals whose characteristics did not match the home schooled sample were discarded. The home schooled students answer sheets were then matched to the first traditional students answer sheet that met the matching characteristics identified above.

Results

Table 1 shows a comparison of the mean scores and standard deviations of the home schooled students and the traditionally educated students on each subscale of the CAS. In addition, mean scores of the standardization group on each subscale are given as a comparison. Due to the absence of any empirical studies supporting a directional hypothesis, a two-tailed dependent t-test was run on each subscale of the CAS. Table 2 shows the results of a matched two-tailed dependent t-test between the two groups on each subscale of the CAS. While the test revealed a significant difference on the AN (anxiety) subscale, t(17)= -2.65, p< .05. Home schooled students were found to be significantly less anxious than traditionally schooled students. All other t-scores were found to be insignificant. However, it should be noted that on the FP (family problems) scale, the level of significance was narrowly missed before rounding (.059).

To assist professionals in using the CAS as a screening instrument, specific cut-off scores were defined by the test developers (Anton & Reed, 1990). These cut-off scores are useful to college mental health professionals to help them determine if a particular student might need further face-to-face evaluation or intervention. T scores between 60T (84^th percentile) and 69T (97^th percentile) are defined as within the borderline range of problem severity. T scores above 70T (98^th percentile) are defined as indicating a significant problem. The developers of the CAS suggest that scores above 60T might suggest the need for further personal interviews by professionals to determine the severity, nature and impact of the problem on the student’s life.

SA

12.39

  .78

12.61

1.29

16.39

5.55

Note. Degrees of freedom is 17

*p < .05

The frequency with which each group achieved a score at or above these cut-off scores was tabulated and compared between the two groups. Among the home schooled group, 5 (28%) had a score within the borderline range on at least one CAS scale. By comparison, 9 (50%) of the traditional students had a score in the borderline range. Comparing the students with a score at or above the significant level, only one student in the home school group and two in the traditionally schooled group had a score that reached the level deemed significant. Table 3 breaks down of the frequency and percentage of students who scored above 60T (84^th percentile) on each of the 9 scales.

Because academic performance is one measure of adjustment to college academic life, the first semester GPAs for each group were examined. Researchers provided separate lists of student I.D.s for each sample and requested the college registrar’s office to provide group data related to ACT scores and first semester GPA’s for each sample. The reported mean GPA of the home schooled group was 3.52 with a range between 1.96 and 4.0 (on a 4 point scale). The mean GPA of the traditionally schooled group was 3.16 with a range between 1.08 and 3.93.

Discussion

As the results indicate, home schooled students reported significantly fewer anxiety symptoms than a matched sample of traditionally schooled students. The items on this subscale address general statements of anxiety as opposed to specific anxieties regarding academics. Close examination of the standard deviations between the two groups also reveals that home schooled students appear to have greater homogeneity than traditionally schooled students related to their reported problems and symptoms measured by the anxiety (AN) and depression (DP) subscales. An overall trend was revealed in the home schooled sample reporting fewer problems or symptoms across all of the 9 CAS scales. In fact, twice as many traditionally schooled students would have been identified by a university counseling service, using the test developers’ cut-off scores, to assess for problems that might require counseling or other interventions. Because students did not complete a CAS prior to arriving at the college, it is unclear whether differences between these two groups existed prior to their entering college, or if the differences found relate to each groups ability or lack of ability to cope with the stressors of their first semester of college life.

The utility of the findings of this study may be most useful in addressing the question raised by critics of home schooling who hypothesize that home schoolers will have difficulty adjusting to typical social environments in society after they finish their home schooling. Firstly, the findings of this study do not support the hypothesis that home schooled students will have a greater difficulty adjusting to traditional academic structures and/or methodology. No significant difference was found between the two groups related to reports of academic problems (AP). Furthermore, the home schoolers’ first semester GPA mean and range compared quite favorably to that of the traditionally schooled students. The home schoolers’ slightly higher mean GPA (3.52 vs. 3.16) is also notable in light of the similarity of the each group’s ACT means of (25 vs. 24.3).

Secondly, this study does not support the hypothesis that home schooled students will have more severe emotional or social difficulties related to the unique social adjustments that they experience entering college life. No difference was found between home schooled Christian freshmen and traditionally schooled Christian freshmen on depression (DP), suicidal ideation (SI), substance abuse (SA), self-esteem problems (SE), interpersonal problems (IP), family problems (FP), or career problems (CP). For those particularly concerned that home schooled students might have significant social problems entering college, this was not supported by these findings. These findings do not demonstrate that home schooled students necessarily have social competence or good interpersonal skills. However, they support the idea that home schooled students do not currently perceive a greater degree of problems in their interpersonal relationships as compared to traditionally schooled students.

Further evidence for the proposition that previously home schooled students are adjusting well to college life can be found by comparing the means of the home schooled sample with the standardized sample obtained by the test developers (Anton & Reed, 1990). One must be cautious interpreting a comparison between such divergent samples, therefore, no statistical comparisons were conducted. However, it is noteworthy that the home schooled group means were lower than the national means across all scales indicating fewer reported problems (Table 1). Therefore, there is some supporting evidence to suggest that home schooled students are adjusting to college life at least as well as the general population of college students.

A key limitation of some previous research related to socialization of home schoolers has been that the sampling methods utilized often required participants to be aware that they were being chosen as subjects due to their status as home schoolers (Francis & Keith, 2000). An important aspect of the methodological design of this study was the fact that the home schooled students were blind to the fact that the study was related to home schooling. Therefore, in this study, it is less likely that subjects in the home school group were purposely seeking to portray themselves in the best light to help demonstrate the validity of home schooling. However, the research design cannot rule out the possibility that home schooled college students might be generally less likely to report problems or distress, thus presenting themselves in a favorable light.

Despite the fact that home schooled students have additional adjustments  (e.g., social relationships, peer pressure, classroom structure, etc.) to higher education as compared to traditionally schooled freshmen (Romanowski, 2002), they appear to be able to adjust as well or better than traditionally schooled freshmen to collegiate life at a Christian college as measured by these various scales of college adjustment. A key issue that needs to be kept in mind is that these Christian freshmen were adjusting to a specific social environment and structure, namely a Christian college. The college to which they were adjusting has approximately 1,000 traditional students living on campus or in the nearby community. Furthermore, Christian commitments, beliefs and values of the administration and faculty were likely to closely match those of many of these students.

However, these home schooled students were in an environment that forced them to adapt to a diverse community similar in many ways to other societal organizations. The student body has over 30% minority students and includes a significant number of international students. The students at the college represent a broad range of denominations, socio-economic classes, and cultural backgrounds. Furthermore, some students do not identify themselves as Christians but attend the college due to other factors. The college does not require that all students attending the college assent to a personal faith in Christ. The previously home schooled students are also confronted by many peers who make lifestyle choices different from their own. Most of the college peers of the home schooled students would be considered less conservative in their dress, entertainment interests, moral values and moral behaviors, than those typically experienced in most Christian home schooled families. Therefore, these students are not entering a homogenous social community that necessarily mirrors their family backgrounds. The home schooled students sampled are being forced to adapt to a social environment significantly divergent from the social environment in many of their homes or home school support groups.

There are important limitations to generalizing the results of this study. The findings of this study may not generalize to previously home schooled Christian freshmen trying to adjust to a college or university with significantly different characteristics. It is also important to remember that this measure was taken during the initial adjustment phase of college, namely the 8^th to 12^th weeks of their first semester. Therefore, the findings are limited to this early transition period. However, these findings are consistent with Lattibeaudiere’s (2000) qualitative sample made up of sophomores, juniors, and seniors. Finally, one must be cautious generalizing from a small sample measuring one group of students on one campus. Further empirical studies measuring college adjustment at multiple sites is recommended to follow up this study.

In conclusion, this study expands the body of empirical evidence examining the effect of home schooling upon socialization. This study examines one of the important real world factors of concern to many parents who might be considering whether home schooling is a viable educational alternative for their children. Because a large portion of home schoolers are conservative Christians (National Center for Education Statistics, 2002), comparing Christian home schooled and Christian traditionally schooled students adjusting to a Christian college has high external validity to address parents’ concerns. Socialization is a broad construct that requires a diverse program of studies to examine the variety of pieces of the mosaic often referred to as socialization. If one adopts Miller’s (2000) populist view of socialization as “the ability the child possesses to interact and adapt to social contexts in a successful manner” (p. 8), then this study, as a piece of the socialization mosaic, supports the notion that home schooled students are able to successfully adapt emotionally, interpersonally, and academically to their first semester of college.

References

Anton, William & Reed, James (1990). College Adjustment Scales:  Professional Manual. Lutz, FL:  Psychological Assessment Resources.

Basham, Patrick (2001). Home schooling:  From the extreme to the mainstream?  The Fraser Institute (Public Policy Sources, Number 51). Vancouver, BC:  The Fraser Institute.

Brower, John, Mc Neil, Kirsten, Park, Jessica. & White, Scott. (2003, September). The assessment of social anxiety in home schooled vs. traditionally educated students:  An analysis of social anxiety and social reticence of college freshmen. Poster session presented at the annual meeting of the Mississippi Psychological Association, Biloxi, MS.

Cizek, Gregory & Ray, Brian (1995). An analysis of home education research and researchers. Home School Researcher, 11, 1-9.

Francis, David & Keith, Timothy (2000). Self-esteem and home schooling socialization research:  A work in progress. Home School Researcher, 14, 1-9.

Francis, David & Keith, Timothy (2004). Social skills of home schooled and conventionally schooled children:  A comparison study. Home School Researcher, 16, 15-24.

Galloway, Rhonda & Sutton, Joe (1995). Home schooled and conventionally schooled high school graduates:  A comparison of aptitude for and achievement in college English. Home School Researcher, 11, 1-9.

Hedin, Norma (1991). Self-concept of Baptist children in the educational settings. Home School Researcher, 7, 1-5.

Jones, Paul & Gloeckner, Gene (2004, Spring). First-year college performance:  A study of home school graduates and traditional school graduates. Journal of College Admission, 17-20.

Kanengiser, A. (2003, March 11). College has niche appeal. The Clarion Ledger, pp. A1, A4.

Kelly, Steven (1991). Socialization of home schooled children:  A self-concept study. Home School Researcher, 7, 1-12.

Kitchen, Paul (1991). Socialization of home school children versus conventional school children. Home School Researcher, 7, 7-13.

Klicka, Chris (2004). Home schooled students excel in college. Home School Legal Defense Association:  Issue  Analysis. Retrieved August, 12, 2005, from http://www.hslda.org/docs/nche/.asp

Knowles, Charles & Muchmore, James (1995). Yep! We’re grown up, home-schooled kids – and we’re doing just fine, thank you!  Journal of Research on Christian Education, Vol. 4(1), 35-56.

Lattibeaudiere, Vivien (2000). An exploratory study of the transition and adjustment of former home schooled students to college life. Unpublished doctoral dissertation, University of Tennessee, Knoxville.

Miller, Bryan (2000). Socialization and home educated children:  An exploratory study. Home School Researcher, 7-14.

Montgomery, Linda (1989). The effect of home schooling on the leadership skills of home schooled students. Home School Researcher, 5, 1-10.

Murray, B. (1996). Home schools:  How do they affect children?  The APA Monitor, 27, pp. 1, 43.

National Center for Education Statistics. (2002). Education Statistics Quarterly. Retrieved April 22, 2003, from http://nces.ed.gov/pubs2002/quarterly/fall/g3-2.asp.

Ray, Brian (1994). A nationwide study of home education in Canada:  Family characteristics, student achievement and other topics. Salem, OR: National Home Education Research Institute.

Ray, Brian (1997). Strengths of their own-homeschoolers across America:  Academic achievement, family characteristics, and longitudinal traits. Salem, OR: National Home Education Research Institute.

Ray, Brian (2002). Worldwide guide to homeschooling:  Facts and stats on the benefits of home school. Salem, OR:  National Home Education Research Institute.

Ray, Brian (2004). Home educated and now adults. Salem, OR: NHERI Publications.

Ray, Brian (2004, Fall). Homeschoolers on to college:  What research shows us. The  Journal of College Admission, 5-11.

Romanowski, Michael (2002). Home schooled students’ perceptions of the transition to public school: Struggles, adjustments, and issues. Home School Researcher, 15, 1-11.

Smedley, Thomas (1992). Socialization of home school children. Home School Researcher, 8, 9-16.

Shyers, Larry (1992). A comparison of social adjustment between home and traditionally schooled students. Home School Researcher, 8, 1-8.

Sutton, Joe & Galloway, Rhonda (2000). College success of students from three high school settings. Journal of Research and Development in Education, 33, 137-146. Abstract retrieved August 12, 2000, from ERIC digest.

Taylor, John (1986). Self-concept in home schooling children. Home School Researcher, 2, 1-8.

Tillman, Vicki (1995). Home schoolers, self-esteem, and socialization. Home School Researcher, 11, 1-6.

Wright, Cheryl (1988). Home School Research:  Critique and suggestions for the future. Education and Urban Society, 21, 96-113.