In 2022, I self-published two books: The Scientific Principles of Reading Instruction and the Scientific Principles of Teaching. Writing a book was always a life goal of mine and I am thrilled that I have now put out multiple books. The goal of these books was to publish summaries of scientific research on each topic, based on the available meta-analyses. However, within the context of science writing, I must say writing a book has two major drawbacks:
1. New studies are constantly coming out and changing the landscape of scientific research, which means science books can become out of date quite quickly.
2. Writing and selling a book, incentivizes you to be unwilling to change your mind, as new evidence comes out. Indeed, authors often feel personally invested in the opinions they write.
I wrote about these problems in each book and stated that I intended to write updated editions as time went on to update the content, as both new research was published and my view points changed. That said updating a book, is a timely process and I recognize that people may not want to purchase multiple editions of these books.
I do not want anyone who has purchased one of my books to feel required, to purchase new editions as updates come out. To rectify this, I have decided to create a page on this blog dedicated to publishing corrections. Below you should find, a list of interpretations that I made within my books that have either been disproven or which I have changed my mind on. This list will likely grow over time, as research continues to evolve. I also hope that this list helps to keep me from becoming to entrenched in my own interpretations of the research.
1. Morphology & dyslexia: When I wrote The Scientific Principles of Reading Instruction, I hypothesized that morphology might be especially important for students with dyslexia. This hypothesis was based on the fact that previous meta-analyses have shown larger effect sizes than meta-analyses on phonics for dyslexic students. However, one week after I published my book, Colby Hall, et al. published the most comprehensive and well controlled meta-analysis ever conducted on the topic of interventions for dyslexic students. Their meta-analysis showed no significant difference between interventions that did or did not use morphology. For more information about this topic, click on this link: https://www.teachingbyscience.com/dyslexia-debat
2. Growth Mindset: In both of my books, I wrote about why I thought it was crucial to teach students to have a growth-mindset. Since then, a meta-analysis was published by Macnamara, et al. which gives me serious reason to doubt this hypothesis. For more information about this topic click on this link:
3. I wrote that longer studies should show larger effect sizes, as students in these studies have more exposure to the learning stimulus. Well this sounds logical, I am fairly convinced now that this hypothesis was completely wrong. In my meta-analysis of language programs, I found strong negative correlations for language studies, regardless of the type of instruction, as can be seen in the following link: https://www.teachingbyscience.com/a-meta-analysis-of-language-programs
4. I interpreted the NRP phonemic awareness results to suggest that phonemic awareness instruction showed diminishing returns after 20 hours, because studies longer than this showed lower effect sizes. However, this analysis ignores the fact that longer studies in general show lower effect sizes. This does not mean that phonemic awareness should be taught across all grades, as there are few studies on phonemic awareness instruction past grade 1. However, it does mean that there is no need to limit phonemic awareness instruction to less than 20 hours in early instruction.
5. When discussing the NRP phonics results, I erroneously used the NRP effect size of .44 for systematic phonics instruction, instead of the overall phonics mean of .41. This error is only in early printings of my book it has been fixed in more recent printings.
6. I referenced two meta-analyses on balanced literacy, from John Hattie's Whole Language meta-analysis data-base. However, while these meta-analyses were on balanced literacy, they did not refer to the common interpretation of the term. To date, I believe my meta-analysis on balanced literacy is the only meta-analysis of this topic. This is problematic, as mine is not peer-reviewed and it is an important topic. However, I have submitted a meta-analysis of this topic for peer-review and I hope that it is published soon. I would also like to note that John Hattie, has also seemed to notice this error, as these studies are no longer listed in his data-base. This error, has been corrected in the most recent printings of my book and only effects the earliest printed copies.
Written by Nathaniel Hansford
Last Edited: 2023-04-17