Academically Adrift Part 4

62. U.S. Department of Education, A Test of Leadership, 14.

63. The Council for Aid to Education is a national nonprofit organization based in New York City. It was originally established in 1952 by a group of corporate leaders including Alfred Sloan (General Motors), Frank Abrams (Exxon Corporation) and Irving S. Olds (U.S. Steel Corporation) as the Council for Financial Aid to Education to advance corporate support of higher education. When the CLA instrument was developed, the Council for Aid to Education was affiliated with the Rand Corporation (see http://www.cae.org).

64. Since spring 2007 we have continued to follow this cohort of students, collecting additional data including their college transcripts and CLA scores in spring 2009. We also plan to follow them up post-graduation in spring 2010 and spring 2011 to examine their transitions to the labor market or further schooling.

65. Richard Hersch, aGoing Naked,a AAC&U Peer Review 9 (2007): 6.

66. Stephen Klein, Richard Shavelson, and Roger Benjamin, aSetting the Record Straight,a Inside Higher Ed, February 8, 2007, http://www.insidehighered.com/views/2007/02/08/benjamin.

67. Hersch, aGoing Naked,a 6.

68. Klein, Shavelson, and Benjamin, aSetting the Record Straight.a 69. Richard Shavelson, aThe Collegiate Learning a.s.sessment,a Ford Policy Forum 2008: Forum for the Future of Higher Education, 20, http://net.educause.edu/forum/fp08.asp.

70. Discussed in Shavelson, aThe Collegiate Learning a.s.sessment,a 20. Prompts reported in posted powerpoints of Richard Benjamin, http://wvhepc.org/resources/RogerPresentation(10.09.03)a.ppt.

71. Council for Aid to Education, Collegiate Learning a.s.sessment Common Scoring Rubric (New York: Council for Aid to Education, 2008).

72. Shavelson, aThe Collegiate Learning a.s.sessment,a 19.

73. Ibid.

74. Ibid., 20.

75. U.S. Department of Education, A Test of Leadership, 23.

76. Anne Grosso de Len, aThe Collegiate Learning a.s.sessment: A Tool for Measuring the Value Added of a Liberal Arts Education,a Carnegie Results, Fall 2007, 3.

77. James Traub, aNo Gr_du_te Left Behind,a New York Times Magazine: The College Issue, September 30, 2007, 106a"9.

78. Doug Lederman, aNo College Left Behind?a Inside Higher Ed, February 15, 2006, http://www.insidehighered.com/news/2006/02/15/testing.

79. Ibid.

80. Ibid.

81. U.S. Department of Education, A Test of Leadership, 23.

82. Organisation for Economic Co-operation and Development (OECD), Directorate for Education, Education Committee, Centre for Educational Research and Innovation (CERI) Governing Board, PISA for Higher Education (Paris: Organization for Economic Co-operation and Development, 2006), 3.

83. Organisation for Economic Co-operation and Development, aa.s.sessment of Higher Education Learning Outcomes (AHELO),a http://www.oecd.org/doc.u.ment/

41/0,3343,en_2649_35961291_42295209_1_1_1_1,00.html.

84. Catherine Hoffman Breyer, aThe Right Way to Measure College Learning: National Standardized Testing Wonat Work,a Christian Science Monitor, April 9, 2007, http://www.csmonitor.com/2007/0409/p09s01-coop.html.

85. Jennifer Epstein, aQuestioning College-Wide a.s.sessments,a Inside Higher Ed, June 21, 2007, http://www.insidehighered.com/news/2007/06/21/a.s.sessments.

86. Stephen Klein, Ou Lydia Liu, and James Sconing, Test Validity Study (TVS) Report, September 29, 2009, http://www.voluntarysystem.org/docs/reports/TVSReport_Final.pdf.

87. Scott Jaschik, aDoes aValue Addeda Add Value?,a Inside Higher Ed, November 3, 2006, http://www.insidehighered.com/news/2006/11/03/a.s.sess.

88. Paul Basken, aTest Touted as 2 Studies Question Its Value: Small Colleges Back Achievement Exam to Measure Accountability,a Chronicle of Higher Education, June 6, 2008, http://chronicle.com/article/Test-Touted-as-2-Studies/23503/.

89. Robert Frank and Phillip Cook, aItas a Winner Take All Market,a Washington Monthly, December 1, 1995.

90. For more information on the Wabash National Study of Liberal Arts Education, see http://www.liberalarts.wabash.edu/study-overview/.

91. See, for example, Astin, What Matters in College, 199f.

92. William Bowen and Derek Bok, The Shape of the River (Princeton, NJ: Princeton University Press, 1998), x.x.xi.

93. Ibid., 77.

94. James Shulman and William Bowen, The Game of Life: College Sports and Educational Values (Princeton, NJ: Princeton University Press, 2001).

95. Camille Charles, et al., Taming the River: Negotiating the Academic, Financial, and Social Currents in Selective Colleges and Universities (Princeton, NJ: Princeton University Press, 2009), 2a"3. See also Douglas Ma.s.sey, et al., The Source of the River: The Social Origins of Freshmen at Americaas Selective Colleges and Universities (Princeton, NJ: Princeton University Press, 2003).

96. Charles et al., Taming the River.

97. Ma.s.sey et al., The Source of The River, 29.

98. Information for College and Beyond is provided in Bowen and Bok, The Shape of the River, Table D3.1. An additional 7 percent of students were estimated to have graduated from inst.i.tutions other than those in which they had initially enrolled, thus producing the overall graduation rate of 92 percent. National averages vary depending on the dataset and year of entry. For some estimates, see Clifford Adelman, The Toolbox Revisited (Washington, DC: U.S. Department of Education, 2006), table 30. Adelman reported that the six-year graduation rate from the same inst.i.tution for students completing high school in 1992 (which is close to the 1989 cohort included in C&B) and entering four-year colleges and universities was 53 percent. An additional 11 percent of students from this high school cohort graduated within six years from a different four-year inst.i.tution than where they had started.

Chapter 2.

1. Michael Hout, aPolitics of Mobility,a in Generating Social Stratification, ed. Alan C. Kerckhoff (Boulder, CO: Westview Press, 1995), 301a"25. Although most of the expansion occurred in public two-year inst.i.tutions, four-year inst.i.tutions expanded as well. Enrollments in public four-year inst.i.tutions as a fraction of the college-age population doubled between 1962 and 1992.

2. National Center for Education Statistics (NCES), Digest of Education Statistics (Washington, DC: U.S. Department of Education, 2008), tables 189 and 265.

3. James Rosenbaum, Beyond College for All: Career Paths for the Forgotten Half (New York: Russell Sage Foundation, 2001). Also, for increase in expectations over time, see John Reynolds et al., aHave Adolescents Become Too Ambitious? High School Seniorsa Educational and Occupational Plans, 1976 to 2000,a Social Problems 53 (2006): 186a"206.

4. NCES, Digest of Education Statistics, table 202. Moreover, this percentage underestimates the ultimate enrollment because more than one-third of college entrants today delay entry by at least one year. See National Center for Education Statistics (NCES), Waiting to Attend College: Undergraduates Who Delay Their Postsecondary Enrollment, NCES 2005-152 (Washington, DC: U.S. Department of Education, 2005).

5. Martin Trow, aReflections on the Transformation from Ma.s.s to Universal Higher Education,a Daedalus 99 (1970): 3a"4.

6. Forty-four percent of students who were cla.s.sified in government reports as marginally qualified or not qualified for college (based on their secondary school GPA, high school rank, test scores, and academic coursework), expected to finish college, and an additional 46 percent expected to attain at least some postsecondary education. National Center for Education Statistics (NCES), Access to Postsecondary Education for the 1992 High School Graduates, NCES 98-105 (Washington, DC: U.S. Department of Education, 1997), 21a"34.

7. Rosenbaum, Beyond College for All, 59a"62.

8. Barbara Schneider and David Stevenson, The Ambitious Generation: Americaas Teenagers Motivated but Directionless (New Haven: Yale University Press, 1999), 79a"85.

9. Jeffrey Jensen Arnett, Emerging Adulthood: The Winding Path from the Late Teens through the Twenties (New York: Oxford University Press, 2004), 125a"26.

10. Secretary of Education Margaret Spellings, aStatement on International Education Week 2008,a http://www.iew.state.gov/2008/docs/2008sec-ed-statement.pdf. Emphasis added.

11. Statement made by Joan Scott, Harold F. Linder Professor of Social Science at the Inst.i.tute for Advanced Study in Princeton, on behalf of AAUP, as part of testimony before the Pennsylvania General a.s.semblyas House Select Committee on Student Academic Freedom, November 9, 2005, http://www.aaup.org/AAUP/GR/state/Academic+Bill+of+Rights-State+Level/Scotttestimony.htm.

12. Higher Education Research Inst.i.tute (HERI), The American College Teacher: National Norms for 2007a"2008 (Los Angeles: HERI, University of California Los Angeles, 2009).

13. This number reflects the difference between CLA performance task score in the fall of 2005 and spring of 2007, divided by the standard deviation of the 2005 score (see table A2.1). Standard error for the 2005 score, adjusted for cl.u.s.tering of students within schools, is 21.11 with a 95-percent confidence interval of [1,090.56, 1,173.32]; standard error for the 2007 score, adjusted for cl.u.s.tering of students within schools, is 23.62 with a 95-percent confidence interval of [1,119.96, 1212.56]. As is discussed in more detail in the methodological appendix, the performance task scores in 2005 were capped at 1,600, while they were allowed to range up to 1,800 in 2007. If we capped the 2007 scores at 1,600, the mean, and consequently the estimated growth, would be slightly lower (the 2007 CLA mean would be 1,162.55, with the growth of 30.61 points or 0.16 standard deviation). Deleting 2007 CLA scores above 1,800 from a.n.a.lysis would decrease the mean, and thus the growth, even further. The limited growth reported in this chapter may thus slightly overstate studentsa gains during their first two years in college.

14. Ernest T. Pascarella and Patrick T. Terenzini, How College Affects Students: A Third Decade of Research (San Francisco: Jossey-Ba.s.s, 2005), 156, 205. It is worthwhile to note that the average improvement of 0.50 standard deviation during college would imply an improvement of 0.06 standard deviation each semester, or 0.18 standard deviation over three semesters, which matches our estimate.

15. This estimate of 45 percent is based on a paired sample t-test of the difference in studentsa performance between fall 2005 and spring 2007. Specifically, it references the proportion of students who gain less than the upper bound of the 95-percent confidence interval (i.e., 1.96 a standard error of the difference). Other, more conservative estimation strategies would place even more students below the cutoff of significant gains. If we adjusted standard error for cl.u.s.tering of students within schools, the percentage of students demonstrating no significant gains in learning would be slightly higher: 47 percent. Moreover, if we used the standard error of the 2005 CLA score, as opposed to the growth, 53 percent of students would fall below the level of statistically significant gains. If we replicate this final calculation using standard error of the 2005 CLA score and a.s.sume constant rates of growth through the senior year, we would still expect 47 percent of students to show no significant gains in critical thinking, a.n.a.lytical reasoning, and written communication by the time they graduate. A test such as the CLA which relies on open-ended prompts may face challenges of reliability, raising the possibility that some of the students showing no gains may actually be learning. However, questions of reliability are likely to pertain to the other half of the distribution as well, meaning that some of the students reporting gains may not actually be learning much.

16. Charles Blaich, aOverview of Findings from the First Year of the Wabash National Study of Liberal Arts Educationa (Wabash College, Center for Inquiry in the Liberal Arts, 2007, http://www.liberalarts.wabash.edu/research/). See also a review of previous research in Pascarella and Terenzini, How College Affects Students, 155a"212.

17. Blaich, aOverview of Findings from the First Year of the Wabash National Study of Liberal Arts Education.a 18. Lamont A. Flowers et al., aHow Much Do Students Learn in College?a Journal of Higher Education 72 (2001): 565a"83.

19. National Survey of Student Engagement (NSSE), Experiences That Matter: Enhancing Student Learning and Success (Bloomington, IN: Center for Postsecondary Research, Indiana University Bloomington, 2007), 42.

20. George D. Kuh, aWhat We Are Learning About Student Engagement From NSSE,a Change 35 (2003): 27.

21. As quoted by David Leonhardt, aThe College Dropout Boom,a New York Times, May 24, 2005, http://www.nytimes.com/2005/05/24/national/cla.s.s/EDUCATION-FINAL.html?pagewanted=1.

22. Michael Katz, ed., School Reform Past and Present (Boston: Little Brown and Company, 1971), 141.

23. Pierre Bourdieu, aCultural Reproduction and Social Reproduction,a in Knowledge, Education, and Cultural Change, ed. Richard Brown (London: Tavistock, 1973), 71a"112. There are other views of social reproduction, such as those articulated in Samuel Bowles and Herbert Gintis, Schooling in Capitalist America (New York: Basic Books, 1976); and Randall Collins, Credential Society (New York: Academic Press, 1979).

24. In addition to parental education, we examined the relationship between parental occupation and CLA scores. However, parental occupation was not statistically significant in any of the models examined. We include it as a control in regression a.n.a.lyses, but do not extensively discuss the results.

25. African-American studentsa 2005 CLA scores were 0.94 standard deviation lower than those of white students. Standard deviations are calculated by dividing the gap between African-American and white students by the standard deviation of the full sample.

26. If we standardized test scores to the 2005 CLA scale, the gap between students whose parents had no college experience and those whose parents held graduate or professional degrees would be 0.64 standard deviation in 2005 and 0.62 standard deviation in 2007.

27. Josipa Roksa et al., aChanges in Higher Education and Social Stratification in the United Statesa in Stratification in Higher Education: A Comparative Study, ed. Yossi Shavit, Richard Arum, and Adam Gamoran (Stanford, CA: Stanford University Press, 2007), 165a"91. See also Yossi Shavit and Hans-Peter Blossfeld, Persistent Inequality: Changing Educational Attainment in Thirteen Countries (Boulder, CO: Westview Press, 1993).

28. If we standardized test scores to the 2005 CLA scale, the gap of 0.94 standard deviation between African-American and white students from 2005 would increase to 1.12 in 2007.

29. NCES, Digest of Education Statistics, tables 188 and 268.

30. Seasonal learning literature indicates that inequality between different groups of students grows particularly during the summer, when students are not in school. For a recent example, see Douglas B. Downey, Paul T. von Hippel, and Beckett A. Broh, aAre Schools the Great Equalizer? Cognitive Inequality during the Summer Months and the School Year,a American Sociological Review 69 (2004): 613a"35.

31. Annette Lareau, Unequal Childhoods: Cla.s.s, Race, and Family Life (Berkeley: University of California Press, 2003), 5.

32. Ibid., 176a"77.

33. Samuel R. Lucas, Tracking Inequality: Stratification and Mobility in American High Schools (New York: Teachers College Press, 1999).

34. Schneider and Stevenson, The Ambitious Generation, 91a"96.

35. Annette Lareau and Elliot B. Weininger, aCla.s.s and the Transition to Adulthooda in Social Cla.s.s: How Does it Work?, ed. Annette Lareau and Dalton Conley (New York: Russell Sage Foundation, 2008), 127a"28.

36. NCES, Access to Postsecondary Education.

37. Students could take the SAT or the ACT. Their respective test scores were converted to an SAT equivalent scale.

38. Lareau and Weininger, aCla.s.s and the Transition to Adulthood,a 127a"28, 139.

39. Price quoted by Kaplan for thirty-two hours of tutoring in the Premier Tutoring: SAT Masters Program. http://www.kaptest.com/College/SAT/comparison.html#.

40. As quoted by Julie Bick, aThe Long (and Sometimes Expensive) Road to the SAT,a May 28, 2006, New York Times, http://www.nytimes.com/2006/05/28/business/yourmoney/28test.html?scp=21&sq=SAT%20prep&st=cse.

41. Indeed, in an extensive literature review Adam Gamoran noted: aThe most important reason for educational inequality between blacks and whites is socioeconomic.a Adam Gamoran, aAmerican Schooling and Educational Inequality: A Forecast for the 21st Century,a Sociology of Education 75 (2001): 137.

42. We focus the discussion on racial composition of the high school. However, we have also collected information on other high school characteristics, such as the proportion of students receiving free or reduced lunch and school control (public versus private). The free / reduced-price lunch variable was missing in approximately 20 percent of the cases and was not significantly related to CLA scores in regression models. Consequently, we did not include it in our presented a.n.a.lyses. Private school students were substantially underrepresented in our sample (including only 59 students, or 2.5 percent of the sample). The low number of cases raised questions about the reliability of the private school estimates; school sector was thus excluded from the a.n.a.lyses.

43. James Coleman et al., Equality of Educational Opportunity (Washington, DC: U.S. Government Printing Office, 1966).

44. Gary Orfield and Susan E. Eaton, Dismantling Desegregation: The Quiet Reversal of Brown v. Board of Education (New York: The New Press, 1996), 53, 64a"70.

45. Douglas Ma.s.sey et al., The Source of the River: The Social Origins of Freshmen at Americaas Selective Colleges and Universities (Princeton, NJ: Princeton University Press, 2003), 93a"97.

46. The same finding is reported for a sample of students attending selective colleges and universities in Ma.s.sey et al., The Source of the River, 105.

47. Rosalyn Mickelson, aSegregation and the SAT,a Ohio State Law Journal 67 (2006): 157a"99.

48. For some explanations of the underlying mechanisms and long-lasting consequences of racially segregated environments on college students, see Camille Charles et al., Taming the River: Negotiating the Academic, Financial, and Social Currents in Selective Colleges and Universities (Princeton, NJ: Princeton University Press, 2009), 150a"71.

49. For a recent review see Min Zhou, aGrowing Up American: The Challenge Confronting Immigrant Americans and Children of Immigrants,a Annual Review of Sociology 23 (1997): 63a"95.

50. The difference in CLA performance between students from English-speaking versus non-English-speaking households was statistically significant in 2007 but not in 2005. Moreover, regression a.n.a.lyses predicting 2007 CLA scores while controlling for 2005 CLA scores reveal a statistically significant relationship between English home language background and learning only in some model specifications (see tables A2.3 and A4.5 in methodolgical appendix). While present, the differences in learning for students from English versus non-English language backgrounds are thus relatively small.

51. Karl L. Alexander and Aaron M. Pallas, aSchool Sector and Cognitive Performance: When is a Little a Little?,a Sociology of Education 58 (1985): 120. Emphasis added.

52. For a statistical explanation of why this formulation can be considered to estimate growth between two time points, see Thomas Hoffer, Andrew M. Greeley, and James S. Coleman, aAchievement Growth in Public and Catholic Schools,a Sociology of Education 58 (1985): 82.

53. If we standardized test scores to the 2005 CLA scale, the African-American-white gap of 136 points would equal almost three-quarters of a standard deviation.

54. The models include sociodemographic characteristics: gender, race / ethnicity, non-English language background, parental education and occupation, household composition (two-parent household and number of siblings), and high school characteristics: region, urbanicity, and 70 percent or more non-white high school.

55. If we standardized test scores to the 2005 CLA scale, the African-American-white gap of 47 points would equal one-quarter of a standard deviation, even after adjusting for family background and academic preparation. This finding is consistent with previous research on the black-white test score gap: social background and academic experiences account for some but not all of the gap between the two groups. See for example, Christopher Jencks and Meredith Phillips, The Black-White Test Score Gap (Washing ton, DC: Bookings Inst.i.tution Press, 1998), 1a"51. Moreover, it is worthwhile to note that in the final model, after including all of the demographic and academic controls, Asian students have lower growth in critical thinking, complex reasoning, and writing skills than white students. This is due in part to the unique sampling strategy of this study, which relied on inst.i.tutions as the primary sampling unit. Asian students in our sample are concentrated in four inst.i.tutions that do not experience the highest gains in CLA. Moreover, although Asian students on average tend to do well academically and have higher high school GPAs and SAT / ACT scores than white students (authorsa calculations based on the Beginning Postsecondary Students (BPS) Longitudinal Study, 2003a"04 cohort), that is not the case for our sample. The Asian students in our sample are thus not necessarily representative of Asian students attending four-year inst.i.tutions nationwide.

56. Christopher Jencks, et al., Inequality: A Rea.s.sessment of the Effect of Family and Schooling in America (New York: Basic Books, 1972), 135.