A review by Brian D. Ray, Ph.D.
Consider donating/supporting NHERI today
It is an important and emotional topic and the title is catchy but this paper in a scholarly journal adds close to nothing to the relevant body of scientific knowledge. Further, the authors commit many errors regarding logic, omission, and a review of literature. The paper is entitled “Homeschooling and child maltreatment: A review of the regulatory context and research evidence in the United States” and is written by Anneliese Luck and five others (2025).
Background and Methodology
The writers’ purpose was to “… understand better the context of child maltreatment in homeschooling environments and the processes, policies, and practices that have potential to mediate any heightened risk of abuse or neglect of children, with the goal of advancing reasonable policy conversations to promote child safety” (p. 2). To do this, they went on a wide-ranging and problematic literature review covering homeschooling definitions, data, and demographics, articles on child maltreatment, policy debate, state-level homeschooling laws and regulations, and various authors’ and organizations’ claims about what the civil government (state) should do to control homeschooling parents and children more.
Findings That Are Supported by the Authors’ Wandering Overview of Literature
The following conclusion by Luke et al. is generally true and follows from the literature reviewed, depending on how one interprets it: “However, the lack of reliable data and highly politicized nature of the topic have produced a body of mixed and inconclusive research, making it difficult to ascertain the true prevalence of abuse and neglect among homeschooled children” (p. 10). It is true that there is not enough empirical research to date to conclusively determine the rates of abuse, neglect, and child fatalities among public school and private school students compared to homeschool students. At the same time, any implication by Luck et al. in their surrounding sentences in the article that more reliable data are needed from the state (civil government) to determine prevalence is philosophically driven.
The authors are also correct that there is a “… lack of concrete evidence …” (p. 10) that public school or private students are at any less risk of child maltreatment than the homeschooled or that more state control over homeschooling is necessary to make home-educated children safer than they are.
False, Misleading, or Highly Dubious Claims and Problematic Strokes by the Authors
Only limited examples of serious problems in this paper will be given here. There are several more for which time and space will not be spent.
Luck et al. wrote that “… a recent analysis of CRHE [Coalition for Responsible Home Education] data, using national child fatality rates between 2000 and 2019 as a comparison … found that the observed fatality rate of homeschooled children is significantly higher than the expected rate (CRHE, 2024).[1] . . . . . The organization [CRHE, 2024] published preliminary findings that fatality data, though incomplete, suggested a significantly heightened risk of dying from child abuse among homeschooled children” (p. 2, 5). However, CRHE (2024) reported the following: “Homeschooling itself is not known to be a risk factor for systematic abuse, and there is no evidence that cases of systematic abuse discussed in this section are representative of abusive homeschooling environments on the whole. Researchers of child torture estimate that it is rare.” Further, CRHE (2024) does not report any finding that “… the observed fatality rate of homeschooled children is significantly higher than the expected rate” compared to anything in general (e.g., non-homeschool children). CRHE, in fact, reports on “… withdrawal [from school] and non-withdrawal cases …” (p. 12) which is a very particular and small-minority set of cases and has nothing to do with homeschool children in general compared to public school or private school children. Luck et al. do not provide any reference for their cited recent analysis of CRHE data. In particular, they do not tell readers the statistical significance of the analysis that “suggested a significantly heightened risk of dying from child abuse among homeschooled children.” Luck et al.’s reporting here is inaccurate and misleading.
This author also wonders why Luck et al. did not reference CRHE (2015), especially when they heavily relied on CRHE’s articles. CRHE (2015) found no statistically significant difference in child fatality rates between homeschool children and other children.
Luck et al. referred to (Connecticut Office of the Child Advocate, 2018) four times and called this report by a state government a study. It is not a scholarly study in any sense and the findings are not representative in any way. In this report, the government mixed theoretical argument with some empirical evidence after they reviewed data, not collected to result in a representative sample, from six “… school districts’ withdrawal/homeschooling practices …” They described case examples of alleged child abuse or neglect from six school districts and presented no finding about the overall rate of abuse or neglect of students who were withdrawn from a n institutional school to be home educated” (Ray, 2024).
Knox et al. (2014) is cited three times by Luck et al. who correctly noted the following: “These cases were not randomly sampled but instead selected by pediatricians …” Unfortunately, Luck et al. claimed that “The study found a link between severe abuse and home education …” which implies that there is a statistical correlation, if not causal relationship, between homeschooling and “severe abuse.” This is an invalid inference from the study they cited.
For simpler and more direct scholarly comments on the Connecticut government’s report (Connecticut Office of the Child Advocate, 2018) and Knox et al. (2014), read Williams (2021) and Ray (2024). The reader will see that Luck et al. shed no new or clear light on the immediately aforementioned state report and study.
It appears by Luck et al.’s approach and the journal in which they published, theirs is supposed to be a scholarly article. However, the authors reference the same newspaper article five times as support to their thesis and conclusions. Luck et al. also reference approximately 23 times an advocacy organization whose main purpose is to promote more state (civil government) authority and control, and less parental authority, over homeschooling. Therefore, they considerably over-weighted one newspaper article that assertively associated homeschooling with child maltreatment by reporting the evil behavior of parents in one homeschool family. And Luck et al. also presented as serving as a mouthpiece for that one advocacy organization that works hard at publicity and lobbying policymakers to control homeschooling more, but is not a research think tank.
One of the most serious errors in Luck et al. is their near omission of two of the most significant studies ever done in the field of homeschooling pertaining to child maltreatment, abuse, or neglect (i.e., Dills, 2022; Ray & Shakeel, 2022) and both in peer-reviewed journals.[2] These are the two most methodologically sound and statistically advanced studies to date and they offer concrete empirical evidence that speak directly to the issue of Luck et al.’s paper. They cited Dills once, within parentheses with seven other references. Luck et al. then never referenced Dills again and never explained Dills’ study or its direct relevance to the issue of child maltreatment and state control of private homeschooling education. This is an inexcusable error on the part of Luck et al.
Luck et al. cited Ray and Shakeel (2023) within parentheses with seven other references. Later, they gave Ray and Shakeel one sentence that was dismissive and disdainful regarding their conceptually and statistically significant findings. At the same time, Luck et al. afforded other studies and reports, with notably more methodological limitations, several sentences or repeated mentions. This is also an inexcusable error on the part of Luck et al.
There are some more subtle problems that appear in Luck et al. For example, it is peculiar and unfitting that they relied heavily on CRHE for an analysis of state laws and regulations when the Home School Defense Association (HSLDA) is a far older organization and recognized across the United States and internationally for being the most reliable source of ongoing and regularly updated information on these topics. This raises the question: Did Luck et al. prefer CRHE’s information and presentation thereof over HSLDA’s because the former better fit the authors’ political and policy philosophy?
Luck et al. make the following patently false claim: “Most research on homeschooling has been funded by advocacy organizations (e.g., Home School Legal Defense Association)” (p. 3). First, they offer no data to support this claim and the burden of proof, in terms of scholarship, is on them. Second, there have been hundreds of research studies on homeschooling across the past 40 years, in the United States and internationally. Homeschool advocacy organizations are normally nonprofits with very small budgets, and typically have no budget for funding research. This author does not have the time to gather all the data on the number of studies and whether they were funded by “advocacy organizations” but this author has been extremely active in the homeschool research field for over 40 years and knows that this claim by Luck et al. is false.
Further, their assertion about funding appears to be a committal of the fallacy of poisoning the well, the rejection of an author’s or researcher’s findings or claims solely because of his or her circumstances or position in life (e.g., funding source). Such a claim and implied poisoning of the well would be a slight, if not an insult, to researchers and scholars across the past over 40 years such as Reed Amussen Benson, Patricia Lines, Stephen Arons, Jane Van Galen, Mary Anne Pitman, J. Gary Knowles, Joseph Murphy, Oz Guterman, Ari Neuman, Cheryl Fields-Smith, Robert Kunzman, Mitchell Stevens, Angela Dills, Angela Watson, Albert Cheng, Daniel Hamlin, Steven Duvall, Richard Medlin, Carlos Valiente, David Sikkink, and Tracy Spinrad. This list could be greatly expanded.
On the note of funding, Luck et al. list four reports/studies by Ray and claim that they were funded by HSLDA. This is a false claim; none of those four reports/studies were funded by HSLDA. The author of this present article is that same Ray, and he knows. In addition, HSLDA has funded a very small minority of all research done on homeschooling in the United States, or anywhere around the world, by anyone.
Further, to be balanced scholars, Luck et al. should report the funding sources of all the scholars and organizations that they cite if they try to identify funding for any of them. One must wonder why Luck et al. do not mention that the funding of negative critics of homeschooling. Many of them, directly or indirectly, are funded by the state-run tax-funded public school system (e.g., university schools of education, state licensure of teachers), other government systems or agencies that are tax-funded, and individuals who are philosophically or emotionally opposed to homeschooling that is free of state controls.
Concluding Comments
Luck et al. gave their paper a catchy title on an important topic and there was potential for them to add something new and valuable to the scholarly knowledge base. However, they fell short, in terms of both what they included and failed to include in their analysis.
One of the most troubling aspects of their paper is that they promoted an implied, if not explicit, theme that if only the civil government would control private homeschooling education more, then the state would be able to gather more data on the children and parents to discover if there might be a statistically significant problem that the state might need to address to maybe prevent parents (and maybe other adults) from doing bad things to children. Consider the following quote, with emphasis added in boldface:
“Although evidence of elevated risk of child maltreatment for homeschooled children is inconclusive, the link among parental stress, social support, and child maltreatment is well established [in society in general] … For some parents, even when elective, homeschooling can be both stressful and isolating, which may pose risks for their children. In particular, children with behavioral issues and disabilities, a population more likely to be homeschooled, may be at heightened risk [in society in general] … Research has shown that parents of such children [in the general population] often have higher levels of stress and disabled children are 3 times more likely to be victims of abuse and neglect [in the general population] … In the absence of traditional child protection mechanisms [i.e., the presence of mandatory reporters in institutional schools], vulnerable homeschooled children whose parents lack proper support or coping mechanisms may be at risk.” (p. 4)
Despite the fact that Luck et al. found there is no empirical evidence that institutional public schooling and private schooling pose any less child maltreatment risk to children than does homeschooling, they pressed their thesis that there is likely a problem with homeschooling that the civil government and researchers need to find and then the state would to and could solve with more laws (or policy) that control homeschooling.
Fortunately, Luck et al. did accurately report that there is no conclusive research that public school and private school students are maltreated, abused, or neglected at any lower rate than homeschooled children.
References
Coalition for Responsible Home Education (CRHE)/Homeschooling’s Invisible Children. (2015). Some preliminary data on homeschool child fatalities. Retrieved April 16, 2025 from https://web.archive.org/web/20150602050333/https:/hsinvisiblechildren.org/commentary/some-preliminary-data-on-homeschool-child-fatalities/
Dills, Angela. (2022). Homeschooling and child safety: Are kids safer at home? Journal of School Choice, https://doi.org/10.1080/15582159.2022.2113355
Ray, Brian D. (2024). Homeschool abuse and neglect research: How many homeschooled kids are abused?, https://nheri.org/homeschool-abuse-and-neglect-research/
Ray, Brian D.; & Shakeel, M. Danish. (2022). Demographics are predictive of child abuse and neglect but homeschool versus conventional school is a nonissue: Evidence from a nationally representative survey. Journal of School Choice, 17(2), https://doi.org/10.1080/15582159.2022.2108879
Williams, Rodger. (2021). Two invalid “studies” promoted by homeschool critics. https://homeschoolingbackgrounder.com/two-invalid-studies-promoted-by-homeschool-critics/
Luck, Anneliese; Chen, Wan-Ting; Cuccaro-Alamin, Stephanie; & Prindle, John; Berrick, Jill Duerr; & Putnam-Hornstein, Emily. (2025). Homeschooling and child maltreatment: A review of the regulatory context and research evidence in the United States. Child Abuse & Neglect 163 (2025) 107355, https://www.sciencedirect.com/science/article/abs/pii/S0145213425001103
[1] Most of the references that Luck et al. cited will not be entered into a reference list in this article. There is no time or space for that redundancy.
[2] This author recognizes that not all papers in peer-reviewed journals are of high scholarly quality and that some papers not in peer-reviewed journals are of high scholarly quality. At the same time, the scholarly, academic, legislative, and judicial worlds continue to give generally higher credibility to peer-reviewed work than to non-peer-reviewed work.