Are IQ Tests Necessary Anymore, & What is a Good Source when Carrying out Research, After All?
What is it like for a person to advance to high school without the ability to read early-elementary level text? How different is it for a student with average intelligence, as opposed to someone with a significant cognitive disability, to face this reality?
Dr. Freeze (2020) describes the ability to read as existential. If a person cannot read, it impacts their sense of self. His description of the struggle to read, for kids like Andy, is insightful: “…The stakes are raised for non-readers when they enter school. More than any other variable, knowing how to read predicts academic and social competence, confidence, engagement and positive outcomes throughout the school years. If children do not learn to read at elementary school, they face an existential crisis. This crisis first appears when children fail to make the transition from “learning to read” in Grades 1, 2 and 3 to “reading to learn” in Grades 4, 5 and 6. During this transition, non-readers’ personal self-doubts, frustrations and reluctance to read evolve into blaming the teacher or the book, resistance to reading activities, pretend reading and learned helplessness. By high school, non-readers feel shame, inferiority and anger…It should be noted that some non-readers find success in work and life. However, I have never met a non-reading adult, successful or otherwise, who did not lament that fact that he or she had never learned to read” (Chapter 5, pp 2-3).
Costello, Foss, King, Mann, Schupack & Wilkins (2015) concur: “One’s language is very important to one’s identity. When children with dyslexia struggle with reading and writing in their native language, it damages their perception of themselves. A student’s frustration with the tasks of language may contribute to low self-esteem and the mistaken belief that he or she is simply unable to learn” (p. 1, Lesson 7).
Truthfully, had it not been for the fact that our school psychologist assessed Andy with a reading disability, and explained to me that, by definition, a person with a reading disability has average intelligence, I would likely have had an entirely different perception of Andy and my expectations for him would have been drastically different. I might have assumed that he had very low intelligence, due to his inability to read, and would have made inaccurate statements and recommendations to his teachers. I am grateful to our school psychologists for their ability to reveal to us so much about how the students we work with think and learn, and what is feasible for our students, in terms of academic and life goals.
Additionally, if I had known more about teaching reading to struggling readers back then, I would likely have been able to do more for Andy, in teaching him to read. Unfortunately, in my own growth as a teacher, I had not yet focused on reading research, and was learning instead about supporting teenagers with significant disabilities in transitioning to the work place, and to assisted living communities. Andy was an anomaly to me. The other students I worked with from day to day had very low intelligence, and were not able to read as a result. These two types of readers are addressed in the article that is the focus of today’s post, a research article entitled IQ is Not Strongly Related to Response to Reading Instruction: A Meta-Analytic Interpretation, (Stuebing, Barth, Molfese, Weiss & Fletcher, 2009).
Now, in the USA at least, it will become much more difficult for teachers to have students assessed for reading disabilities or dyslexia. Reasons for this change are explained below, but I have to say first of all, that I personally feel that not having this information puts teachers and students at a disadvantage in some ways. It allows teachers, like myself, to jump to conclusions about a student’s intelligence and abilities, and removes the ability to check in with a psychologist to learn the true nature of a student’s strengths and weaknesses. It allows for huge misperceptions on the part of everyone involved with the student. I have heard arguments in opposition to the use of IQ testing, which warn us against labelling students, but I find that a label would not to be as detrimental as the alternative. What if, for lack of an IQ test, we completely misunderstand a student’s cognitive ability, and set the bar much too low, or impossibly high? What are your thoughts on the issue?
Prior to the revisions to Individuals with Disabilities Act (IDEA) in 2004, schools used something called the “Discrepancy Model” to determine who received special education supports. Essentially, this model involved checking whether there was a difference, or discrepancy between a child’s IQ scores, and how well the student was doing in school, or whether they could meet grade level outcomes. A student who scored in the average range on an IQ test, but was struggling to read at grade level, could be diagnosed with a specific learning disability in the area of reading. Once given this diagnosis, students were provided with special education services. (Rosen, n.d.)
Now, in place of this model, many schools in the USA use the Response to Intervention (RTI) model. This is the model that is used commonly here in Manitoba, as well. In this model, there is no longer a significant need for IQ tests and specific diagnoses, when students are found to be struggling at reading. Students are provided with support when they are first seen to be experiencing challenges. The Response to Intervention (RTI) model “looks at all students’ reading, writing and math skills early in the school year. Then it provides targeted support to those who are struggling. Children who don’t respond to increasing support may then be considered for special education. The benefits of RTI: Students get help early. And they don’t have to wait to prove eligibility in order to get support” (Rosen, n.d. section 5, paras 1-2).
Teachers who work with struggling readers usually work with two different types of students, according to Joseph (2002), those with “IQ-reading achievement discrepancies,” (those described above, like my student Andy), and “students with a combination of low ability and low reading achievement” (para 5). “Low ability readers make up the largest number of poor readers. They tend to have lower than average IQ and have below grade level listening comprehension, word recognition, and reading comprehension performance” (para 6)
Many research studies have been conducted over the past decades that look at how effective IQ scores are in predicting student reading achievement, and how responsive students will be to reading intervention (Stuebing et al., 2009, pp. 31-32).
Steubing et al., (2009) looked at the results from a large number of studies of this kind, some which indicated that IQ was helpful is predicting student outcomes, and some which determined that there was very little difference in the learning outcomes of struggling readers with and without diagnoses of reading disabilities. The goal of the Steubing meta-analysis was to look at the multiple studies that had been carried out over time, to determine for certain, whether IQ scores could tell us how students would respond to reading interventions.
Sometimes when meta-analysts gather data, they do not include all of the data that is available to them, even if that data is valid. If a study shows a small effect, or if the results are considered “nonsignificant”, that study may not be sent in to be published, and the results may be ignored. When only studies that show significant effects are included in a meta-analysis, the researchers are not seeing the whole picture. When this happens, a positive bias can occur in the research (Steubing et al., 2009, p. 36).
The meta-analyses carried out by Steubing et al. was unique in that it included data that is sometimes left out, for the reasons described above. The methods they used involved estimation.
When reading this, I asked myself, “Since when is it more accurate to make an estimation? Aren’t estimates by definition less precise?”
Yet, the authors state that this method results in more accurate data, since a larger number of studies and therefore, more participants, are included and analyzed. It is an approach that was advocated for by researchers Lipsey and Wilson (2002, as cited in Stuebing et al.). It is a more difficult approach, that involves looking closely at studies that were unreasonably ignored in previous meta-analyses.
Apparently, in many meta-analyses, the researchers take the easy road, and look only at published studies that show a significant impact or difference. it is simpler to leave out studies in which the data is less easy to analyze or studies which are not as easily retrieved, since they were never published. This is something so common in fact, that it has been termed the “file drawer problem” by Rosenthal. (as cited in Steubing et al., 2009, p. 36). Again, a red flag rose in my mind, as I have been taught that the way to determine whether a source is reputable is whether or not it has been published by a respectful journal.
The researchers found, after looking at this much larger amount of data, including studies previously left out, that there were only very minor differences among the two groups of struggling readers. Thus, there was very little basis for requiring IQ tests (p. 44). If the IQ test tells us that a learning disability exists or does not exist, but fails to impact how we teach reading to kids, I can see why people may question the need for IQ tests.
Essentially it was concluded that there is no difference in how these two groups of students learn to read: IQ scores do not have a role in planning interventions, or in matching interventions to readers, since there is no important difference in how we should approach teaching reading, to the two types of students. There is no difference in “growth patterns” and “no significant differences between these two groups of readers on how they develop reading precursor skills” (Wristers, Francis, Foorman, Fletcher, & Swank, as cited in Joseph, 2002).
What do we use then, if we do not use IQ scores, when planning reading interventions? Instead, teachers can use tests of phonological awareness or reading comprehension, and be guided by the data they collect on their own. An excellent article on how to support struggling readers with and without reading disabilities can be found in the following link:
I suppose that leads me to conclude that requesting an IQ test to determine which reading intervention to use with a particular student, may not be necessary. I do, however, see the value in IQ tests when it comes to long range planning for students who struggle to read and learn. These tests help teachers and parents to understand how an individual child’s brain works, what methods might be especially effective in supporting a specific learner, and what to expect in terms of progress over time. I would like to invite you to share your own thoughts on IQ tests, below.
I would also like to invite you, colleagues of mine, who may have more experience with research and meta-analyses, whether the methods used in this study are in fact proper, or whether the results should be questioned, due to the use of estimation and non-published studies. I am tempted to believe that the methods used by Steubing et al. are suitable and appropriate, considering that their results were published in a well-respected journal, even if the data from some of the studies they consulted was not. Also, it makes sense to me that leaving out data can skew results, especially if that data is properly acquired. What do you think? Please share your thoughts and knowledge on the topic below!
References
Costello, S., Foss, J. M., King, D. H., Mann, M., Schupack, H. & Wilkins, A. (2015). Academy of Orton-Gillingham Practitioners and Educators. AOGPE Subscriber Course – Lesson 7: History of the English Language, Part 1. Retrieved February 13, 2020 from http://courses.ortonacademy.org/
Freeze, R. (2020). Precision Reading: Instructors’ Handbook (3rd Edition). Winnipeg, MB: D. R. Freeze Educational Publications (www.precisionreading.com).
Joseph, L. (2002). Best practices in planning interventions for students with reading problems. Reading Rockets. Retrieved February 22, 2020 from https://www.readingrockets.org/article/best-practices-planning-interventions-students-reading-problems
Rosen, P. (n.d.). The discrepancy model: What you need to know. Retrieved February 22, 2020 from https://www.understood.org/en/school-learning/evaluations/evaluation-basics/the-discrepancy-model-what-you-need-to-know
Stuebing, K. K., Barth, A. E., Molfese, P. J., Weiss, B., & Fletcher, J. M. (2009). IQ is not strongly related to response to reading instruction: a meta-analytic interpretation. Council for Exceptional Children, 76 (1), 31-51.
The Iris Center. Star Legacy Modules. Retrieved February 22, 2020 from http://www.ideapartnership.org/documents/IQ-RTI.pdf
Recent Comments