Adding Friction. A Librarian Asks, “What about Using a Database like Newsela or ReadWorks for Research?”

As schools implement more problem- and project-based learning and participate in science fairs and history competitions, students are encountering more esoteric and specialized sources that are hard for them to understand. Can a database of leveled articles serve librarians helping students do research? Let’s compare two of these databases: ReadWorks and Newsela.

Leveling schemes emerged from language arts teachers’ need to differentiate reading selections in their classrooms. ReadWorks and Newsela rewrite the same content for different levels. The free version of Newsela, a for-profit company with well-known investors, uses a freemium model in which students can read articles, take quizzes, and annotate sources. But, most of the teaching benefits (e.g., constructing writing assessments, viewing and replying to student and teacher annotations, monitoring individual and school progress and tracking it against Common Core standards) are available only in Newsela PRO (Davis). In contrast, the entire ReadWorks platform is free and directed by a highly-rated nonprofit ("ReadWorks: Score") with an advisory board of literacy academics.

Scope

One of Newsela’s strengths is its curation of current “high-interest news and nonfiction articles,” including some primary sources. ReadWorks hires “professional writers and journalists” (Arabo et al. 16) to compose half their articles, while the rest comes from 13 content providers. Two partners, History.com and the Gilder Lehrman Institute of American History, supply content to both platforms.

ReadWorks describes the process of creating leveled rewrites as “lovingly handwritten” and assures us that “all of the important knowledge of the original article, as well as the key academic vocabulary, rich syntax, word count, and beauty of writing” are preserved("StepReads"). Newsela ninth level (“Max”) is a lightly edited version of the published article which, in turn, is rewritten into four (out of potentially eight) levels “by hand…with attention to the original author's voice, background knowledge of the reader, and many other aspects of text complexity” (Palenski). While we are assured that “no Newsela text is adapted by computers or via machine text simplification,” the company’s “About” page claims it uses natural language processing software to do “proprietary, rapid text-leveling.” Although Newsela has sent me numerous emails since I began a trial, a number of attempts to contact Newsela to clarify this apparent inconsistency have failed.

Searching

To get the most benefit from tailored content in both products, use the advanced level options to limit by discipline, content partner, and comprehension subtopics. In ReadWorks, you can select from 18 skills and strategies including author’s purpose, genre, compare and contrast, main idea, inferencing, and vocabulary in context. Searches can locate added reading supports such as paired texts or audio read by a human vs. synthesized voice.

Newsela’s advanced search limits by content provider, format (essay, primary source, reference), and by reading skills, assessed in pairs by four multiple-choice questions per article. One can identify texts with instructional supports like a Spanish translation or embedded definitions. PRO users can locate and make use of activity packages with texts, as well as sources that contain pre-created annotations to reuse or modify.

Writing in Levels

In comparing revision decisions on both platforms, one begins to appreciate the writing challenges and tradeoffs. For example, Newsela adds subheadings that chunk text, presumably so that students can readily understand the content that follows. ReadWorks’ content is not subdivided. As a result, readers are being challenged to make inferences in order to understand the main idea.

In both programs, a lower-level rewrite may actually be longer than the original. One Newsela article logged an additional 83 words in order to clarify terms and ideas presented in the original that might be unfamiliar to 9th graders. Likewise, ReadWorks added 44 words to a StepRead article targeted to 8th graders.

Vocabulary choice and sentence structure and length are key factors in reading comprehension. For example, an original article from the Washington Post about the National Park Service’s decision to rescind the bottled water ban reads:

“According to the report, drafted largely during the Obama administration, eliminating sales of water bottles prevented up to 112,000 pounds of plastic from being sold and discarded each year, along with up to 140 metric tons of carbon dioxide emissions. In addition, between 276 cubic yards and 419 cubic yards of landfill space stood to be saved.” (Fears)

In Newsela’s simplest version, all this data is replaced by a single, easily understood, numerical estimate:

“The bottle ban may have prevented more than 40 dump trucks of trash.” (“National Park” 590L)

However, this visualization of plastic water bottles in terms of numbers of dump trucks (described as a range in the passage below) is not present in the original Washington Post article. Another version of the article reads as follows:

Ban Was Good for the Environment
“Ending sales of water bottles stopped up to 112,000 pounds of plastic from being sold and trashed each year. It also stopped the release of up to 140 metric tons of carbon dioxide. In addition, between 276 cubic yards and 419 cubic yards of landfill space stood to be saved. That's somewhere between 28 and 42 dump trucks full of trash. ("National Park” 900L)

The subtitle telegraphs to the reader what should be concluded from the statistics, which are couched in shorter sentences. Other modifications include word substitutions (stopped vs. prevented; trashed vs. discarded) and the omission, unfortunately, of the only scientific term: emissions. Theoretically, the text might be rearranged for clarity but, if that were the case, the dump truck concept should come after the first sentence to which it relates.

Like Newsela, ReadWorks shortens paragraphs to chunk ideas and revises sentence structure and vocabulary for readability. However, ReadWorks’ aims to retain some complexity so that readers must still face challenging text and use reading strategies. For example, in the versions of an 8th grade science article on wetland destruction, an eight-sentence paragraph ends with these three sentences:

“Although the state has only 40 percent of the country’s wetlands, it also bears 80 percent of the country’s wetland losses. Every 38 minutes, the equivalent of a football field is lost. This has serious implications for the region’s wildlife and economy, as well as the ability of the region to withstand natural disasters.” (Lexile 1210L)

The rewrite consists of eleven rephrased sentences split into two paragraphs with the addition of 25 simpler words. The conclusion still consists of three sentences, the first of which ends rather awkwardly:

“Louisiana may have 40 percent of the country’s wetlands, but it’s also where 80 percent of the country’s wetland loss is happening. Every 38 minutes, a section of wetlands in Louisiana as big as a football field is lost. This loss has serious consequences for the state’s wildlife, as well as the people and businesses there.” (Lexile 1010L)

In both platforms the rewriting is serviceable and, occasionally, either inspired or insipid.

Standards and Assessment

Both Newsela and ReadWorks append Common Core anchor standards to their articles and curated sets (paired texts, text sets). Newsela’s science text sets, built from Scientific American articles, are pegged to the disciplinary core ideas of the Next Generation Science Standards.

In the free version of Newsela, article-level quizzes assess these standards, but only the PRO version allows the teacher to track individual and class progress in attaining those standards and to view, sort, and filter the data. Also in PRO, teachers can see Newsela’s suggested answers to quiz questions as they are grading students’ responses. Other assessment features include the ability of both teachers and students to highlight a section of the text and ask their own questions. The dialogue that emerges, visible to both, can form a basis for formative-, summative- and self-assessment.

In the free ReadWorks platform, the instructor can aggregate all student work by class or individual student for assessment. Students cannot highlight and annotate text, but they can take notes in a personal Book of Knowledge. ReadWorks has developed questions for paired and text sets, in addition to article questions, so one can assess a student’s ability to integrate information from multiple articles. ReadWorks has a more nuanced approach to assessment; teachers are able to grade written responses within a range rather than as just correct or not.

Leveling Books, Not Students

ReadWorks advises teachers to encourage students to tackle the original, either before or after StepReads. In Newsela, the teacher initially assigns a level to the entire class but each student can switch to another level or even to the Spanish translation, when available. Transparent leveling and easy switching among levels and styles can build student’s sense of agency and independence (Belle).

One of the most intriguing aspects of Newsela is that it customizes the reading level presented to each student based on their performance within the software (Davis). One wonders if, in the minds of both the teacher and students, this conflates text level with student labels. It’s important to remember that leveling measures were created “to help teachers think more analytically about the characteristics of texts and their demands on the reading process” (Parrott "Fountas and Pinnell Say”), not to level or label students.

Library Research Potential

Since the purpose of both products is to scaffold the teaching of reading, they’ll be welcomed by language arts teachers or educators with learning-differenced students or those whose native language isn’t English. Reflecting its up-to-date news focus, articles available only in Newsela include information on transgender, fracking, and gerrymandering. No doubt, Newsela’s curation of high-interest, current news could engage struggling readers and support basic research. Wide-ranging student inquiry, problem-based learning or science and humanities competitions need more range and scope than even the text sets. Since Newsela identifies Gale Cengage as a content provider, content depth could increase dramatically in a short period of time. One would welcome better searching on phrases. For example “judicial activism” and “Executive Order” frustratingly default to OR searches (judicial OR activism; Executive OR order).

ReadWorks’ serviceable writing is well-crafted for comprehension instruction but lackluster in animating the student-driven curiosity that should be generated by problem-based learning and competitions. Like an online sister to those colorful but inauthentic print sets about trucks, turtles and Tutankhamun, the platform falls short of the topical depth and range of sources we see in school library database aggregations from vendors like ProQuest, Gale, and EBSCO.

Leveling is an attractive prison for school library researchers. Slotting students into categories by a Lexile score, whether it’s “just right reading” or at their instructional level, hobbles their exposure to the varied and rich resources that build reading fluency, stamina, and comprehension (Shanahan). That doesn’t mean you won’t teach reading through research. Watch where a student researcher gets stuck—and coach comprehension strategies rather than announce the answer. Model “search alouds” (Walker) that reveal your own reading strategies—and missteps. Show how to milk the literature review or chain backward from an article’s references to find needed background information. Becoming more skillful at differentiation—and more strategic at interventions—is not readily accomplished by leveling the library or systematically teaching within a leveling platform. “Reading in the wild,” while more instructionally demanding, holds richer, more authentic promise for both learning to read and reading to learn.

Works Cited (MLA format)

"About Newsela." Newsela, newsela.com/company/. Accessed 13 Oct. 2017.

Arabo, Melody, et al. The Right Tool for the Job: Improving Reading and Writing in the Classroom. Edited by Victoria McDougald, Thomas B. Fordham Institute, 14 Mar. 2017. Thomas B. Fordham Institute, edex.s3-us-west-2.amazonaws.com/publication/pdfs/%2803.14%29%20The%20Right%20Tool%20for%20the%20Job%20-%20Improving%20Reading%20and%20Writing%20in%20the%20Classroom.pdf.

Belle, Alysia. "FAQs: How Newsela Thinks about Reading Levels." Newsela, 2 Oct. 2017, support.newsela.com/hc/en-us/articles/217495666.

Bloomfield, Kathy. "Response to Your Query about ReadWorks." Received by the author, 17 Oct. 2017.

"Content Providers." Newsela, newsela.com/company/partners/. Accessed 13 Oct. 2017.

Davis, Nicole. "How Newsela Determines a Student's Reading Level." Newsela, 2 Oct. 2017, support.newsela.com/hc/en-us/articles/201219075.

---. "Newsela vs. Newsela PRO." Newsela, 11 Oct. 2017, support.newsela.com/hc/en-us/articles/201218935.

Fears, Darryl. "National Park Service Showed Its Bottled Water Ban Worked - Then Lifted It (1220L)." Washington Post, WP Company, 26 Sept. 2017, www.washingtonpost.com/news/energy-environment/wp/2017/09/26/the-national-park-service-showed-that-its-bottled-water-ban-worked-then-lifted-it/?utm_term=.a237d619f7df.

"National Park Service Showed Its Bottled Water Ban Worked - Then Lifted It (590L)." Newsela, 1 Oct. 2017, newsela.com/read/nps-water-bottle-ban/id/35839/.

"National Park Service Showed Its Bottled Water Ban Worked - Then Lifted It (1220L)." Newsela, 1 Oct. 2017, newsela.com/read/nps-water-bottle-ban/id/35841/.

"National Park Service Showed Its Bottled Water Ban Worked - Then Lifted It (900L)." Newsela, 1 Oct. 2017, newsela.com/read/nps-water-bottle-ban/id/35840.

"Newsela - Privacy Policy." Newsela, 22 Oct. 2016, newsela.com/pages/privacy-policy/#/interest-based-or-online-behavioral-advertising.

Palenski, Ted. "Citing a Newsela Article." Newsela, 16 Aug. 2017, support.newsela.com/hc/en-us/articles/203226905-Citing-a-Newsela-article. Accessed 10 Oct. 2017.

Parrott, Kiera. "Fountas and Pinnell Say Librarians Should Guide Readers by Interest, Not Level." School Library Journal, 12 Oct. 2017, www.slj.com/2017/10/literacy/fountas-pinnell-say-librarians-guide-readers-interest-not-level.

---. "Thinking outside the Bin: Why Labeling Books by Reading Level Disempowers Young Readers." School Library Journal, 28 Aug. 2017, www.slj.com/2017/08/feature-articles/thinking-outside-the-bin-why-labeling-books-by-reading-level-disempowers-young-readers.

"Privacy Policy." ReadWorks, 30 Sept. 2017, www.readworks.org/privacy.

"ReadWorks: Score and Rating." Charity Navigator, 1 July 2017, www.charitynavigator.org/index.cfm?bay=search.summary&orgid=16941.

Shanahan, Timothy. "New Evidence on Teaching Reading at Frustration Levels." Shanahan on Literacy, 17 May 2014, shanahanonliteracy.com/blog/new-evidence-on-teaching-reading-at-frustration-levels.

"StepReads." ReadWorks, about.readworks.org/stepreads.html. Accessed 14 Oct. 2017.

Walker, Courtney. "Search Alouds: Showing Not and Telling the Search Process." Independent Ideas, Association of Independent School Librarians, 9 Oct. 2017, aislnews.org/?p=5888.

"Wetlands and Habitat Loss." ReadWorks, 2013, www.readworks.org/article/Wetlands-and-Habitat-Loss/be342287-efca-48ee-a982-1a2b0eb5738e#!articleTab:content/.

"Wetlands and Habitat Loss: StepRead 1." ReadWorks, 2013, www.readworks.org/article/Wetlands-and-Habitat-Loss/be342287-efca-48ee-a982-1a2b0eb5738e#!articleTab:content/.