217 Mann was deeply impressed: Mann (1844). The original documents are available through “The Online Books Page.”
217 two large-scale programs: “About PISA,” OECD; NAEP: “National Assessment of Education Progress,” National Center for Education Statistics.
Oh, another cross-national assessment in which the US comes out pretty much like in other such comparisons. I include only to emphasize that the PISA is not the only basis for concerns about US performance.
218 “high-stakes testing”: NCLB mandated that each state measure every child’s progress in reading and math in grades three through eight and at least once during grades ten through twelve. See “Introduction: No Child Left Behind,” US Department of Education.
219 Congressionally mandated program: The first NAEP data were collected on a trial basis in 1969. The assessments went through several major revisions before reaching the modern format in 1992. The data from the early assessments are rarely cited because they are not directly comparable to the post-1992 results. See “Grading the Nation’s Report Card: Evaluating NAEP and Transforming the Assessment of Educational Progress,” National Academies Press, http://www.nap.edu/catalog/6296.html.
219 Participating schools and children: Sampling methods: “About the 2015 Reading Assessment,” Nation’s Report Card, http://goo.gl/ubd5f6.
220 The NAEP website also provides: All data cited here are from two Department of Education websites. A simple overview of the basic results is found on the Nation’s Report Card website. The National Center for Education Statistics website has tools for exploring the data, as well as summaries and technical reports.
220 Really? reflex: Seth Meyers and Amy Poehler, “Weekend Update: Really!?! Congress’ Birth Control Hearing,” NBC.com, February 18, 2012.
220 Claims that depend: Trust, but verify, as they said in the Reagan era.
223 I mainly focus: For the blog, see http://www.dianeravitch.net. The blog is also a forum for like-minded critics. The book is Reign of Error (Ravitch 2013). For the talk, see “Conversation: Diane Ravitch,” uploaded to YouTube by The Nation, July 19, 2010, https://goo.gl/UXAaq0.
223 “The educational foundations”: National Commission on Excellence in Education (1983).
223 two reports: National Endowment for the Arts (NEA), 2004, 2007.
223 resulted from a toxic combination: Ravitch: “Venture capitalists and for-profit firms are salivating over the exploding $788.7 billion market in K–12 education” (Lee Fang, “Venture Capitalists Are Poised to ‘Disrupt’ Everything About the Education Market,” Nation, September 25, 2014, https://goo.gl/xCg5yP).
223 The reports on the state of reading: The NEA analyses of trends in reading were not competently conducted. The authors did not understand basic data handling and analysis (see, e.g., here). The data were from surveys about reading habits; the declines were in “leisure” and “literary” reading, ignoring reading on digital platforms, amount of reading for school or employment, and other reading. Whereas the first two reports reported alarming decreases in reading, the third one, “Reading on the Rise”, documented a sudden turnaround, “a significant turning point in recent American cultural history,” some slight upticks in the amount of literary reading in the survey data. These conclusions cannot be taken at face value.
224 independent operations: NAEP is federally mandated and funded but operates independently. NCLB mandated that schools receiving Title I funding participate in the NAEP assessments if selected under its sampling procedure. The purpose was to be able to use the NAEP to benchmark the states’ own tests. See Hombo (2003).
224 Hence the many attempts to explain: Examples: Martin Carnoy and Richard Rothstein, “What Do International Tests Really Show About U.S. Student Performance?,” Economic Policy Institute, January 28, 2013, http://www.epi.org/publication/us-student-performance-testing; Nick Pearce, “Pisa Panic: Being Honest About What PISA Really Shows,” Left Foot Forward, December 3, 2013, http://goo.gl/Ivpr0a; Diane Ravitch, “David Berliner on PISA and Poverty,” Diane Ravitch’s Blog, April 12, 2014, http://goo.gl/fPa4eo.
225 “Don’t believe anyone”: Ravitch (2013), 50.
225 Flynn effect: Flynn (1987).
225 general equivalency diploma (GED) holders: Heckman & LaFontaine (2008) note that GEDs account for 15 to 20 percent of high school diplomas each year and that those with GEDs perform at the level of dropouts in the US labor market. By inflating graduation rates, “The GED program conceals major problems in American society.” See also Heckman & LaFontaine (2010).
225 The US graduation rate: See Murnane (2013). The author also documents a true (non-GED-dependent) but unexplained uptick in high school graduation rates between 2000 and 2010.
226 “In my recent book”: Diane Ravitch, “Four Lessons on New PISA Scores—Ravitch,” Washington Post, December 3, 2013. See also Ravitch (2013), 70–72.
226 Hanushek and Woessmann: Hanushek & Woessmann (2012).
226 FISS, FIRS, SIMS, SISS, SIRS, TIMSS, and PIRLS: Mmmmm, you could Google them (try “_________ cross national assessment”). They are identified in the Hanushek & Woessmann article, p. 304, and, as promised, here.
226 analyses indicated “a close relationship”: Hanushek & Woessmann (2012), 267.
226 Hanushek and others have pursued: See, e.g., Murnane et al. (2000) and Hanushek & Woessmann (2008); for a different view, see Ramirez et al. (2006).
226 A five- or eight-point change: Sean Cavanagh & Kathleen Kennedy Manzo, “NAEP Gains: Experts Mull Significance,” Education Week, October 3, 2007.
227 They matter enough: Buckley (2001).
228 A smaller percentage of US students: OECD (2013b). For the lowest level, 1, scores are further divided into three sublevels. The < 2 refers to all level 1 data.
228 A comparison of the results: Amazingly, a large representative cohort of students took both the NAEP (in 2007 when they were thirteen) and the PISA (in 2009 when they were fifteen). This allowed Peterson et al. (2011) to make closer comparisons between the tests. They concluded that the NAEP’s proficiency standard was higher for reading, but the PISA standard was higher for math.
228 “It is fair to say”: Diane Ravitch, “Every State Left Behind,” New York Times, November 7, 2005.
229 “That 27 percent”: Garrison Keillor, “Where’s St. Michael When You Really Need Him?,” Prairie Home Companion, January 29, 2008, http://goo.gl/s2dSC8.
229 only a rough index: The National School Lunch Program, established by President Harry S. Truman in 1946, is a federally assisted meal program operated in public and private nonprofit schools and residential child-care centers. Students from households with an income at or below 130 percent of the federal poverty guideline are eligible for free lunch; they qualify for subsidized lunch if the household income is between 130 and 185 percent of the guideline. Peterson et al. (2011). For 2013–2014, children from families of four with incomes less than $30,515 qualified for free lunch; the cutoff for subsidized lunch was $43,568. Children above this cutoff are in the highest SES group, which includes many of modest means.
230 Low achievement in reading: See “Closing the Opportunity Gap: What America Must Do to Give Every Student a Chance,” Stanford Center for Opportunity Policy in Education. The press release provides a good summary.
230 Gaps in educational opportunity: For a grim reminder of poverty’s impact on education, see Jaclyn Zubrzycki, “Detroit Studies Illuminate Problem of Lead Exposure,” Education Week, September 26, 2012.
231 Every PISA assessment includes: To get a fuller picture of how factors such as parental education, educational spending, and percentage of low-income participants influence country scores, I recommend looking at pp. 34–36 in a 2010 PISA in-depth report. 
Graphs (taken from here) showing how GDP, parental education, and other factors relate to PISA reading scores. A brief explanation of what the figures show will open in a new tab if you click here.
231 example from close to home: Correction for SES and US versus Canada data: OECD (2011, 34–35). According to this document, 17 percent of the variation in student performance is explained by students’ socioeconomic background in the US, compared to 9 percent in Canada (also Japan). The report also notes that in the US, “The relationship between socio-economic background and learning outcomes is far from deterministic. For example, some of the most socio-economically disadvantaged schools match the performance of schools in Finland. Furthermore…a quarter of American 15-year-olds enrolled in socio-economically disadvantaged schools reach the average performance standards of Finland, one of the best-performing education systems.”
231 “U.S. students lag”: For Duncan’s remarks about the 2009 results, see “Secretary Arne Duncan’s Remarks at OECD’s Release of the Program for International Student Assessment (PISA) 2009 Results,” US Department of Education, December 7, 2010, http://goo.gl/PAKnO. Many educators took issue with them and with his comments on the 2012 results. See Cameron Brenchley, “Duncan Calls for Higher Standards and Expectations Following PISA Results,” Homeroom, December 3, 2013, http://goo.gl/JUS5n8; Christopher H. Tienken, “Problems with PISA,” Chris Tienken, February 11,, 2014, http://christienken.com/2014/02/11/pisa-problems; and Diane Ravich, “David Berliner on PISA and Poverty,” Diane Ravitch’s blog, April 12, 2014, http://goo.gl/fPa4eo.
232 2013 report from the Economic Policy Institute: Carnoy & Rothstein (2013).
232 “damaging world-wide effects”: On the letter from educators, see “OECD and Pisa tests Are Damaging Education Worldwide—Academics,” Guardian, May 6, 2014, http://goo.gl/nl4epU. For a response from OECD, see Peter DeWitt, “What is PISA?”, Education Week, September 25, 2015, http://goo.gl/gGXCA0.
232 A much easier way to improve: Kay McSpadden, “Public Schools Aren’t Failing,” Charlotte Observer, January 30, 2015. http://goo.gl/8aeAWF.
233 “Less than 25 percent”: Kena et al. (2015).
234 “resilient” children: OECD (2011), 37.
234 “One of the basic principles”: “Educational Policy in Finland,” Finnish Ministry of Education and Culture.
235 particularly disturbing: On eligibility for free/reduced-price lunch, see “Percent Low Income Students: 2013–2014,” ED.gov.
235 educational redlining: “Redlining is the practice of arbitrarily denying or limiting financial services to specific neighborhoods, generally because its residents are people of color or are poor”: D. Bradford Hunt, “Redlining,” Encyclopedia of Chicago. I am familiar with the concept from growing up on the South Side of Chicago, at a time when the practice was rampant. Educator Linda Darling-Hammond has also described ESSA’s retention of high-stakes testing for the lowest-performing schools, which are in low-income communities, as a form of educational redlining: “Why Is Washington Redlining Our Schools?,” Nation, January 30, 2012.
236 Many factors contribute: Several perspectives on the achievement gap: Hoff (2013); Richardson (2008); Magnuson & Duncan (2006); Barton & Coley (2009); Washington, Terry, & Seidenberg (2013); Seidenberg (2013).
236 “East Versus West”: Sharon Begley, “East Versus West: One Sees Big Picture, Other Is Focused,” Wall Street Journal, March 28, 2003.
237 A good place to look is the reading data: Motoko Rich, Amanda Cox, and Matthew Bloch, “Money, Race and Success: How Your School District Compares,” New York Times, April 29, 2016, http://goo.gl/yVrTRF.
239 Studies of African American youth: Gosa & Alexander (2007).
239 These kinds of data: In analyses using a large longitudinal survey, Fryer & Levitt (2006) observed that the achievement gap present in kindergarten can be accounted for by a small number of factors related to SES. However, the gap also increased through the first several years of school for reasons they could not identify. It may be that standard SES measures are less valid for African Americans than whites (Rothstein & Wozny 2013) or that effects of SES multiply as school demands increase. It may also be that other factors such as language background start to assume more importance in school.
240 Children from lower-income families: Hart & Risley (1995). They used the labels “professional,” “working-class,” and “welfare,” which I’ve renamed as SES levels. The follow-up study on reading achievement: Walker et al. (1994).
240 Variability in spoken-language acquisition: Fernald, Marchman, & Weisleder (2013).
240 Recognition of the importance: Duncan & Sojourner (2013).
240 a study of fifty-three low-income mothers: Weizman & Snow (2001).
240 Similar variation: Kena et al. (2015); Horton-Ikard & Miller (2004); Gosa & Alexander (2007).
240 “Parents’ verbal engagement”: Fernald & Weisleder (2015). The study of the impact of teachers’ speech to children is Dickinson & Sprague (2003).
241 Most African American children: Woolfram (2004).
241 The question is not: AAE is linguistically unremarkable, an example of dialect variation as it occurs in all languages. The breakthrough research establishing this fact was conducted by sociolinguist William Labov (1972) and extended by many others.
241 Part of the “achievement gap”: These conditions were first pointed out to me by Julie Washington.
241 This occurs for several reasons: In fact the conditions governing the deletions of final phonemes are complex, and deletions are always optional. Basic facts about how often phonemes are omitted for a given word or by individual speakers are not known because the questions have not been adequately studied.
242 For an AAE speaker: Some of these phenomena also occur in the Southern White English dialect, and the impact on reading would be expected to extend to those speakers. See Wolfram (1974); Oetting & Kent (2004).
242 The impact of such pronunciation: Brown et al. (2015).
242 Acquiring this additional linguistic expertise: Bialystok et al. (2009).
242 A speaker of a minority dialect: It is also more difficult to identify AAE speakers with developmental spoken-language impairments because features that are indicative of an impairment in MAE can be grammatical in AAE: Oetting & McDonald (2001).
242 For verbally proficient children: Terry et al. (2010).
243 The impact of dialect differences: The issue was examined in the 1970s and 1980s, but the findings were inconclusive and the methods used do not hold up by modern standards. In recent years researchers have started to investigate the issue again, building on advances in research methods and findings about reading and language.
243 The evidence that speaking: Some representative studies: Terry et al. (2010); Charity, Scarborough, & Griffin (2004); Terry et al. (2012).
243 The findings are particularly sensitive: History of racist characterizations of black speech: Baratz & Baratz (1970).
243 I have used computational models: See Brown et al. (2015).
244 Both Australia and Canada have: Siegel (2010). For similar issues in Arabic, see Levin et al. (2008), and in Finland, see Latomaa & Nuolijärvi (2002).
244 Preservice teacher education rarely includes: Moats (1999, 2009).
244 negative perceptions of AAE: Blake & Cutler (2003).
244 Having to learn two dialects: From a biography of Condoleezza Rice (Mabry 2008, 13): “Then as now, many African American parents told their children, ‘You have to be twice as good.’ Meaning, they had to be twice as good as white people to receive the same level of respect, opportunity or status. ‘You were taught that you were good enough, but you might have to be twice as good given you’re black,’ Condoleezza Rice often recalled.” New data from the National Bureau of Economic Research (http://www.nber.org/papers/w21612) shows this to be essentially true with respect to getting and keeping jobs: Gillian B. White, “Black Workers Really Do Need to Be Twice as Good,” Atlantic, October 7, 2015.
244 Compelling children to learn MAE: Dudley-Marling & Lucas (2009).
244 “Often powerful”: Steiner & Rozen (2004).
245 with particular attention: For an example of casual bias, see Jennifer Holladay, “The Character of Our Content: A Parent Confronts Bias in Early Elementary Literature,” Rethinking Schools 27, no. 2 (winter 2012–2013), http://goo.gl/aFuCH. On tenacity, perseverence, and grit, see Duckworth et al. (2007). On the impact of poverty and racial bias: Wilson (2011).
245 As agents of social justice: Sleeter (1996, 152), for example, sees multicultural education as “a collaborative process involving dialog and bonding across racial and ethnic boundaries for the purpose of forging greater equality and social justice.” See also Stinson et al. (2012), who discuss pedagogical practices that link the teaching of mathematics to social justice issues.
245 “Using a theoretical approach”: Hassett (2006).
246 For lower-SES individuals: Heritability of reading depends on SES: Hart et al. (2013). Similar effects have been found for parental education (Friend, DeFries, & Olson 2008) and educational quality (Taylor et al. 2010). Note that although heritability is a measure of genetic influence on behavior, these heritability differences do not arise from the genome but rather from the impact of the environment on gene expression, brain, and behavior.