Podcast draft

Reading level systems, such as Accelerated Reader's ATOS, Lexile, and Fountas & Pinnell, attempt to determine the complexity of a text and assign it a reading level. However, these systems vary in their factors and expressions of level. For example, ATOS considers word length, sentence length, and vocabulary, while Lexile focuses on sentence length and word frequency. Fountas & Pinnell analyze multiple factors, including content and themes. These systems can lead to inconsistencies and inequities in education. The validity of these systems is also debated, with some texts being assigned unexpected levels. As a book-obsessed child, I remember feeling proud of my AR level, checking out thick books labeled as 10.3 from my elementary school library. I racked up AR points competitively, collecting pizza parties, water park trips, and once, a $50 Marshalls gift card from my 6th grade English teacher. Maybe I was too proud. Reading level systems faded from my concern after middle school, but now, as a prospective librarian working in a children's department at Chicago Public Library, they are back with a vengeance. Diligent caretakers ask for books at a certain AR, Lexile, or Fountas & Pinnell level for their children, and I attempt to balance their needs while explaining that the public library does not label books this way, and that perhaps, the library should. All the different branded reading level systems attempt to quantify and streamline a student's reading progression from early elementary through high school. All of them are designed by for-profit companies. All of them produce widely different results for the same text. My name is Lil, and today, we're going to talk about how reading levels report to work, why they don't really work, and how these classifications migrate from text on to students. There are several popular reading level systems. I'm going to focus on Accelerated Reader's ATOS levels, formerly AR levels, the Lexile framework for reading, and Fountas & Pinnell's levels. I chose these due to my familiarity with them. My school used Accelerated Reader's tools throughout my education, and now, in my work at the public library, I most frequently encounter Lexile and Fountas & Pinnell's systems. All three systems identify factors that they take into consideration in order to break down the complexity of a text and assign it a reading level, but the factors and expressions of level vary. Let's break down how each one works, starting with Accelerated Reader's ATOS. ATOS, known in my time simply as AR, takes into consideration average word length, average word grade level, average sentence length, and book length, according to the Renaissance website. It draws from a graded vocabulary list, which includes more than 100,000 words, based on data gathered over 2.5 billion words and more than 170,000 books, according to Dr. Andrew B. Miller in 2014, Writing for Renaissance. The numerical output corresponds to grade level, so books range from 0.0 to 13.0, increasing intervals to 0.1 for each month of the school year, meant to encompass children's reading from kindergarten through high school. For example, 5.0 to 5.9 are intended to map on to the progression of a fifth grader's reading from the beginning of the year to the ninth month of classes. The 2024 Newberry Honor winner, Simon Sort of Says, written by Aaron Bao, is assessed at 5.2, meaning ATOS expects the text to be readable for a student in their second month of fifth grade. Accelerated Reader explains that the ATOS level is meant to be used in conjunction with interest levels, which relate to content and age-appropriateness based on publisher recommendations. And furthermore, a student's reading level is intended to be taken into consideration, as ascertained through standardized tests. Accelerated Reader does offer its own test. Interestingly, Accelerated Reader dedicates space on its webpage to explaining the Lexile framework for reading, through a partnership with MetraMetrics, Inc., which owns Lexile. On the current website, Accelerated Reader positions Lexile as a complementary or equal tool to consider. But in a document published by Accelerated Reader in 2007, Renaissance Learning argued that ATOS is more accurate and better than Lexile. Now, 17 years is a long time, but the complete 180 Accelerated Reader made on the validity of Lexile gives me pause. Anyway, I'll turn to Lexile's own website to define its framework, rather than letting its competitor partner do so. In 2020, their director of educator engagement, Matt Copeland, gave a talk on Zoom titled An Introduction to the Lexile Framework for Reading. He explains that the Lexile framework encompasses both the complexity of text and the ability of individual readers, although, as he asserts, there is no such thing as a Lexile test for readers. Students' reading levels are read through a variety of other standardized tests that the education system requires. Copeland compares a Lexile scale for books to a thermometer, and, just like a thermometer, Lexile can level a reader and text below zero. He justifies it by clarifying that they don't actually use negative numbers or the negative sign. No one wants to stigmatize a child or a book in such a manner. Instead of negative 100, a text or child will be designated BR 100, which, of course, in effect, accomplishes the same undesirable classification. More on these kinds of negative identifications later. Lexile bases its determinations on sentence length and word frequency, and the scale ranges from BR 300L to 2000L. Obviously, this is messier to map onto grade levels and add to us the simple numbers, with added letters BR for Beginning Reader and L for Lexile, which feels redundant. Now one cannot see a Lexile score and intuit which grade level it might correspond to, instead needing to rely on another interpretive tool. According to a chart from Lexile developed to demonstrate the connections between reading levels and grades, ranging from 25th to 75th percentile students, Lexile locates BR 460L to 150L for Kindergarten, 650L to 1260L for 5th grade, and 795L to 1410L for 7th grade. The gradation also allows for overlap between grades. Lexile says the students should strive for the reading comprehension sweet spot of 100L below to 50L above the reported Lexile measure when looking for books. To give an example, another 2024 Newberry Honor winner, Eagledrums, by Nassib Grok-Rainey-Hobson, weighs in at 910L, or near the 50th percentile of 5th grade students' Lexile levels. And then we reach Fountas and Pinnell's reading levels. Instead of a numerical scale, they employ an alphabetic one, from A to Z+. In order to access the document which explains how Fountas and Pinnell divides text, I had to become a member. I forked over my email address, employment information, and location, just to learn how the system works. They use 10 factors to determine text difficulty—genre, text structure, content, themes and ideas, language and literary features, sentence complexity, vocabulary, words, illustrations, and book and print features. The single-page document listing these factors defines them in the abstract, but does not indicate how Fountas and Pinnell determines these qualities of text, nor how some of them might impact the Fountas and Pinnell level assigned. Encompassing way more metrics than Lexile or Atos, Fountas and Pinnell's system explicitly analyzes the content and themes of text, inserting subjective, ideological decisions about age appropriateness and morality into their reading levels. Fountas and Pinnell's alphabet extends from kindergarten through eighth grade, with designations of high school and above lumped together in Dizzee+. Kindergarten levels are expected to be between A and D, fifth graders between T and V, and both seventh and eighth grades find themselves lumped together on Dizzee. The winner of the 2024 Newbery Medal, The Eyes in the Impossible, by Dave Eggers, is granted a Y by Fountas and Pinnell, ideally positioning it for sixth-grade readers. In addition to having to bless Fountas and Pinnell with my free data, just to learn about the actual process behind their system, I could not access the official database which connects Fountas and Pinnell levels to text. Both Accelerated Reader and Lexile's databases were easy to find and use, which they should be in order for caregivers to accurately assess their child's readings within the confines of the system. Fountas and Pinnell, however, lock away this information behind a $25 annual subscription and make clear that the Fountas and Pinnell leveled books website is the only official source for the Fountas and Pinnell level of books. Briefly, to put them all together with an example, Access places fifth-grade-level texts between 5.0 and 5.9, Lexile locates a fifth-grade text and reader between 795L and 1100L, and Fountas and Pinnell place fifth-grade-level texts into the categories T, U, and V. It's a lot to keep track of, and students evaluated across different systems are held to different standards, which direct them to different materials, creating inequities between educational environments. The discontinuities between reading-level systems are not under-documented. In recent years, Ricky Ginsberg and Kelly Moye have written for the Washington Post on these issues in 2019, and Wayne DiOrio weighed in for the School Library Journal in 2020. While I would love to draw up my own list of inconsistent examples, I am not paying Fountas and Pinnell $25 to access their database. So let's review Ginsberg, Moye, and DiOrio's findings. Ginsberg and DiOrio first looked at some admittedly dramatic examples of the Lexile framework's departure from any kind of common-sense classification. The Sun Also Rises, by Ernest Hemingway, gets to 610L, placing it at the third-grade level, while the first installment of Dario Bawimpi Kid, from Jeff Kinney, falls at 950L, between a fifth-grade and ninth-grade level. A quantitative analysis of the words used in a text and the length of sentences cannot account for all of the literary devices authors harness to complicate their text. DiOrio chooses to compare Atos, Lexile, and Fountas and Pinnell levels for the same text to one another. On one hand, Harry Potter and the Sorcerer's Stone is assigned an 880 from Lexile, a 5.5 from Atos, and a V from Fountas and Pinnell. In other words, all the systems agree that the text corresponds to a fifth-grade level, with Lexile's range broadening from fourth to sixth grade, depending on the child. However, they don't always cohere. Twilight is given a 720L by Lexile, a 4.9 by Atos, and a Z-plus by Fountas and Pinnell. While Lexile and Atos place its readability in late elementary, Fountas and Pinnell declare it appropriate for high school students, presumably due to Fountas and Pinnell's built-in content filters.