The determinants of electronic textbook use among college students.
Miller, Jon R. ; Nutting, Andrew W. ; Baker-Eveleth, Lori 等
I. Introduction
Electronic media appear to be winning the battle for readership in
the United States, as bookstores close (Roso & Rosenwald 2011) and
print newspapers downsize or go out of business entirely (Perez-Pena
2009; Richman & James 2009). According to a recent sales report from
the Association of American Publishers, adult e-book sales rose 49.4% in
January 2012, more than triple the growth in mass market paperback
sales. Sales of children's and religious e-books rose 475% and
151%, respectively (Association of American Publishers 2012).
Sales figures for electronic textbooks (e-textbooks), a specific
type of e-book, are harder to find, but estimates suggest that their use
is rising. According to the National Association of College Stores
(2010), national e-textbook revenue in 2010 was 3% of total textbook
revenue, but that share was expected to grow to 10 or 15 percent by 2012
(Foderaro 2010; Schuetze 2011; Krapfl 2011).
Rapid evolution characterizes the e-textbook market. In this
research, we take a snapshot of this evolution by analyzing e-textbook
use at a particular university at a particular time. Using data on
individual students at the University of Idaho, we use logit analysis to
answer questions about the determinants of e-textbook use. Why have some
students used e-textbooks while others have not? Do different college
majors influence use? Do students' backgrounds and socioeconomic
characteristics affect e-textbook use?
The remainder of this paper is organized as follows. Section II
discusses providers of e-textbooks. Section III surveys the scant
literature regarding students' propensities of using electronic
textbooks. Section IV discusses the data collection process and summary
statistics. Section V discusses the estimation strategy and results.
Section VI concludes.
II. E-textbook Providers
The number of providers of e-textbook content continues to expand.
Digital publisher flatworld-KNOWLEDGE has been in business since 2007.
In May 2011, they listed 35 titles with an expected doubling in the
number of titles by spring 2012, according to Jon Williams, Chief
Technology Officer at flatworldKNOWLEDGE (Hampson 2011). The
flatworldKNOWLEDGE textbooks are free to view online, or a student can
pay for sections to be printed or the entire book for around $29.95.
CourseSmart also began in 2007, and is a partnership of traditional book
publishers, such as Cengage, Pearson and McGraw-Hill. Through the
partnership, CourseSmart provides not only e-textbooks, but also an
online catalog of electronic resources and other digital course material
(Graham 2011). The e-textbooks are available to purchase as a PDF for a
180-day subscription. Another e-textbook supplier, VitalSource, is
supported by publisher Macmillan and provides a customizable textbook
using a platform called DynamicBooks. These e-textbooks can be read on
e-readers such as a Kindle or an iPad.
Universities are experimenting with e-textbook programs. The
University of Phoenix consolidated all course textbooks in an electronic
library, charging $75 a semester for electronic access to any textbook
(Blumenstyk 2008). Northwest Missouri State University ran a pilot
program with 240 students who were loaned e-book reading devices and
provided with electronic textbooks (Young 2009). The University of Idaho
has tested a method where professors provide an electronic, custom text
tailored to a specific course and charge for it with a course fee
(Baker-Eveleth et al. 2011). Open source programs, such as GNU Free
Documentation License (GFDL), allow the authors of the content as well
as readers to make suggestions and note errors and corrections (Beezer
2009).
III. Literature on Determinants of E-Textbook Use
In spite of the rapid growth and development of e-books and
e-textbooks, little research on it exists. What we know comes from
student surveys and anecdotal evidence in the discussion of the pros and
cons of e-textbooks and their print alternatives. For example, a report
on research sponsored by the National Association of College Stores
(NACS) noted that three-quarters of students surveyed preferred a bound
book to a digital version (NACS 2010; Foderaro 2010). Student surveys
indicate a strong preference for hard copy texts, even though the
current generation of college students is very comfortable with
electronic information and communication (Nelson 2008a; Dudley 2011).
Some authors note that e-textbook sales growth could be limited by
unique factors not facing regular e-Books. Graphics and mathematics,
common in college texts, are more difficult to transmit on popular
e-readers such as Amazon's Kindle. Hard copy books don't get
blank screens or viruses and they allow the reader to flip quickly
through chapters and write easily in the margins (Foderaro 2010). And if
e-textbooks are not available on special readers, students resist being
"tethered" to a computer and suffering eye strain and back and
neck problems from reading on computer screens (Carlson 2005; Crawford
2006; Nelson 2008b).
Some point to advantages of e-textbooks. They allow an environment
where students can interact and engage with the material in a different
way. Most digital books are searchable, can be highlighted like a
traditional textbook, and often have a comment box or annotation ability
on the pages (Ravid, Kalman and Rafaeli 2008). E-textbooks are easier to
update and edit. Traditional hard copy textbooks from publishers are
limited in how quickly they can correct errors or make updates to the
textbook (Larson 2002; Stewart 2009).
The rising cost of hard copy textbooks appears to be causing
students to consider electronic alternatives to traditional textbooks
(Marklein 2012). According to the U.S. Government Accountability Office
(2005), between 1987 and 2004 the average price of a college textbook
increased an average of 6% per year, twice as fast as the consumer price
index. On average, a college textbook costs $125, and many are not
resalable on the used textbook market. The total textbook cost for a
typical year of classes is close to $1,000 (Christopher 2009; Rampell
2008). This increased cost creates financial pressure for college
students and their families. Students, parents, state legislatures,
federal agencies, university bookstores and book publishers have
responded to rising textbook costs by making it an important and
controversial issue on college campuses (Chaker 2006; Kang 2004;
Kingsbury & Galloway 2006; Roberts 2006).
IV. Data
Data were collected via a web-based questionnaire in November 2009
conducted at the University of Idaho. Though our data are
campus-specific, answers should be extendable to other four-year
colleges. The University of Idaho is a public land-grant, largely
residential institution in Moscow, Idaho. It consists of approximately
12,000 students, over 80 percent of whom are undergraduates. Its
Carnegie Classification of "research university" with
"high undergraduate enrollment" is shared by 63 other public
universities. (1)
An invitation to participate in the survey was distributed via a
listserv to 11,957 University of Idaho students. (2) A total of 1,382
responses were received, producing an 8.65% response rate, and almost
1,100 responses possessed complete enough answers to be included in this
paper. Summary statistics for relevant variables derived from the
questionnaire are presented in Table 1. A total of 44.5% of students had
purchased or used an electronic textbook. Almost 60 percent of students
were female, whereas under half of degree-seeking undergraduates at the
University of Idaho are female (US News and World Report 2012). The
average student's high school had a graduating class of
approximately 270, (3) and 7.4% of the sample graduated from a private
high school. Approximately 65 percent of respondents received financial
aid, and over two-thirds received scholarships. A majority, over 55
percent, worked either full-time or part-time while in school. Almost
half of respondents received tuition assistance from parents or family.
Almost 93 percent of respondents owned a laptop computer. Fewer
than 40 percent owned a desktop. Survey respondents were asked to rate
on a scale of 1-5 (from "Strongly Disagree" to "Strongly
Agree") whether they agreed with the statement "My professors
think I should use an electronic textbook." About 8.5 percent of
students reported that their professors "strongly disagreed"
with the use of electronic textbooks, while 5.9 percent reported that
their professors "agreed" or "strongly agreed" with
their use. The College of Letters, Arts, and Social Science accounted
for over one quarter of respondents, while the Colleges of Business and
Economics, Education, and Engineering each comprised over 10 percent of
respondents. (4)
Table 2 shows percentage of respondents, by college, who had ever
purchased or used an electronic textbook. Four of every five students in
the College of Business and Economics (CBE) had used an electronic
textbook. Between 30 and 50 percent of students in each of the other
colleges had used an electronic textbook.
V. Estimation Strategy and Results
Estimations determining what affects a student's likelihood of
having purchased or used an electronic textbook take the form
[Y.sup.*.sub.ic] = [beta][X.sub.ic] + [alpha][Z.sub.ic] +
[[delta].sub.c] + [[epsilon].sub.ic] (1)
where i is individual student and c is college of enrollment.
[Y.sub.ic.sup.*] is the latent propensity for a student to have
purchased or used an electronic textbook. [X.sub.ic] represents
student-level demographic and background controls. (In estimations, the
linear variable AGE is set equal to zero for students whose AGE_OVER_40
variable is equal to 1.) [Z.sub.ic] represents controls regarding the
attitudes of each student's professors towards electronic
textbooks. [[delta].sub.c] represents fixed effects controls for college
and [[epsilon].sub.ic] is a random error term.
Since propensity to purchase an electronic textbook is latent and
unobserved, [Y.sub.ic] is defined as a dummy variable equal to 1 if
student i purchases an electronic textbook and 0 otherwise.
[[epsilon].sub.ic] in Equation (1) is logistically distributed and
maximum likelihood logit estimations of the equation
P([Y.sub.ic] = 1|[X.sub.ic], [Z.sub.ic]) =
[LAMBDA]([beta][X.sub.ic] + [alpha][Z.sub.ic] + [[delta].sub.c]) (2)
are estimated.
Table 3 shows marginal effects from logit estimations of Equation
(2) on the population of all surveyed students. (5) In Column 1, only
student demographic and background controls are included as explanatory
variables. Significantly negative marginal effects on AGE and
AGE_OVER_40 indicate that younger students are more likely to have
purchased or used an electronic textbook. Students from larger high
schools, students financing their educations through financial aid
and/or scholarships, and students who own either a laptop or a desktop
computer are significantly more likely to have purchased or used an
electronic textbook. (6)
Column 2 adds fixed effects for college, to control for the
between-college differences in use of an electronic textbook shown in
Table 1b. The dummy representing College of Letters, Arts, and Social
Sciences (CLASS) is omitted for identification purposes. Students from
four colleges--the College of Business and Economics, the College of
Engineering, and College of Science, and the College of Agricultural
& Life Sciences--are significantly more likely to have used an
electronic textbook than students from CLASS. Many controls which were
significant in Column 1 are insignificant in Column 2. AGE, AGE_OVER_40,
and SCHOLARSHIPS become insignificant, suggesting that younger students
and students who receive scholarships are more likely to have used
electronic textbooks because they are enrolled in colleges where
students are more likely to use them. Ownership of either a desktop or a
laptop computer is no longer significantly correlated with use of an
electronic textbook. Students receiving financial aid and students from
larger high schools remain significantly more likely to have purchased
or used an electronic textbook.
Column 3 adds controls for professors' attitudes towards
electronic textbooks. Students who have had professors who either
"agree" or "strongly agree" with the use of
electronic textbooks are more likely to have used electronic textbooks,
and those whose professors have "strongly disagreed" with the
use of electronic textbooks are less likely to have used them.
FINANCIAL_AID and HIGH_SCHOOL_ CLASS_SIZE remain significantly positive.
OWN_ DESKTOP becomes significantly positive again, after having been
insignificant in Column 2.
Table 3 thus suggests that students on financial aid, students who
own desktop computers, and students from larger high schools are more
likely to have purchased or used electronic textbooks. These all almost
certainly reflect demand-side factors. That students on financial aid
are more likely to have used an electronic textbook suggests that
electronic textbooks are an inferior good, i.e. while students with more
resources demand traditional bound textbooks (a normal good), students
with fewer resources demand their less expensive electronic
counterparts. The frequently significant coefficient on OWN_DESKTOP
indicates that desktop computers and electronic textbooks are
complementary goods. Laptops, being so widely owned, are not the
specific complement to electronic textbooks that desktop computers are.
That students from larger high schools more frequently used electronic
textbooks suggests that larger high schools expose their students to
more computer-intensive modes of learning, giving their graduates a
higher willingness-to-pay for using electronic textbooks in college. (7)
Table 4 re-estimates Equation (2) on two subsets of the original
sample. All 4 columns include fixed effects controls for college.
Columns 1-2 show results when omitting College of Business and Economics
(CBE) students, who are much more likely to have used electronic
textbooks. In both columns, HIGH_SCHOOL_CLASS_SIZE, FINANCIAL_AID and
OWN_DESKTOP remain significantly positive. AGE_OVER_40 is significantly
negative in both columns.
Columns 3-4 include CBE students but limit the sample to students
under the age of 30. Larger high schools and financial aid status are,
once again, significantly positive in all three columns, and OWN_DESKTOP
is significant in Column 4. Among these younger students, graduates of
private high schools and students receiving scholarships are
significantly more likely to have used electronic textbooks. Column 4
also shows that female students are significantly more likely to have
used electronic textbooks. (8)
In sum, Tables 3 and 4 show strong, significant, and persistent
evidence that students on financial aid and students from larger high
schools are more likely to have purchased or used electronic textbooks.
These findings suggest, respectively, that e-textbooks are inferior
goods, and that students from larger high schools (which are usually
located in cities or suburbs) may have been exposed to more
computer-intensive modes of learning. There is some evidence that
students who own desktop computers are more likely to have used an
electronic textbook, suggesting that desktops and e-textbooks are
complementary goods. And university-level factors--such as college of
enrollment and professor attitudes towards electronic
textbooks--significantly influence whether students have used them.
VI. Conclusions and Future Research
This paper examines what factors affect college students'
likelihood of having purchased or used an electronic textbook. There are
strong, significant, and persistent results showing that students from
larger high schools and students who own desktop computers are more
likely to have used an electronic text, and some less-persistent results
showing that younger students are more likely to have used one. These
results jointly suggest that experience with and technical competence
with information technology increase students' use of e-textbooks.
Additionally, students paying for their education with scholarships and
loans are more likely to have used an electronic text, suggesting that
e-textbooks are inferior goods, in the strict economic sense. Putting
these findings together, as technical competence with information
technology expands, as it surely will, and as the cost of college
continues to rise, we would expect the demand for e-textbooks to rise.
There is a strong relationship between college of enrollment and
having used an electronic textbook. Students in four arguably
more-technical colleges--Business and Economics, Science, Engineering,
and Agriculture and Life Sciences--are more likely to have used an
electronic text than are students in the College of Letters, Arts and
Social Sciences. Results are especially strong for the College of
Business and Economics, where over 80% of the students report use of an
electronic text, possibly because some required courses in the college
often use electronic text materials.
Results also persistently show that professor attitudes towards
electronic textbooks significantly affect students' use of them.
The high rate of e-textbook use in the College of Business and
Economics, where use of some e-textbooks are required, and evidence of a
distaste among professors in general for e-textbooks (Nelson 2008b),
jointly suggest that e-textbook use may increase as professors'
attitudes towards them change. Therefore, trends in competence with
information technology, college costs, and professorial and
institutional attitudes towards electronic textbooks suggest that rates
of e-textbook use will increase over time. We are interested in this as
a topic of future research, perhaps with a panel dataset.
There is also the chance that technological advances have changed
the market for electronic textbooks since this survey was taken in Fall
2009. Electronic textbooks are more commonly featuring interactive
features (Young 2010), and the 2010 introductions of the Apple iPad
(Stone and Miller 2012) and the fourth edition of Amazon's Kindle
(Miller 2011) may have made electronic textbooks more easily utilized.
If students from higher-income backgrounds are disproportionately more
likely to use electronic textbooks under these conditions (Savitz 2011),
our findings that students on financial aid are more likely to use
electronic textbooks could be weakened or even reversed.
APPFNDIX TABLE 1A.
Sample Population
AGE (Sample: if 40 or under) 22.8 24.0
AGE_OVER_40 0.07
FEMALE 0.58 0.46
HIGH_SCHOOL_CLASS_SIZE 266.7 129.0
FINANCIAL_AID * 0.65 0.58
AGRICULTURE_ 0.10 0.08
LIFE_SCIENCES
ART_ARCHITECTURE 0.05 0.07
BUSINESS_ECONOMICS 0.13 0.10
EDUCATION 0.13 0.14
ENGINEERING 0.16 0.14
LETTERS_ARTS_SOC_SCI 0.27 0.28
NATURAL_RESOURCES 0.07 0.06
SCIENCE 0.10 0.09
* = Population Mean observed for Undergraduates Only
References
Association of American Publishers. 2012. "AAP monthly
snapshot." January 2012. Retrieved April 12,2012
(http://www.publishersweekly.com.)
Baker-Eveleth, Lori, Jon R. Miller and Laura Tucker. 2011.
"Lowering Business Education Cost with a Custom Professor-written
Online Text." Journal of Education for Business, 86(4): 248-252.
Beezer, Rob. 2009. "The truly free textbook." EduCause
Review, 44(1): 23-24.
Blumenstyk, Goldie. 2008. "To cut costs, ought colleges look
to for-profit models?" Chronicle of Higher Education, June 13. pp.
A19-20.
Carlson, Scott. 2005. "Online textbooks fail to make the
grade." Chronicle of Higher Education, February 11, pp. A35-36.
Chaker, Anne Marie. 2006. "Efforts mount to cut costs of
textbooks: As prices rise at twice the rate of inflation, states pass
laws to encourage cheaper alternatives." The Wall Street Journal,
September 28, pp. 1-3.
Christopher, Carol L. 2009. "Academic Publishing: Digital
alternatives to expensive textbooks." Seyhold Online, 8(19): 11-14.
Retrieved April 6, 2009 (http://www.seyboldreport.com/
book-publishing/academic-publishing-digitalalternatives-expensive-textbooks.)
Crawford, Walt. 2006. "Why aren't ebooks more
successful?" EContent 29(8): pp. 44.
Dudley, Brier. 2011. "Kindle So-So for students. UW Study
Concludes." Seatle Times. Retrieved Jan. 7, 2012
(http://seattletimes.nwsource.com/
html/technologybrierdudleysblog/2014937738_
kindle_so-so_for_students_uw_s.html.)
Foderaro, Lisa W. 2010. "In a Digital Age, Students Still
Cling to Paper Textbooks." The New York Times. Oct. 20. Retrieved
Jan. 7, 2012 (http://www.nytimes.com/2010/10/20nyregion/ 20textbooks.)
Graham, Jefferson. 2011. "Inkling opens textbooks on the
iPad." USA Today. August 3. Retrieved Feb. 26, 2012
(http://usatoday30 .usatoday.com/tech/products/2011-08-02-inklingipad-e-textbooks_n. htm.)
Hampson, Keith. 2011. "Interview: Ed tech leader, Jon Williams
of Flat World Knowledge." Retrieved May 2, 2012
(http://bx.businessweek .com/higher-education/view?url=http%3A%2F
%2Fhighereducationmanagement.wordpress.com %2F2011
%2F05%2F02%2Finterview-ed-techleader-jon-williams-of-flat-world-knowledge%2F).
Kang, Stephanie. 2004. "New options for cheaper textbooks:
Under fire for high prices, publishers push alternatives; renting your
chem book." The Wall Street Journal, August 24. p. 1.
Kingsbury, Alex and Lindsey Galloway. 2006. "Textbooks enter
the digital era." U.S. News and World Report, 141, October 16, pp.
63-65.
Krapfl, Anne. 2011. "E-textbook Sales off to a Good
Start." Inside Iowa State. Retrieved Jan. 7, 2012
(http://www.iastate.inside.edu/2011/ 1017/text.php.)
Larson, Randol. 2002. "E-enabled textbooks: Lower cost, higher
functionality." Campus Technology, 05. Retrieved April 9, 2009
(http://campus technology.com/articles/39043/.)
Marklein, Mary Beth. 2012. "Groups target textbook prices to
rein in college costs." USA TODAY, February 2 2012. Retrieved April
12, 2012 (http://www.usatoday.com.)
Miller, Clarie Cain. 2011. "In Price War, New Kindle Sells for
$139." The New York Times, July 28, pp. B1.
National Association of College Stores. 2010. "E-Books,
E-Readers slow to catch on with college students." Press Release,
October 27. Retrieved Feb. 20, 2011 (http://www.nacs.org/
advocacynewsmedia/pressreleases/ebooksereaders
slowtocatchonwithstudents.aspx.)
Nelson, Mark R. 2008a. "E-books in higher education: Nearing
the end of the era of hype?" EduCause Review, 43(2). Retrieved
April 12, 2012 (http://www.educause.edu.)
Nelson, Mark R. 2008b. "Is higher education ready to switch to
digital course materials?" Chronicle of Higher Education 55,
November 28: pp. A29. Retrieved April 12, 2012 (http://www
.chronicle.com.)
Office, U.S. Government Accountability. 2005. College Textbooks:
Enhanced Offerings Appear to Drive Recent Price Increases. Report to
Congressional Requesters, July.
Perez-Pena, Richard. 2009. "Rocky Mountain News Fails to Find
Buyer and Will Close." The New York Times. February 26. Retrieved
Feb. 15, 2012 (http://www.nytimes.com/2009/02/
27/business/media/27paper.html?ref=us&_r=0).
Rampell, Catherine. 2008. "Free textbooks: An online company
tries a controversial publishing model." Chronicle of Higher
Education, May 2, pp. 14-15.
Ravid, Gilad, Yoram M. Kalman, and Sheizaf Rafaeli. 2008.
"Wikibooks in Higher Education: Empowerment through Online
Distributed Collaboration." Computers in Human Behavior, 24,
1913-1928.
Richman, Dan and Andrea James. 2009. "Seattle P-I to publish
last edition Tuesday." The Seattle Post-Intelligencer, March 16.
Retrieved Nov. 11, 2011 (http://www.lexisnexis.com/.)
Roberts, Shearon. 2006. "Costly textbooks draw scrutiny of
lawmakers." The Wall Street Journal, April 25. p. 2. Retrieved
April 7, 2011 (http://www.lexisnexis.com/.)
Roso, Larissa and Michael S. Rosenwald. 2011. "At
Borders' closing, everything on shelves is priced to go. The
shelves, too. "The Washington Post, April 17. Retrieved Nov. 11,
2011 (http:// www.washingtonpost.com/local/at-borders-closingeverything-on-shelves-is-priced-to-go-the-shelves-too/
2011/04/17/AFvuRAwD_story.html).
Savitz, Eric. 2011. "Who Buys Applie iPads? Rich, Male,
Pet-Owning Geeks." Forbes. November 17. Retrieved June 15, 2012
(http://www.forbes.com/
sites/ericsavitz/2011/11/17/who-buys-apple-ipadsrich-male-pet-owning-geeks/.)
Schuetze, Christopher F. 2011. "Textbooks finally take big
leap to digital." The New York Times, November 23. Retrieved April
12, 2012 (http:// www.nytimes.com/2011/11/24/world/americas.)
Stewart, Robert. 2009. "Some thoughts on free textbooks."
EduCause Review, 44(1), 24-26.
US News and World Report Best Colleges 2012 Edition (2012).
Retrieved April 18, 2012 (http://colleges.usnews.rankingsandreviews.com/
best-colleges.)
Stone, Brad and Clarie Cain Miller. 2012. "Doing the iPad
Math: Utility + Price + Desire." The New York Times, April 1, pp.
B1.
Young, Jeffrey R. 2009. "6 Lessons One Campus Learned about
E-Textbooks." The Chronicle of Higher Education. June 4. Retrieved
October 31 2012 (http://chronicle.com/article/6-Lessons-OneCampus-Learne/44440).
Young, Jeffrey R. 2010. "Format War Heats Up Among Publishers
of Electronic Textbooks." The Chronicle of Higher Education.
February 22. Retrieved October 31 2012 (http://chronicle.com/
article/Format-War-Heats-Up-Among/64323/).
Notes
(1.) This analysis was performed on the website
classifications.carnegiefoundation.org.
(2.) Both undergraduate and graduate students received the
invitation. Data for this research are part of a larger study examining
the psychological underpinnings of e-textbook use.
(3.) Students who had a high-school graduating class size of 1 were
home-schooled.
(4.) Appendix Table 1 shows differences between sample means and
population means for select variables. The authors thank the University
of Idaho office of Institutional Research and Assessment for providing
population means.
(5.) Marginal effects are taken at estimations' sample means.
(6.) Preliminary estimations included an interaction of OWN_DESKTOP
and OWN_LAPTOP. The coefficient on the interaction term was
insignificant.
(7.) Larger high schools are more frequently found in cities and
suburbs. It is possible that students who attended such schools might be
more exposed to computers, even in non-scholastic settings, than
students who attended rural high schools.
(8.) We also performed estimations that included a measure of each
student's personal price elasticity, which was created via a
one-factor loading from four different survey questions asking about
students' willingness to search for low prices on textbooks. The
measure of personal price elasticity was never significant in any
estimation.
by Jon R. Miller *, Andrew W. Nutting **, and Lori Baker-Eveleth
***
* Professor of Economics, College of Business and Economics,
University of Idaho, Moscow, ID 83844-3161. Email: jrmecon@uidaho.edu
** Associate Professor of Economics, College of Business and
Economics, University of Moscow, ID 83844-3161. Email:
anutting@uidaho.edu Corresponding author.
*** Associate Professor of Information Systems, College of Business
and Economics, University of Idaho, Moscow, ID 83844-3161. Email:
leveleth@uidaho.edu
TABLE 1.
Variable Descriptions and Summary Statistics (n = 1,098)
Dependent Variable Variable Definition Mean Std. Dev
ETEXT = 1 if ever purchased 0.45 0.50
or used electronic
textbook;
= 0 otherwise
Explanatory Variables
Demographics
AGE = age in years 22.8 4.9
(if 40 or under)
AGE OVER_40 = 1 if over 40 0.07 0.25
years old;
= 0 otherwise
FEMALE = 1 if female; 0.58 0.49
= 0 otherwise
HIGH_SCHOOL_CLASS-SIZE = size of high 266.7 196.6
school graduating
class
HIGH_SCHOOL_PRIVATE = 1 if high school 0.07 0.26
was private;
= 0 otherwise
Financial Status
SHARE_OWN_RESPONSIBILITY = share (between 0.66 0.37
0 and 1) of
college expenses
paid by student
FINANCIAL_AID = 1 if received 0.65 0.48
financial aid;
= 0 otherwise
SCHOLARSHIPS = 1 if received 0.68 0.47
scholarship(s);
= 0 otherwise
PARENTS-FAMILY = 1 if received 0.46 0.50
financial assistance
from parents
or family;
= 0 otherwise
WORK_FULL OR_PART_TIME = 1 if worked 0.56 0.50
full-time or
part-time;
= 0 otherwise
Computer Experience
OWN_LAPTOP = 1 if owned laptop 0.93 0.26
computer;
= 0 otherwise
OWN_DESKTOP = 1 if owned desktop 0.38 0.48
computer;
= 0 otherwise
PROFS STRONGLY_DISAGREE = 1 if professors 0.08 0.28
"strongly disagreed"
with electronic
textbooks;
= 0 otherwise
PROFS_AGREE_OR = 1 if professors 0.06 0.24
STRONGLY_AGREE "agreed" or
"strongly agreed"
with electronic
textbooks;
= 0 otherwise
College of Enrollment
AGRICULTURE_LIFESCIENCES = 1 if enrolled in 0.10 0.30
College of
Agriculture &
Life Sciences;
= 0 otherwise
ART_ARCHITECTURE = 1 if enrolled in 0.05 0.22
College of Art &
Architecture;
= 0 otherwise
BUSINESS-ECONOMICS = 1 if enrolled in 0.13 0.33
College of Business
& Economics;
= 0 otherwise
EDUCATION = 1 if enrolled in 0.13 0.34
College of
Education;
= 0 otherwise
ENGINEERING = 1 if enrolled in 0.16 0.37
College of
Engineering;
= 0 otherwise
LETTERS_ARTS_SOCIAL_SCIENCE = 1 if enrolled in 0.27 0.44
College of Letters,
Arts, & Social
Science;
= 0 otherwise
NATURAL _RESOURCES = 1 if enrolled in 0.07 0.25
College of Natural
Resources;
= 0 otherwise
SCIENCE = 1 if enrolled in 0.10 0.30
College of Science;
= 0 otherwise
TABLE 2.
ETEXT Usage By College
College Mean
Agriculture & Life Sciences (n = 107) 0.45
Art & Architecture (n = 53) 0.40
Business & Economics (n = 137) 0.80
Education (n = 146) 0.32
Engineering (n = 173) 0.46
Letters, Arts & Social Science (n = 293) 0.37
Natural Resources (n = 75) 0.39
Science (n = 106) 0.42
TABLE 3.
Full Sample Results: Marginal Effects from Logit Estimations
Variable 1 2
AGE -0.009 -0.006
-2.20 ** -1.47
AGE_OVER_40 -0.311 -0.191
-2.52 ** -1.47
FEMALE -0.003 0.031
-0.09 0.87
HIGH_SCHOOL_CLASS_SIZE (100s) 0.020 0.021
2.43 ** 2.48 **
HIGH_SCHOOL_PRIVATE 0.077 0.096
1.29 1.56
SHARE_OWN_RESPONSIBILITY -0.035 0.002
-0.68 0.03
FINANCIAL_AID 0.081 0.108
2.34 ** 2.93 ***
SCHOLARSHIPS 0.064 0.056
1.76 * 1.46
PARENTS_FAMILY -0.004 -0.003
-0.11 -0.07
WORK_FULL_OR_PART_TIME 0.034 0.034
1.04 0.98
OWN_LAPTOP 0.135 0.091
2.12 ** 1.39
OWN_DESKTOP 0.081 0.060
2.24 ** 1.57
AGRICULTURE_LIFE_SCIENCES 0.099
1.72 *
ART_ARCHITECTURE 0.038
0.49
BUSINESS_ECONOMICS 0.510
7.93 ***
EDUCATION -0.009
-0.17
ENGINEERING 0.123
2.36 **
NATURAL_RESOURCES 0.059
0.87
SCIENCE 0.097
1.64 *
PROFS_STRONGLY_DISAGREE
PROFS_AGREE_STRONGLY_AGREE
Observations 1,092 1,084
Pseudo-[R.sup.2] 0.024 0.083
Variable 3
AGE -0.006
-1.45
AGE_OVER_40 -0.207
-1.55
FEMALE 0.035
0.98
HIGH_SCHOOL_CLASS_SIZE (100s) 0.020
2.32 **
HIGH_SCHOOL_PRIVATE 0.084
1.33
SHARE_OWN_RESPONSIBILITY 0.006
0.10
FINANCIAL_AID 0.109
2.91 ***
SCHOLARSHIPS 0.047
1.20
PARENTS_FAMILY -0.001
-0.03
WORK_FULL_OR_PART_TIME 0.046
1.32
OWN_LAPTOP 0.101
1.51
OWN_DESKTOP 0.067
1.74 *
AGRICULTURE_LIFE_SCIENCES 0.101
1.73 *
ART_ARCHITECTURE 0.066
0.82
BUSINESS_ECONOMICS 0.485
7.44 ***
EDUCATION -0.013
-0.24
ENGINEERING 0.131
2.46 **
NATURAL_RESOURCES 0.057
0.82
SCIENCE 0.105
1.75
PROFS_STRONGLY_DISAGREE -0.154
-2.47 **
PROFS_AGREE_STRONGLY_AGREE 0.240
3.06 *
Observations 1,070
Pseudo-[R.sup.2] 0.095
Absolute value of z statistics included
* significant at 10%; ** significant at 5%; *** significant at 1%
TABLE 4.
Subsample Results: Marginal Effects from Logit Estimations
NON-CBE
Variable 1 2
AGE -0.007 -0.007
-1.57 -1.51
AGE_OVER_4O -0.220 -0.227
-1.68 * -1.70 *
FEMALE 0.027 0.033
0.74 0.89
HIGH_SCHOOL_CLASS_SIZE (100s) 0.023 0.021
2.63 *** 2.44 **
HIGH_SCHOOL_PRIVATE 0.092 0.078
1.49 1.25
EXPENSES_SHARE_YOUR_RESP 0.015 0.020
0.26 0.36
FINANCIAL_AID 0.111 0.112
2.98 *** 2.98 ***
SCHOLARSHIPS 0.035 0.027
0.91 0.69
PARENTS_FAMILY -0.002 0.000
-0.05 0.00
WORK_FULL_OR_PART_TIME 0.019 0.031
0.53 0.88
OWN_LAPTOP 0.078 0.087
1.24 1.35
OWN_DESKTOP 0.067 0.073
1.75 * 1.87 *
AGRICULTURE_LIFE_SCIENCES 0.096 0.098
1.73 * 1.75 *
ART-ARCHITECTURE 0.034 0.061
0.45 0.79
BUSINESS_ECONOMICS
EDUCATION -0.008 -0.012
-0.15 -0.22
ENGINEERING 0.116 0.124
2.29 ** 2.41 **
NATURAL_RESOURCES 0.057 0.055
0.87 0.83
SCIENCE 0.092 0.101
1.63 1.74 *
PROFS_STRONGLY_DISAGREE -0.150
-2.44 **
PROFS_AGREE_STRONGLY_AGREE 0.230
2.78 ***
Observations 948 935
Pseudo-[R.sup.2] 0.029 0.042
UNDER AGE 30
Variable 3 4
AGE -0.010 -0.007
-1.39 -1.03
AGE_OVER_4O
FEMALE 0.057 0.068
1.46 1.69 *
HIGH_SCHOOL_CLASS_SIZE (100s) 0.020 0.020
2.09 ** 2.04 **
HIGH_SCHOOL_PRIVATE 0.136 0.121
2.02 ** 1.77 *
EXPENSES_SHARE_YOUR_RESP 0.021 0.029
0.36 0.49
FINANCIAL_AID 0.100 0.095
2.44 ** 2.27 **
SCHOLARSHIPS 0.072 0.072
1.69 * 1.66 *
PARENTS_FAMILY 0.007 0.011
0.17 0.26
WORK_FULL_OR_PART_TIME 0.037 0.047
0.98 1.24
OWN_LAPTOP 0.019 0.020
0.25 0.26
OWN_DESKTOP 0.066 0.074
1.55 1.71 *
AGRICULTURE_LIFE_SCIENCES 0.111 0.112
1.79 * 1.79 *
ART-ARCHITECTURE 0.051 0.077
0.62 0.93
BUSINESS_ECONOMICS 0.522 0.501
7.74 *** 7.32 ***
EDUCATION -0.038 -0.042
-0.59 -0.63
ENGINEERING 0.134 0.145
2.35 ** 2.50 **
NATURAL_RESOURCES 0.061 0.054
0.81 0.70
SCIENCE 0.121 0.124
1.87 * 1.89 *
PROFS_STRONGLY_DISAGREE -0.120
-1.84 *
PROFS_AGREE_STRONGLY_AGREE 0.275
3.26 ***
Observations 912 901
Pseudo-[R.sup.2] 0.085 0.098
Absolute value of z statistics included
* significant at 10%; ** significant at 5%; *** significant at 1%
