Homeschooling Information and Resources
all about homeschooling - from Home Education Magazine

HEM can
be found in
thousands of
public libraries.
Is it in yours?

Do you have a question?
We have several answers!


Search HEM's 10,000+ page knowledge base.


HEM Subscriptions


For over 28 years every issue of HEM has been full of help, guidance, validation, and support.

We are proud of defending the 1st amendment and standing up to a frivolous lawsuit, however, this civil liberties exercise temporarily ground HEM to a halt, we are coming back strong with the May-June/12 issue.

HEM Subscriptions

One Year Sub $12.95

One Year Sub $26.00

Digital & Print
One Year Sub $36.00




Home Education Magazine

July-August 1999 - Columns

Taking Charge HSLDA Study: Embarrassing and Dangerous - Larry and Susan Kaseman

Suppose you heard about a new study that claimed to prove that homeschoolers do well academically, and, in fact, perform much better than students who attend public school. Would you be glad to have your doubts about homeschooling put to rest once and for all? Would you be eager to share a copy with Aunt Lucy who asks all those awkward questions? Would you want another copy for State Senator Get-um who's threatening to introduce a bill requiring that all homeschoolers take state-mandated tests and send the results to the state?
A new study designed, financed, and promoted by the Home School Legal Defense Association (HSLDA) might at first glance seem to be just that study. But hold on to your horses. A little investigation will show that this study reports on the academic achievements of only a very narrow group of homeschoolers and is not at all representative of the homeschooling community as a whole, so its findings do not apply to homeschoolers in general. If Aunt Lucy is at all perceptive, she will realize that this study is poorly done and draws unwarranted conclusions. In fact, her opinion of homeschoolers may even take a nose dive if she thinks this study represents the kind of thinking and supposed research that homeschoolers typically do or are associated with. And Senator Get-um? Don't send him the study! It will play right into his hands by giving him just the "information" he's been looking for to support his testing bill.
"What's going on here?" you may be wondering.

A study of test scores and some demographic characteristics (such as income and parents' education) of a narrow segment of the homeschooling community was recently designed, financed, and promoted by HSLDA. Both common sense and an analysis based on criteria used by social scientists show that this study is seriously flawed. Despite this fact, most of the report of the study, including its "Summary of Major Findings," says the findings apply to homeschoolers without qualification. Part II of this column examines the problems with the study and the report.

Okay, but can't we simply ignore a study that is so obviously and ridiculously flawed? Unfortunately, no, for several reasons. One, the inaccuracies either are not obvious to or are being overlooked by the social scientists who participated in the peer review of the study, the journal that published it, and newspapers such as The Washington Post and the Christian Science Monitor that are reporting on it.

But even more serious, this study has the potential to undermine our homeschooling freedoms by presenting a very narrow and limiting stereotype of what a homeschooler should be and by providing assistance to critics of homeschooling who want to increase the regulation of homeschoolers by increasing the testing requirements for homeschoolers in states that already require testing and by requiring testing in states that currently do not require it. Therefore, even if you are not interested in the analysis of the study in Part II, please read Part III. How This Study Undermines Homeschooling Freedoms and Part IV. What We Can Do To Minimize the Damage Done By This Study.

(Note: Of course, we homeschoolers know from observing our own children and other homeschoolers that homeschooling works and homeschoolers generally do well academically. Criticisms of this study are not to say that homeschoolers whose parents decided to have them tested or who were required to take tests, wouldn't score well. After all, as homeschoolers we can help our children prepare for tests and teach to the test just as conventional school teachers do, and we are dealing with many fewer students.)

Part I. Background: The Study and How It Came About

Parents of homeschoolers who took the Iowa Achievement Tests through Bob Jones University Press Testing and Evaluation Service in the spring of 1998 were asked to complete a two-page questionnaire covering topics such as family income; number of children; parents' "highest academic attainment," marital status, and religious preference; race/ethnic background; students' computer use, library visits, and television viewing; etc. Data from questionnaires were correlated with test scores when possible (there were some administrative problems).

The results were published online as "Scholastic Achievement and Demographic Characteristics of Home School Students in 1998," by Lawrence M. Rudner, in Education Policy Analysis Archives, Vol. 7, No. 8, March 23, 1999, (The report is referred to as "Rudner, 1999" in this column. Page numbers assume the report has been downloaded in 12 point Bookman using 7.5 inch columns.) The report was publicized at an Online Press Conference on March 23, 1999. Transcripts of both the Opening Remarks and Chat Session are available at HSLDA's web site: (Page numbers in this column are based on the PDF layout.)

In light of the serious problems with the study and the ways in which it undermines homeschooling freedom (both of which are discussed below), one would assume that it was created and conducted by critics or opponents of homeschooling. A review of the history of the study shows HSLDA was responsible for it.

* HSLDA arranged with BJU Testing Service to use homeschoolers who ordered Iowa tests as the target population for a study, to gain their agreement to participate in the study, to have the Service send HSLDA's questionnaire along with the tests to families who had agreed, and to have the Service receive the test answer sheets and completed questionnaires and forward them to other organizations for scoring and data entry.

* HSLDA designed the questionnaire.

* HSLDA wrote and signed the letter asking parents to complete the questionnaire. The letter included a P.S.: "If you have ANY questions about this survey, please contact HSLDA (not BJUP)."

* "Home School Legal Defense Association gave a grant to Dr. Larry Rudner to study the demographic characteristics and achievement levels of home schooled students," according to Michael Farris, president of HSLDA, in an online press conference on March 23, 1999. (Opening Remarks, p. 1)

* In the Acknowledgments of the report, Rudner writes, "This report relied on the creativity and expertise of several individuals. Michael Farris, Esquire, President of the Home School Legal Defense Association conceived the study and secured the cooperation of the author and three contributing organizations - Bob Jones University Testing Service, National Computer Systems, and HSLDA. Earl Hall of HSLDA worked out the numerous logistics and details of working, was the primary architect of the background questionnaire, and provided responses to my thousand questions regarding home schooling." (p. 29)

* HSLDA held an online press conference to publicize the report on March 23, 1999; distributed the findings to members of the U. S. Congress; and gave interviews about the study.

Part II. What Is Wrong With This Study?

Obvious Problems with the Sample

Both common sense and the criteria used by social scientists tell us that a study cannot draw conclusions about homeschoolers in general unless it either includes all homeschoolers or is based on a representative sample of homeschoolers. However, this study was based on an unrepresentative sample: homeschoolers who took the Iowa tests through BJU Testing Service in the spring of 1998, whose parents completed the questionnaire, and whose questionnaires were not eliminated because of administrative problems described below.

The only homeschoolers in the study were the following:

(1) Only homeschooling families who wanted to have their children take standardized tests were included. This means that the study automatically excluded the many homeschooling families who do not have their children take standardized tests because they do not want their children subjected to such tests; or because they feel that such tests are biased, unfair, inaccurate, or meaningless; or because they do not want to spend the time, money, and effort the tests require; or for other reasons. Each family should certainly have the right to decide whether or not to have their children take standardized tests. But since the study was limited to families that have their children take standardized tests, a large number of homeschoolers were omitted from the study, and even more important, a whole element of the population (those who don't have their children take tests) was omitted. It cannot be assumed that homeschoolers who don't take tests do not differ from those who do.

(2) More specifically, only families who wanted to have their children take Iowa tests were included. Any family that chose a different test, for whatever reason, was excluded. Again a whole element was excluded.

(3) Even more specifically, only families who used the BJU Testing Service were included. Because BJU has a strong religious affiliation, it seems reasonable to assume that people who share this affiliation would be more likely to have their children tested through BJU while people who do not share this affiliation might choose a different testing service.

(4) Only qualified administrators can give Iowa tests through BJU Testing Service. According to the "1999 Catalog and Order Form," to qualify, a person must "1. Be a teacher who has been certified by a state department of education to teach in a public school or other conventional school (home school or local association certification does not qualify) OR 2. Have a bachelor's degree OR 3. Be or have been a full-time academic classroom teacher in a conventional school (not a home school)." (p. iii) Obviously, families in which a parent does not qualify to administer the tests can find someone who does. But testing is easier if a parent qualifies. Therefore, the sample is likely to contain a higher percentage of parents who are certified teachers or have bachelor's degrees or teaching experience than the general homeschooling population.

(5) Participation in the study was voluntary even within the very narrow group of people targeted for study. That is, families of children who took the Iowa tests through BJU Testing Service were not required to complete the questionnaire.

Farris' letter to parents soliciting participation in the study said, "While participation in this survey is strictly voluntary and is not a required part of your child's test, it is ONLY by your willing participation that we can realize the maximum impact of this study, proving that parental involvement and hard work are the keys to academic success. Research projects such as this promote home schooling as a viable-and often superior-method of education. Please help 'make the case' for current and future home schoolers by taking a few moments to fill out this survey." Parents who did not want to support HSLDA's study probably would not have participated. Also, parents who wanted to support HSLDA's study but thought their children might not score well on the tests might have decided not to participate, so that their children's lower scores would not detract from the results.

(6) The sample probably included a higher percentage of homeschoolers who are or have been associated with HSLDA than does the general homeschooling population. This is partly because in 1993, "BJU Press assumed the testing that the Home School Legal Defense Association had been providing for its members." (Rudner, 1999, p. 3) Also, BJU Testing Service gave a discount to HSLDA members.

In the examples given above, it seems likely that the homeschoolers who were excluded from the study differ in important ways from those who were included. For example, families who choose to have their children take standardized tests are likely to have a different approach to learning, curriculum, and other aspects of homeschooling than families who choose not to have their children tested.

(7) Although parents of 39,607 students had reportedly agreed to complete the questionnaire, the final report included data on only 20,790. In other words, 48% were not included in the findings of the report. No doubt some parents did not follow through and fill out the questionnaire. However, there were also major administrative problems. To protect families' privacy, parents were instructed not to put their names on the questionnaire. Test administrators were supposed to keep each child's questionnaire with his or her answer sheet and send them together to BJU Testing Service, so eventually data from the questionnaires could be correlated with test scores. However, we talked with an employee at BJU Testing Service who said he was "intimately familiar" with the data collection for the HSLDA study. This employee said that there were a variety of problems in getting questionnaires matched with students' answer sheets. Perhaps the largest problem was that some test administrators separated the test answer sheets from the questionnaires and sent in two separate stacks. This reportedly happened more frequently when large numbers of students were being tested by a single test administrator. These questionnaires could not be included in the study because there was no way to correlate them with test scores.

To understand more concretely how having such an unrepresentative sample leads to inaccurate findings, let's consider why the report concluded, "Almost one out of every four home school students (23.6%) has at least one parent who is a certified teacher." (p. 12)

* Giving children the Iowa tests through BJU Testing Service is easier for parents who are certified teachers (or have a bachelor's degree or have taught in a conventional school) than it is for parents who have to find someone else to administer the tests, make special arrangements, and perhaps pay the administrator. (Contrary to what Farris said in the press conference on March 23, 1999, qualified parents could administer Iowa tests to their own children without persons outside their immediate family being present.)

* Because teacher training stresses the value and importance of standardized tests. certified teachers are probably more likely to want their children to take standardized tests than are many parents who are not certified teachers. Parents who are certified teachers may want test scores as confirmation of their teaching.

* As described above, many of the questionnaires submitted were separated from the answer sheets. This happened more often when tests were being administered to a group of children from different families, probably often because the children's own parents were not certified teachers or otherwise qualified to administer the tests. Many parents who were certified teachers or otherwise qualified would have administered the tests to their own children and mailed in only their family's answer sheets and questionnaires, minimizing the possibility of a mix-up. Therefore, questionnaires were more likely to be included for children with a parent who is a certified teacher and eliminated (because of the administrative problems) for some children whose parents are not certified teacher. This contributed to the unrealistically large figure for the percentage of children with at least one parent who is a certified teacher.

Failure of the Study to Meet Criteria of Social Scientists

Great care and precision are required to develop and conduct accurate surveys. To make this process more manageable, social scientists have developed criteria for determining accuracy and ways of avoiding errors. Priscilla Salant and Don A. Dillman's book How to Conduct Your Own Survey, a widely respected resource on survey research, outlines four major errors that can easily lead to inaccurate surveys and ways in which they can be avoided. Let's evaluate this study in terms of these criteria.

(1) "Coverage error occurs when the list-or frame-from which a sample is drawn does not include all elements of the population that the researchers wish to study." (p. 16) Obviously, surveys of homeschoolers will have very serious coverage error because there is no way to come up with a good list that includes all elements of the homeschooling population or even a list that includes more than a small fraction of the total homeschooling population.

This study has particularly serious coverage error. The list-or frame-from which the sample was drawn was limited to homeschooling families who had their children take Iowa tests through BJU Testing Service in the spring of 1998 and agreed to complete a questionnaire. This list certainly "does not include all elements of the population that the researchers wish to study." It does not include any homeschoolers who decided not to have their children tested, or who chose a different test or a different testing service, or who did not agree to complete the questionnaire. (The report is unprofessionally vague about how the sample was obtained. However, according to the employee of BJU Testing Service with whom we talked, parents of the 39,607 children had agreed in advance to complete the questionnaire.) These are major omissions for a study that claims to be about homeschoolers.

(2) "Sampling error occurs when researchers survey only a subset or sample of all people in the population instead of conducting a census." (p. 17) Sampling error is inevitable unless a census is conducted in which every member of the population is contacted. However, sampling errors can be minimized if enough people are "sampled randomly to achieve the needed level of precision." (p. 15)

Although the sample selected was woefully inadequate, an attempt was made to have everyone in the sample complete the questionnaire. Again, it is not clear why 48% of the questionnaires were not included in the final report, although administrative problems discussed above account for part of this. Since the excluded questionnaires were "selected" on the basis of administrative problems and other factors and were not randomly selected, this study has sampling errors. In any case, avoiding sampling error would not have compensated for the extraordinary coverage error discussed above.

(3) "Measurement error occurs when a respondent's answer to a given question is inaccurate, imprecise, or cannot be compared in any useful way to other respondents' answers." (p. 17) For a survey to avoid measurement error, "Clear, unambiguous questions would be asked so that respondents are both capable of and motivated to answer correctly." (p. 15)

Questions on the questionnaire were generally clear, so measurement error was probably not a serious problem. However, the questions about parents' "highest academic attainment" included "Graduated from college with master [sic] degree" and "Graduated from college with doctorate degree." This is confusing since universities rather than colleges award master's degrees and doctorates.

(4) "Nonresponse error occurs when a significant number of people in the survey sample do not respond to the questionnaire and are different from those who do in a way that is important to the study." (p. 20, authors' italics)

Not enough information is given about why only 52% of the sample was included in the report to assess whether nonresponse error occurred. If nonresponse error was not a problem, the researcher should have made this clear.

Therefore, in addition to the obvious factors discussed above that make the sample so unrepresentative as to render the study worthless, the study does not stand up when judged by the criteria used by social scientists. And yet Farris claimed that the report was published in a "very prestigious peer-reviewed online journal." (Press conference, March 23, 1999, Opening Remarks, p. 1) One can only wonder how a study with such major problems and inadequacies survived the peer review and was published. Perhaps some social scientists need a homeschooled child to shout, "The emperor has no clothes."

Obviously, with a sample so limited, the findings are of no value. The most that can be said is that the findings apply to the 20,760 homeschooled children who took the Iowa tests through BJU Testing Service in the spring of 1998, whose parents completed the questionnaire, and who were not among the 48% not included in the report. It cannot even be said that the findings apply to all homeschoolers who took the Iowa tests through the BJU Testing Service in the spring of 1998 and whose parents completed the questionnaire.

Actually, the survey is worse than worthless because the findings give very misleading and inaccurate information about homeschooling and undermine our homeschooling freedoms. (See Part III.)

Problems with the Report

* The researcher does not seem to understand how serious the coverage error was and how unrepresentative the sample was.

During an online press conference with the media on March 23, 1999, Rudner said, "I wound [sic] rather have studied a random sample of 3,000 students. Unfortunately that is not possible. We don't have a well defined [sic] universe of home school students from which to draw a random sample. The fact that this was a large sample assures accurate estimates." Shortly thereafter, he said, "I only claim to generalize to groups like the one Studied [sic]. There is nothing to indicate that the group is atypical. The fact that it was large assures good estimates." (Chat Session, pp. 2, 3) (Note: Typos are common and understandable in online press conferences. They are included here simply so the reader can read exactly what was on the Internet.)

Rudner admits that a random sample of homeschoolers would have been better than the sample he used. But he seems unaware of the major problems with the sample he used. For example, at the beginning of his report, he says, "While the academic levels of home school students are described in terms of public and private school norms, this study is not a comparison of home schools with public or private schools. . . . This study seeks to answer a much more modest set of questions: Does home schooling tend to work for those who chose [sic] to make such a commitment? That is, are the achievement levels of home school students comparable to those of public school students? Who is engaged in home schooling? That is, how does the home school population differ from the general United States population?" (P. 2) Rudner fails to either realize or acknowledge that these questions could not possibly be answered using this sample.

How could Rudner possibly say, "There is nothing to indicate that the group is atypical." As explained above, the sample is unrepresentative in many ways. Also, it is the responsibility of a researcher to demonstrate that a sample is representative, not to simply proclaim that it is not atypical.

More disconcerting is Rudner's claim that the study was accurate because the sample was large. (Actually the 20,000 participants represent only about 2% of the estimated 1,000,000 homeschoolers in the U.S.) What Rudner fails to acknowledge is that even having a relatively large sample is not enough to overcome serious coverage error. This is a pretty basic point for a researcher to fail to communicate.

Rudner does point out that this is not a controlled study. His report says, "This study does not demonstrate that home schooling is superior to public or private schools. It should not be cited as evidence that our public schools are failing. It does not indicate that children will perform better academically if they are home schooled." (p. 27) But that is very different from pointing out that the sample is not representative.

* Rudner trusts his sample to be representative and makes wild guesses to account for bogus findings in his biased data. For example, he says, "Compared to the national data a relatively small percentage of home school students are enrolled in high school. Possible reasons for this lower participation for high school students may be the relative newness of the home school movement, early graduation from high school, and possibly a desire on the part of some home school parents to enroll their children in a traditional high school." (Rudner, 1999, p. 6) (Actually, many people homeschool for high school. For example, in Wisconsin dating back to 1995-96, the total of homeschoolers in ninth, tenth, and eleventh grades has been greater than the total in first, second, and third.)

* Rudner's speculations about the data generated through this study demonstrate a woeful ignorance of homeschooling, including state laws. In attempting to account for why three states (Georgia, Ohio, and Virginia) were over represented in the sample, Rudner states in his report, "This is probably due to the fact that these states require testing of home school students." (p. 5) This explanation fails to take into account two important and obvious facts. First, there are many states with more demanding testing requirements that have much lower representation in the sample. Second, Ohio and Virginia do not require homeschoolers to take tests. Ohio homeschoolers can either take a test or submit a written assessment from a certified teacher or someone agreed upon by the parents and the local school superintendent. Virginia homeschoolers can submit an assessment, take a test, or claim a religious exemption from testing. Georgia only requires testing once every three years, and the test results do not need to be reported to the state.

* Rudner makes unwarranted generalizations in his report and in the online press conference on March 23, 1999. In the body of the report, he explains that 65.3% of the fourth-grade students in the study watched one hour or less of television per day, compared to 25.1% of fourth graders nationwide, according to information collected by the National Assessment of Educational Progress. But in the major findings at the end of the report (p. 25) and in the online press conference (Opening Remarks, p. 4), he says, "Home school students watch much less television than students nationwide; 65% of home school students watch one hour or less per day compared to 25% nationally" without reminding the reader of the report or telling the reader of the press conference that this sweeping statement is based on the reported television viewing of homeschooled fourth graders from a very unrepresentative sample.

In sum, it is embarrassing to have homeschooling associated with a study and a report that have so many problems.

Part III. How This Study Undermines Homeschooling Freedoms

Despite the fact that this study is so misleading, inaccurate, and based on such an unrepresentative sample of homeschoolers, it has the potential to undermine our homeschooling freedoms for a number of reasons, including the following.

* The study has been reported in the mainstream media as if it applied to all homeschoolers. The Christian Science Monitor for March 25, 1999, reported, "Of parents who teach at home, 88 percent have studied beyond high school (compared with 50 percent of parents nationwide). Indeed, 1 in 4 children taught at home have a parent who's a certified teacher." The Washington Post for March 24, 1999, reported, "The median income for home-school families was $52,000 a year, compared with $36,000 for all U.S. families with children." These published reports are not surprising since Rudner's "Summary of Major Findings" makes the same statements without a disclaimer about how unrepresentative the sample was and without a statement that the findings cannot be assumed to apply to all homeschoolers.

In addition, some researchers studying homeschooling are so eager for data that they will use even a study this poorly done. For example, although Welner and Welner are very critical of Rudner's report, they also write, "We applaud Rudner's contribution to building a greater understanding of the homeschooling movement." (Welner, Kariane Mari and Kevin G. Welner, "Contextualizing Homeschooling Data: A Response to Rudner," Education Policy Analysis Archives, Vol. 7, No. 13, April 11, 1999, p. 1.

We cannot assume this study will just go away. * The study gives the impression that homeschoolers are willing to take standardized tests, that, in fact, we go out of our way to have our children tested. This allows people who want to increase state regulation of homeschooling to say, "Obviously, homeschoolers are willing to take standardized tests used in many public schools. So why don't we require that homeschoolers take state-mandated tests?" This could lead to requirements for testing in states that do not currently require it and increased requirements for testing in states that now have testing. For example, states that now allow homeschoolers a choice between tests or portfolio assessment might begin requiring tests of all homeschoolers. States that now do not require that parents report their children's test scores to the state might begin requiring reporting.

When Farris was interviewed on Wisconsin Public Radio about the study, a strong response came from a school principal calling for increased regulation of homeschooling, including testing. (Unfortunately, Farris suggested that officials prosecute homeschoolers for "educational neglect." Homeschoolers in Wisconsin have worked hard to keep the term "educational neglect" out of Wisconsin statutes and were very frustrated to have it introduced by someone from outside the state, especially a homeschooler.)

As homeschoolers, we understand that there is a big difference between state-mandated standardized tests and parents choosing to have their children take tests, deciding which tests they will take and when and how, and determining with whom they will share the scores. However, proponents of state-mandated testing are likely to ignore or fail to understand this difference.

* This study increases the likelihood that homeschoolers will be included in national data bases and requests for information about us as homeschoolers. Rudner explains in the report, "Where possible, questions and responses were made to match those used by the U.S. Census, U.S. Department of Labor and the National Assessments of Educational Progress to facilitate comparisons of home school students with students nationwide." (p. 3) How can HSLDA have failed to realize that this opens the door for the federal government to collect much more information from homeschoolers by including us in requests for demographic data; collecting information on homeschooling as part of the census; suggesting that we provide testing data for fourth, eighth, and tenth grades just as public school students are required to do under Goals 2000 standards; etc. We have been working against being included in state and federal data bases to protect our privacy and our right to homeschool according to our principles and beliefs.

* The study implies that our homeschooling freedoms are based on test scores. This implication was made explicit in an online press conference on March 23, 1999, when Farris said, "We believe that our [homeschoolers'] good academic performance warrants greater freedoms even though every year the education establishment attempts to curtail our freedoms in a significant number of states." (Chat Session, p. 1)

This is a misleading and dangerous statement. Our homeschooling freedoms and responsibilities are based on our inalienable right as parents to choose for our children an education consistent with our principles and beliefs, which has been upheld by several U.S. Supreme Court decisions. This right comes from God or nature, not the state. It predates and is independent of our children's test scores. It means that we have the right to homeschool even if our children score below grade level on standardized tests designed to accompany conventional school curriculums. Our right to choose an education consistent with our principles and beliefs means we do not have to adopt the curriculum used in conventional schools so our children will score well on conventional school tests. To link our homeschooling freedoms with test scores undermines these basic freedoms and means that we have to earn from the state the right to homeschool.

* This study could change the standards that homeschoolers are expected to meet. People reading the conclusions of the study are likely to conclude that it's okay for families to homeschool as long as they meet the following criteria: There are two parents, preferably with the father working "at a job for pay" and the mother not working "at a job for pay". At least one parent has "some college education" and ideally is a certified teacher. The family has a relatively high income. The children score well above grade level.

The more the general public expects homeschoolers to have such qualifications, the more likely it is that states will increase their regulation of homeschooling to make sure that only properly qualified families homeschool.

* Many families choose homeschooling because they want an alternative to conventional schools. They may want their children to have a chance to learn without being bound to conventional curriculums and tests. They may want their children's education to emphasize religious, moral, or philosophical values. They may want their children to learn from the real world and not be limited to so-called educational materials.

This study pulls homeschooling into to the arena of conventional education. It says that what matters are the usual measurements of conventional education: scores on standardized tests, parents' educational background in conventional schools, family income, money spent on education, teacher certification. This study undermines the work being done by homeschoolers to provide our children and our society with a desperately needed alternative to conventional schools and government education.

Part IV. What We Can Do To Minimize the Damage Done By This Study

* We can refuse to use the study in any attempt to gain support for homeschooling, whether we are writing a homeschooling newsletter, dealing with legislators, talking with friends and relatives, facing a court case, or whatever. When would it ever be a good idea to cite a study that is fundamentally flawed and inaccurate? In addition, the study sends the message that homeschoolers wouldn't mind taking state-mandated tests. It says that homeschoolers should be measured by conventional school standards instead of being able to develop our own standards. If we cite the study, we give it power and importance.

* If we see this study being used, we can describe its problems and inaccuracies and voice our concerns about any statements made about homeschooling or homeschoolers based on this study.

* We can decide not to participate in such studies and encourage others not to participate in them. Such studies pretend to identify, label, and quantify who homeschoolers are, what we do, how we perform, and what we believe. Moreover, they attempt to do this using conventional measures. Participating in such studies lends credence to their importance and can lead to homeschoolers being further regulated based on conventional school standards, tests, and terms.

* We can ask any testing service we use not to collaborate with organizations, including homeschooling organizations, to use homeschooling test results or personal or other vital data about our homeschools to develop data bases for research purposes. We can also ask the testing service itself not to rent, sell, or use for research any information concerning our families, including our addresses, telephone numbers, ages or grades of our children, test scores, etc.

* We can tell BJU Testing Service that we are concerned about this study, the inaccurate information about homeschooling that it presents, and the fact that it could be used by critics of homeschooling to push for state-mandated testing of homeschoolers which would require that homeschoolers follow a curriculum geared to the state tests rather than a curriculum based on our principles and beliefs.

* We can tell HSLDA that this and similar studies feed right into the state's arguments for requiring testing of homeschoolers, including state-mandated tests. We can ask them not to make the data base of this study available to researchers, federal agencies, or others.

* We can keep the problems with this study in mind as we decide which homeschooling organizations to join and support as we continue the work we must to do maintain our homeschooling freedoms. This is especially important since this is not the first flawed research study that HSLDA has commissioned or the only step HSLDA has taken that offers a misleading and inaccurate view of the homeschooling community and then leaves the reader with the challenge of figuring out who really benefits.


The findings of this study cannot be used to make accurate statements about homeschoolers in general. Citing this study will support those who want to require that homeschoolers take state-mandated tests and will force homeschools to become more like conventional schools. Therefore, we should not use this study ourselves and should be prepared to correct those who cite it.

1999, Larry and Susan Kaseman


Online Press Conference on March 23, 1999. Transcripts of both the Opening Remarks and Chat Session are available at HSLDA's web site:

Survey and Lobbyists Cause Problems for Homeschoolers, Taking Charge column HEM, Sept./Oct., 1997

"Homeschooling Freedoms At Risk"

If you wish to share this information with others via your support group or local library, we request that you share it in its entirety, referencing source and credits.

....(articles list) | columns list)....

HEM General Information

Subscribe to HEM


Become a part of our
Resource Guide
Crafts for Learning

The New 3R's - Burns

Resource Guide

Become a part of our
Resource Guide

Children's Magazines
Homeschool Consulting

Online Programs
Parenting Support
Support Groups
Writing Programs

Become a part of our Resource Guide

Find Us on Face Book

Home Education Magazine, PO Box 1083, Tonasket, WA 98855; 800-236-3278
Contents © Home Education Magazine 1996 - 2012