Research in the field of home education has logically focused on the academic achievement of students who have been home schooled, usually comparing their performance on standardized tests to the national norms. These studies consistently report findings that demonstrate that home schooled students perform as well as, if not better than, their conventionally schooled counterparts (Lines, 1987; Maarse Delahooke, 1986; McAvoy, 1986; Schmidt, 1989; Wartes, 1990; Ray, 1990, 1992, 1993).
However, the measure of the success or failure of the home school movement lies in the success or failure of the adults who have been home schooled. What happens to these children who complete their education in a home school setting? Do they go on to lead productive lives? Do they get jobs? Do they go to college? Toch (1991) reports that indeed an “estimated 50 percent of home-schooled students attend college, about the same rate as their public-school counterparts” (p. 74).
Yet, it is success in, not just admission to, college that indicates the adequacy of academic preparation for that endeavor. Therefore, a two-fold question becomes relevant to the analysis of the success of home education: (a) do the home schooled students who desire a college education succeed in college; and (b) what is the extent of their success compared to college students from conventional educational settings?
Obviously, success in college can be measured by completion of a college degree program. Yet, it is a fact that not all students who enroll as freshmen in college will ultimately graduate. Pantages and Creedon (1978) found that only 7 out of 10 students who enter college will ever obtain a college degree. Although there are undoubtedly many reasons that students drop out of college, research has shown that aptitude for college-level study and academic achievement in college courses play a significant role (Hudson, 1989; Nelson, 1985; Pantages & Creedon, 1978). Research also shows that academic achievement in college is closely aligned with the mastery of certain academic skills: namely, reading, writing, and thinking skills (Adelman & Reuben, 1984; Bauer, 1987; Kallingal, 1989; Pascarella, 1987; Raburn & VanSchuyver, 1984; Wood, 1988). And the one college course that seems to focus on the development of all three of these skills is freshman English. Typically, a freshman-level college English course emphasizes reading and other cognitive skills in the development of effective writing ability. Lazere (1982) explains this close relationship between English courses and the rest of the college curricula: “the process of writing . . . [is] not an isolated end in itself but [it is] a means of assimilating knowledge and analytic skills that are integral to . . . other studies and general intellectual growth” (pp. 3-4).
Because the skills necessary to succeed in a college English course are representative of the skills necessary to succeed in most other college-level courses, freshman English is a logical focus for a study of home schooled students’ potential to succeed in college. Therefore, this study was designed to investigate home school graduates’ potential for success in college by comparing their performance with that of students who have graduated from conventional schoolsCspecifically, conventional public schools and conventional Christian schools (i.e., institutions whose philosophies and traditional methodologies harmonize with the central tenets of the Christian faith). The basis for comparison was their aptitude for and achievement in a college-level English course.
Review of the Literature
Research investigating the success of home education has emphasized academic achievement on both the elementary and secondary levels. The following studies show that home schooled students from all regions of the country perform consistently above national norms.
Greene (1984) conducted a study in Alaska that compared the scores on standardized achievement tests by 193 home schooled students in grades 1 through 8 with the academic performances of their public schooled counterparts. In this case, the home schooled students “consistently outperformed” the other students in all the academic subtests (p. 18). Rakestraw (1987) conducted a study in Alabama that produced similar results. The 84 home schooled, elementary-aged students in this study performed above national norm levels on all the standardized achievement tests. Frost (1987) conducted a study in Illinois that reported the scores of 74 home schooled students of elementary age and found that they scored higher than the national norms in all subtests of a nationally recognized standard achievement test. Tipton (1990) investigated the success of 81 home schooled children in West Virginia and found that they consistently performed significantly higher than their public school counterparts in vocabulary, reading comprehension, mathematics concepts, science, and total language means.
Wartes (1990) summarized the results of four years of home school testing in Washington state. The results based on 2,911 sets of test scores showed that in each of the four years of testing the median score of home schooled students fell well above the national norm. Richman, Girten, and Snyder (1990) replicated Wartes’ study in Pennsylvania. The 171 home schooled students who were tested consistently scored well above the national norm for the same tests. The first nationwide study of academic achievement among home schooled students was conducted by Ray (1990) and involved 1,471 students. The data collected on all grade levels (K-12) showed that the home schooled students scored at or above the 80th percentile on all subtests, again well above the national average. As Taylor (1992) claims, “There is yet to be found any study which suggests that, as a group, home schoolers perform below average on any kind of academic measure” (p. 25).
However, very little research has investigated the results of home education by examining its product: the adult who has been home schooled. What happens to these students after they complete the home school experience? Can they succeed in the post-secondary academic environment away from their parents and in a more traditional educational setting?
Ray (1990) included a survey of home schooled adults in his nationwide study mentioned earlier. Of the 99 subjects who responded to his survey regarding post-high school activities, 50.5% reported that they had attended either a junior college or a four-year college after completing their high school requirements. These results substantiate Toch’s (1991) claim that approximately 50% of students who have been home schooled go on to attempt college-level academic work.
Barnebey (1986) investigated university admission requirements for home schooled applicants and found that different schools had varying expectations of home schooled students “with Private universities being more likely to accept home schooled applicants than the Public universities” (p. 142). Interestingly, the survey uncovered varying attitudes among the admissions officers regarding the potential for success of the home schooled applicants. Of the admissions officers from non-accepting universities, 70.4% believed home schooled students would be less successful than conventionally educated students and that 73% advised home schooled applicants to attend a junior college or community college before attempting university-level work. Obviously then, there is some doubt in the minds of some educators that home schooled students can compete on this advanced academic level.
Two recent studies have included college students who have been home schooled. Oliveira (1994) investigated the differences in critical thinking skills among college freshmen who had graduated from different types of high schools, including home schools, conventional Christian schools, public schools, and Accelerated Christian Education (ACE) schools. He found no significant differences among these groups on any of the measures of these skills. Mexcur (1993) compared the college-level academic achievement of freshman college students who had graduated from public schools, conventional Christian schools, Accelerated Christian Education (ACE) schools, and home schools. Comparing American College Testing Program (ACT) scores, final semester grades for two freshman level English courses, and both the first and second semester grade point averages, Mexcur concluded that “graduates of all types of high schools achieve at very similar academic levels” (p. 110). However, because his study used only 10 students who had been home schooled, the results contribute little to the body of research on this topic.
Therefore, existing literature in this field reveals that the investigation of home schooled adults’ potential for academic success in college is a legitimate area of inquiry. Thus, the purpose of this study was to determine differences in academic preparedness as well as academic achievement among college students who had attended home schools, public schools, and conventional Christian schools.
The subjects of this study included 180 first-time freshman students enrolled during the 1992-93 and 1993-94 academic years at a large, private Christian university located in the Southeast. These students were divided into three groups based on the type of high school they had attended: (a) home school graduates (n=60), (b) public school graduates (n=60), and (c) conventional Christian school graduates (n=60). The home school group consisted of all of the available freshman students who had completed their entire high school education in a home school and who had taken the English course that was the focus of the study. The other two groups resulted from a random sample selected from the available population of freshman students who had completed all of their high school education in either a public school or a conventional private school and who had taken the English course required for this study. Chi-square analyses were used to determine that the groups did not differ significantly (at an alpha level of .05) on any of the selected demographic variables (See Table 1).
–This table is missing in this web page presentation; you may find the PDF format elsewhere on this web site with the table included–
Data for this study were collected from the academic records for both the 1992-93 and 1993-94 academic years at the university that was the site for this study and from freshman English classes in the same academic years.
The aptitude for college level English was measured by the American College Testing Program (ACT). Administered to the majority of college-bound juniors and seniors in 38 states, it is the more popular of the two major college entrance examinations offered nationally (Brasel, 1991). The ACT “measures educational development in English, mathematics, reading, and science reasoning. Scores for each of the four subtests are reported, as well as a composite score” (p. 5), the mathematical average of the four subtests (Brasel, 1991). The complete ACT assessment program is constructed “to measure as directly as possible [the] mastery of knowledge and skills required for success in college studies” (ACT Assessment User Handbook, 1991, p. 4). This study used the scores from the English subtest and the composite scores for all four subtests. Several studies have shown the ACT composite score to be a reliable predictor of first year college student grade point averages (Dvorak, 1989; Nelson, 1985; Schade, 1977; Snyder & Elmore, 1983; Thornell & Jones, 1986) and the subscores for the specific tests to be valuable in predicting specific course performance (Bauer, 1987; Goldman & Slaughter, 1976; Schade; 1977).
The specific English course chosen as the focus of this study was a required course for all of the students in the study. Entitled Composition and Rhetoric, it is an introduction to academic writing emphasizing argumentation, research, documentation, and style, centering on the production of a library research paper. The course emphasizes skills that are intended to enable students to communicate effectively in their college work beyond this specific English course. The course is also representative of most college courses in that it is conducted in a lecture format, a potential area of concern for home educated students who have depended on individualized instruction rather than a structured educational situation in which they, along with scores of other students, must listen to an oral presentation of material, take notes, and assimilate information over which they will be tested at a later date. It is also within this structured situation that the students must develop the skills necessary to succeed in this course and in their other college-level courses.
The instruments used to measure academic achievement in this study consisted of measures of composition skills and usage skills mastered in this required English course. The composition skills were measured by the grade earned on the major writing assignment for this course (the second draft of a library research paper) and the number of correct responses to multiple choice items that deal with principles of composition taken from the quizzes, tests, and final exam constructed by the professor of the course (a total of 162 multiple choice items). The composite composition scores, therefore, were a combination of the multiple choice items and the library research paper score. The usage skills were measured by the number of correct responses to multiple choice items that test knowledge of vocabulary taken from the quizzes, tests, and final exam (a total of 144 items) and the number of correct responses to multiple choice items that test knowledge of grammatical principles taken from the quizzes, tests, and final exam (a total of 64 items). The composite usage score then was a combination of the scores on multiple choice items reflecting vocabulary and grammar.
Research Design and Statistical Analysis
The independent variables for this study were the prior high school setting in which the subject was educated (either home school, conventional Christian school, or public school). The dependent variables were the norm-referenced test scores reflecting aptitude for a college-level English course (specifically the relevant ACT test scores) and the criterion-referenced scores reflecting achievement in skills necessary for success in a college-level English course (specifically the library research paper as well as the quiz and test scores reflecting composition and usage skills). Because of similarities among the dependent variables, they were divided into three meaningful sets: (1) the measures reflecting aptitude for college English: the ACT English subtest scores and the ACT Composite test scores; (2) the measures reflecting achievement in composition skills: the grades on the library research paper and the scores on test items that measure knowledge of composition principles; and (3) the measures reflecting achievement in usage skills: the scores on test items that measure vocabulary skills and the scores on test items that measure grammar skills.
The purpose of this study was to determine the differences in aptitude for and achievement in college English among home school graduates, conventional Christian school graduates, and public schools graduates. Specifically, this study addressed the following questions:
1. Do the groups of college freshmen differ significantly on the variables dealing with aptitude for college English?
2. Do the groups of college freshmen differ significantly on the variables dealing with composition skills learned in a college level English course?
3. Do the groups of college freshmen differ significantly on the variables dealing with usage skills learned in a college level English course?
Results and Findings
Because this research design studied the effects of three independent variables on six dependent variables that were clustered into three sets, the multivariate analysis of variance (MANOVA) was deemed an appropriate statistical test for this study. Three separate MANOVA tests were run to reflect the three sets of dependent variables: (1) one MANOVA analyzed the relationships of the ACT English subtest scores and the ACT composite test scores on the three independent variables; (2) a second MANOVA analyzed the relationships of the research paper grades and test items reflecting composition skills on the three independent variables; and (3) a third MANOVA analyzed the relationships of the test scores reflecting vocabulary and grammar skills on the three independent variables. A .05 alpha level of significance was used.
Resulting Wilks’ Lambdas were followed with univariate statistics to determine which specific dependent variables the groups of freshman students differed on. The Tukey HSD Multiple Comparison test was used to determine which groups differed on the particular dependent variable. The statistical program SYSTAT (Wilkinson, 1989) was used for all computations and analysis of data.
Research Question One
Table 2 lists the means and standard deviations of the ACT English subscore and the ACT composite score for the three groups of college freshman students. The MANOVA (See Table 3) revealed a significant difference among the groups of freshmen in their mean scores reflecting aptitude for college English (Wilks’ Lambda(4, 352)=.88, p=.02). Follow-up univariate tests (See Table 4) showed that only one of the two variables (ACT English subtest score and ACT composite score) had a significant F, which was the ACT English subtest score (F(2, 177)=4.22, p=.02). The univariate tests were followed by a Tukey HSD Multiple Comparison (See Table 5), which revealed that the significant difference on the ACT English subtest variable existed between the home school graduates and the conventional Christian school graduates. The home schooled students (M=24.73) scored significantly higher than the Christian school students (M=22.43) on the ACT English subtest.
Research Question Two
Table 2 lists the means and standard deviations of the scores on the library research paper and the test scores reflecting writing principles for the three groups of freshman college students. The MANOVA (See Table 6) revealed no significant difference among the groups of freshmen in their mean scores on the library research paper or the test items reflecting writing principles.
Research Question Three
Table 2 lists the means and standard deviations of the scores on test items reflecting vocabulary skills and grammar skills for the three groups of college freshman students. The MANOVA (See Table 7) revealed no significant difference among the groups of freshman students in their mean scores on test items reflecting vocabulary skills and grammar skills.
–These tables are missing in this web page presentation; you may find the PDF format elsewhere on this web site with the tables included–
Discussion of Findings
The purpose of this study was to determine the differences in academic aptitude for and achievement in a college-level English course among home school graduates, public school graduates, and conventional Christian school graduates. Based on the results of this investigation, one may conclude that home schooled students demonstrate similar academic preparedness for college and similar academic achievement in college as students who have attended conventional schools.
Aptitude for College-Level English
The statistical analysis revealed that the home schooled students’ ACT composite scores were comparable with the scores of the other groups of students; however, the home schooled students scored significantly higher on the ACT English subtest than did the conventional Christian school graduates. These findings are consistent with the existing research involving the performance of home schooled students on standardized tests taken before they reach college-level study. Almost without exception, home schooled students in both elementary and secondary levels perform as well as, if not better than, their conventionally educated counterparts in both statewide and national comparisons.
In light of the research that identifies the predictive value of the ACT test, these home schooled students seem to demonstrate an aptitude for success in college, specifically in college-level English. Schade (1977), Snyder and Elmore (1983), and Thornell and Jones (1986) all concluded from their studies the value of the ACT subtest scores and the ACT composite score in predicting academic achievement in college. Although the ACT scores may not be helpful in predicting success in specific courses, they do seem to serve as a predictor of general academic achievement. Therefore, this study suggests that home schooled students are as likely as conventionally educated students to succeed academically in college-level work. In fact, their ACT scores reveal that these home schooled students may be more likely to succeed than the students who had graduated from conventional Christian schools.
Achievement in Writing Skills
The second MANOVA revealed no significant difference among the groups of students in their mean scores on the library research paper or the quiz and test items reflecting comprehension of writing principles. Even though the results were not significant, the groups differed in their mean scores on writing variables: (1) the library research paper: home school graduates, M=48.83; public school graduates, M=48.40; conventional Christian school graduates, M=48.32; (2) the test items reflecting writing principles: home school graduates, M=115.27; public school graduates, M=115.02; conventional Christian school graduates, M=114.23.
These findings indicate that the home schooled students in this study demonstrate similar abilities in writing skills on the college level as conventionally educated students. Because writing skills are foundational to success in most college courses, it is likely then, that these home schooled students are just as capable as conventionally educated students to succeed in that aspect of their college courses.
Achievement in Usage Skills
The third MANOVA revealed no significant difference among the groups of students in their mean scores on the quiz and test items reflecting vocabulary and grammar skills. Although not significant, the groups differed in their mean scores across usage variables: (1) the test items reflecting vocabulary skills: home school graduates, X=112.13; public school graduates, X=112.03; conventional Christian school graduates, X=113.50; (2) the test items reflecting grammar skills: home school graduates, X=47.40; public school graduates, X=47.03; conventional Christian school graduates, X=46.50.
These findings demonstrate a similar command of the English language among these home schooled students and conventionally educated students. Foundational not only to writing skills but also to speaking and reading skills, the verbal acuity reflected by a command of academic vocabulary and correct grammar is undoubtedly instrumental in a student’s success in college. This study reveals that these home schooled students are just as capable as conventionally educated students to achieve the verbal skills necessary to perform well in college. In addition, the results of the students’ performances in this particular course demonstrate similar levels of academic achievement on the measures selected to determine success in the course: namely, papers, quizzes, and tests. The home schooled students performed as well as, if not better than, their conventionally educated counterparts.
Home education has become in recent years an educational alternative for many families in a variety of circumstances. As a relatively new, yet significant, force in modern education, home schools have attracted the attention of educators, legislators, and even the general public. Not all of this attention, however, has been in admiration of the accomplishments of home educators. Many doubt, and even contest, the legitimacy of this educational alternative.
Research in the field of academic achievement, however, supports the home schoolers’ claim that their strategy is working. Lines (1986, p. 11) summarizes this research by reminding educators that “high test scores provide no evidence that all children would benefit from the experience of home instruction ….. The test data are evidence only that, in general, children in home-schools are having successful experiences.” Certainly, home schooling is not appropriate, nor possible, for all students.
Yet, based on the results of this investigation, one may conclude that the home schooled students in this study demonstrate similar academic preparedness for college and similar academic achievement in college as the students who have attended conventional schools. These findings are consistent with the existing research that reveals that home schooled students perform as well as, if not better than, conventionally educated students on academic achievement measures on the elementary and secondary levels. It seems that home schooled students can continue to perform adequately in the different, and more advanced, academic setting of college-level study.
ACT assessment user handbook: 1991-92. (1991). Iowa City, IA: American College Testing Program. (ERIC Document Reproduction Service No. ED 338 715)
Adelman, C., & Reuben, E. (1984). Starting with students: Promising approaches in American higher education. Washington, DC: National Commission on Excellence in Education. (ERIC Document Reproduction Service No. ED 257 411)
American College Testing Program (1991). Enhanced ACT assessment program: 1991. Iowa City, IA: American College Testing Program.
Barnebey, L. F. (1986). American university admission requirements for home-schooled applicants. Doctoral dissertation, Brigham Young University, Salt Lake City, UT.
Bauer, A. J. (1987). Academic performance in English: A report and comprehensive examination of high school grades, achievement test scores, college level placement, and subsequent college course grades. (ERIC Document Reproduction Service No. ED 292 827)
Bingaman, D. E. (1989). The relationship of cognitive and affective variables to college grade point average among highly successful high school graduates. (ERIC Document Reproduction Service No. ED 320 512)
Brasel, M. D. (1991, January). New Mexico Enhanced ACT and SAT results: School year 1989-1990. New Mexico State Department of Education, Santa Fe, NM. (ERIC Document Reproduction Service No. ED 334 219)
Dvorak, J. (1989, Fall). Publications experience as a predictor of college success. Journalism Quarterly, 66, 702-706.
Frost, E. A. (1987). A descriptive study of the academic achievement of selected elementary school-aged children educated at home in five Illinois counties. (Doctoral dissertation, Northern Illinois University, 1987). Dissertation Abstracts International, 48(7), 1589A.
Goldman, R. D., & Slaughter, R. E. (1976, February). Why the college grade point average is difficult to predict. Journal of Education Psychology, 68(1), 9-14.
Greene, S. S. (1984, April). Home study in Alaska: A profile of K-12 students in the Alaska Centralized Correspondence Study Program. (ERIC Document Reproduction Service No. ED 255 494)
Hoyt, D. P. (1968, October). Generalized academic prediction in four-year colleges. Personnel and Guidance Journal, 47(2), 130-136.
Hudson, J. B. (1989, July 21). An analysis of ACT scores,placement tests, and academic performance in reading, English, and mathematics courses. Louisville, KY: Louisville University. (ERIC Document Reproduction Service No. ED 334 916)
Kallingal, A. (1989). Critical factors in college level course achievement. (ERIC Document Reproduction Service No. ED 305 843)
Lazere, D. (1982). Composition for critical thinking: A course description. Washington, D. C.: National Endowment for the Humanities. (ERIC Document Reproduction Service No. ED 273 959)
Lines, P. M. (1986, November). Home instruction: An overview, policy issues paper. Charleston, WV: Appalachia Educational Lab. (ERIC Document Reproduction Service No. ED 296 453)
Lines, P. M. (1987, March). An overview of home instruction. Phi Delta Kappan, 68(7), 510-517.
Maarse Delahooke, M. (1986). Home educated children’s social/emotional adjustment and academic achievement: A comparative study. (Doctoral dissertation, California School of Professional Psychology, 1986). Dissertation Abstracts International, 47(2), 475A.
McAvoy, R. (1986, october). Home-bound schools: An alternative for rural areas. Paper presented at the Annual Conference of the National Rural and Small Schools Consortium, Bellingham, WA. (ERIC Document Reproduction Service No. ED 280 633)
Mexcur, D. (1993). A comparison of academic performance among public school graduates, conventional Christian school graduates, accelerated Christian education graduates, and home school graduates in three Christian colleges. Unpublished doctoral dissertation, Bob Jones University, Greenville, SC.
Nelson, L. F. (1985). The study of freshman retention and withdrawal at Bob Jones University. Unpublished doctoral dissertation, University of South Carolina, Columbia, SC.
Nice, T. N. (1992, October). Home schooling. The Education Digest, 58, 37-38.
Oliveira, P. C. M. (1994). Differences in selected critical thinking skills among Christian college students who graduated from various high school settings. Doctoral dissertation, Bob Jones University, Greenville, SC.
Pantages, R. J. & Creedon, C. F. (1978). Studies of college attrition: 1950-1975. Review of Educational Research, 48(1), 49-101.
Pascarella, E. T. (1987, November). The development of critical thinking: Does college make a difference? Paper presented at the Annual Meeting of the Association for the Study of Higher Education. (ERIC Document Reproduction Service No. ED 292 417)
Raburn, J., & Van Schuyver, B. (1984). The relationship of reading and writing to thinking among university students taking English composition. (ERIC Document Reproduction Service No. ED 273 978)
Rakestraw, J. F. (1987). An analysis of home schooling for elementary school-age children in Alabama. (Doctoral dissertation, University of Alabama, 1987). Dissertation Abstracts International, 49(4), 725A-726A.
Ray, B. D. (1990, November 16). A nationwide study of home education: Family characteristics, legal matters, and student achievement. Paeonian Springs, VA: National Center for Home Education.
Ray, B. D. (1992). Home education in Oklahoma: Family characteristics, student achievement, and policy matters. Salem, Oregon: National Home Education Research Institute.
Ray, B. D. (1993). Learning at home in North Dakota: Family attributes and student achievement. Salem, Oregon: National Home Education Research Institute.
Richman, H. B., Girten, W., & Snyder, J. (1990). Academic achievement and its relationship to selected variables among Pennsylvania homeschoolers. Home School Researcher, 6(4), 9-16.
Schade, H. C. (1977, November). The ability of ACT and MCPT to predict the college grade point average. Neosho, MO: Crowder College. (ERIC Document Reproduction Service No. ED 188 683)
Schmidt, S. J. (1989). A report of North Carolina home school parents on certain variables associated with social studies instruction. Unpublished masters thesis, Wake Forest University. (ERIC Document Reproduction Service No. ED 320 798)
Snyder, V. & Elmore, P. B. (1983, April). The validity of the ACT and descriptive tests of language skills for developmental students over a four-year college program. (ERIC Document Reproduction Service No. 228 309)
Taylor, J. W., 5th. (1992, Autumn). The self-concept of home schoolers: a nationwide study. Journal of Research on Christian Education, 1(2), 23-34.
Thornell, J., & Jones, R. (1986, November). The college admissions equation: ACT scores versus secondary school grade performance. Paper presented at the Annual Meeting of the Mid-South Educational Research Association, Memphis, TN. (ERIC Document Reproduction Service No. ED 278 687)
Tipton, M. (1990). An analysis of home-schooled children’s comprehensive test of basic skills results and demographic characteristics of their families. Unpublished masters thesis, Antioch University. (ERIC Document Reproduction Service No. ED 336 208)
Toch, T. (1991, December 9). Schooling in family values. U. S. News & World Report, 111, 73-74.
Wartes, J. (1990). Recent results from the Washington homeschool research project. Home School Researcher, 6(4), 1-7.
Wilkinson, L. (1989). SYSTAT: The system for statistics. Evanston, IL: SYSTAT, Inc.
Wood, P. H. (1988, September 17). Predicting college grades and helping colleagues to assist poor readers to succeed in college courses. (ERIC Document Reproduction Service No. ED 309 699; 1989).