Content uploaded by Richard Duran
Author content
All content in this area was uploaded by Richard Duran
Content may be subject to copyright.
31
© 2007 The University of North Carolina Press
Redefining the Digital Divide:
Beyond Access to Computers
and the Internet
James R. Valadez
California Lutheran University
Richard Duran
University of California at Santa Barbara, CA
This study critiqued the notion that a binary
“digital divide” between high and low resource
schools describes accurately the technology
disparity in U.S society. In this study, we sur-
veyed teachers from six southern California
schools. Five of the schools were low resource
schools and one school, chosen for compara-
tive purposes, was characterized as a high
resource school. We found that high resource
school teachers had significantly more physi-
cal access to computers and the Internet (C&I),
more frequent use of C&I, more creative uses of
C&I for instruction, communicated by email
more often with students, and engaged more
frequently in professional activities such on
on-line communication with other teachers.
The study lent modest support to previous
researchers (Natriello, 2001; Warschauer,
2003a, 2003b, 2003c; Wenglinksy, 1998) who
claimed that high resource students are more
likely to use C&I for more experimental and
creative uses than students from low resource
schools. In addition the findings contribute to
a broader definition of the “digital divide” that
includes social consequences including the
impact of social networks and wider use of
technology to improve instruction.
While the total number of U.S. residents pur-
chasing computers and connecting to the
Internet increases daily, large segments of the
population are being passed over in the
Information Age. In a recent study NTIA (2002)
showed that Whites and Asian Americans have
higher rates of both computer and Internet use
than Blacks and Latinos. NTIA found computer
use to be highest for Asian Americans (71.2 per-
cent) and Whites (70.0 percent), followed by
Blacks, (55.7) and Latinos (48.8). Regarding
Internet use, 60% of Whites and Asian
Americans use the Internet compared to Blacks
(39.8 percent) and Latinos (31.6 percent) who
use the Internet at much lower rates (NTIA,
2002).
U.S. schools mirror the nation’s trends in com-
puter ownership and Internet connectivity. By
2002 over 99% of U.S. schools owned comput-
ers and had Internet connections (NCES, 2004).
32
income individuals exhibited gains, Internet
use was growing more slowly for low-income
individuals compared to higher income groups.
Martin concluded that in order to compensate
for the gap, the lag in technology diffusion
between rich and poor should be as brief as pos-
sible to ameliorate the current inequality in C&I
use.
Others, meanwhile, have argued that declaring
the “digital divide” closed based on a wider
availability of computers oversimplifies the
construct. For example, Bruce (1999) stated that
the “digital divide” is not so much that certain
groups have less access to the Internet; it is that
they have a different kind of access. In other
words students from low-income backgrounds
often find their access is restricted to computer
labs where they are limited to instructional soft-
ware that emphasizes low-level drill and prac-
tice routines. Healy (1998) divided the techno-
logical world into the “interacting” and the
“interacted.” The “interacting” includes those
who can take advantage of sophisticated appli-
cations and research possibilities available on
the Internet, while the “interacted” are those
individuals who must settle for the most sim-
plistic offerings designed for lower level users.
In what follows we develop a framework for
redefining the concept of the “digital divide.” In
particular we propose how a re-envisioning of
the “digital divide” concept addresses more
accurately how high and low poverty schools
differ in terms of access, use, and support of the
integration of information technology in teach-
ers’ classrooms.
Framing the Study
Popular acceptance and reaction to the concept
“digital divide” led to well-meaning but incom-
plete attempts to solve the technology gap
between rich and poor students by simply
installing computers and providing access to
the Internet in low resource schools (Attewell,
2001; Duran, 2001; Natriello, 2001; Warschauer,
2003a; 2003b; 2003c). California’s Digital High
School Grants Program (DHSGP) was an exam-
ple of this effort. DHSGP succeeded in increas-
ing access to computers in resource-poor
schools, but California’s schools still have not
achieved equity in terms of C&I use. Schiff and
Solmon (1999) reported that DHSGP offered
Despite these findings, there is still evidence of
an economic and racial divide among school
children and their use of computers and the
Internet (C&I). To illustrate, the ratio of students
to computers in high poverty schools1is much
higher than it is in more affluent schools. NCES
(2003) estimated that in high poverty schools,
the student to computer ratio was 5.5 students
per instructional computer compared to 4.6 stu-
dents to a computer in more affluent schools. In
addition, use of the Internet by children from
the varying social strata differs markedly.
Overall Internet use at home and school for
Latino and Black children is 47.8% and 52.3%
respectively, compared to Asian American
(79.4%) and White children (79.7%) who are
far more likely to use the Internet (NTIA, 2002).
Becker and his colleagues (1999) pointed out
that teachers are more likely to assign C&I work
when their students have ready access to com-
puters. They showed that teachers with ratios of
four or fewer students to a computer were three
times more likely to assign computer work to
students than those teachers with less favorable
ratios of six or more students to a computer.
Given that higher socioeconomic status (SES)
schools are more likely to have low student to
computer ratios, they provide a distinct advan-
tage over low SES schools in gaining the expe-
rience and practice necessary for using the
Internet as an educational resource.
It is evident that computer availability and
Internet access have improved over the last sev-
eral years, however the studies cited above con-
cur that the poor and racial minorities are lag-
ging behind society’s dominant groups in terms
of computer ownership and Internet connectiv-
ity. Rapid advancement in computer availabili-
ty however, has spurred some commentators to
proclaim the closing of the “digital divide”
(Simons, 2000; Thierer, 2002). Compaigne
(2000) argued against any legislation to ensure
universal access because present market forces
and government programs currently in place
are already achieving that goal. While affirming
a digital gap between rich and poor, NTIA’s
report (2002) indicated a remarkable level of
“catching up” among individuals with low fam-
ily incomes. Martin (2003), in his re-analysis of
the NTIA (2002) data, found that while low-
The High School Journal – Feb/ Mar 2007
33
schools the opportunity to provide students
with computer hardware, but did not provide
feasible plans to support teachers in integrating
technology into their curricula.
In Warschauer’s (2003a) view, initiatives such
as DHSGP were aimed at closing the computer
gap, but failed to address differences in how
technology was used in high and low SES
schools. Cuban, Kirkpatrick, & Peck’s (2001) cri-
tique of the technology movement sought to
address some of its shortcomings by character-
izing schools as too eager to “jump on the tech-
nology bandwagon.” Cuban et al. (2001) note
how schools make use of special funding to
purchase equipment, but fail to develop coher-
ent plans for implementation, support, and pro-
fessional development of teachers to integrate
technology into the classroom.
Natriello’s (2001) commentary centered on how
schools joined in the rush to bridge the divide
by connecting to the Internet, but the effort my
have inadvertently contributed to patterns that
have exacerbated disparities between rich and
poor. Natriello argued that schools shaped stu-
dents’ computer activities by purchasing soft-
ware that supported the school’s ongoing peda-
gogical practices. In other words low SES
schools acquired software that promoted drill
and practice routines while more affluent
schools purchased software for more creative
and experimental uses. This is in line with
Warschauer’s (2003b) warning that computer
use in schools could actually worsen inequality
by failing to address the key issue regarding the
way computers are used rather than merely pro-
viding more physical access to technology.
Cuban’s (1986) critique was similar. Cuban con-
cluded that teachers take from the computing
world what they find immediately useful and
jettison the rest, often relying on low-level drill
and practice strategies that fit easily within their
existing pedagogical approaches. Wenglinsky’s
(1998) extensive analysis of the National
Assessment of Educational Progress data sup-
ports to some extent, Natriello’s (2001),
Warschauer’s (2003b) and Cuban’s (1986) asser-
tions. Wenglinsky (1998) found that African
American students were more likely than White
or Asian students to use computers for lower-
order activities and were more likely to be
taught by teachers who were unprepared to use
computers in their classrooms. Other findings,
ironically, have shown that economically disad-
vantaged students are more apt than White or
Asian students to use the computer daily, but
are also more likely to use it for drill and prac-
tice routines (Coley, et al., 1997; NCES, 2003a;
2003b).
Redefining the Digital Divide
This study critiques the notion that a binary
“digital divide” between the “haves” and “have
nots” describes accurately the technological
inequality that exists between high and low-
resource schools. The weakness of the “digital
divide” framework lies in its overemphasis on
the importance of the physical presence of C&I
connectivity to the exclusion of other factors
that allow students to use computers for mean-
ingful ends (Kling, 1998). DiMaggio et al. (2001)
argue that a new view of the “digital divide”
must include an attempt to redefine access in
social as well as technical terms. Their view
was that a social definition encompasses not
only the pressing question of who can find
access, but also includes what teachers are able
to do when they go on-line, including on-line
communication and their ability to locate
appropriate instructional materials.
Martin (2003) suggested that to overcome the
binary definition of the “digital divide” it must
be replaced with a multidimensional view of
access to C&I. Martin’s (2003) model partitions
access into three dimensions, including motiva-
tion, possession, and skills. Motivation refers to
the willingness of individuals to use technology
and to include it in their home, work, and edu-
cational efforts. Possession describes a more
concrete definition of access including physical
access to C&I and the ability to use the technol-
ogy. Skills refer to the ability to use the technol-
ogy, and the degree of support available to
instruct individuals in its use. DiMaggio et al.’s
(2001) contribution to a new view of the
“divide” emphasizes social aspects including
the development of peer networks and the
social support necessary to institute innovative
uses of the technology. Warschauer’s (2003b)
critique of the concept “digital divide” calls for
a deepening of the concept to include the social
consequences of the “divide.” Warschauer
Redefining the Digital Divide
34
level of local and administrative commitment
and support to integrate computers and tech-
nology into classroom activities.
We include social consequences as part of our
re-definition of “access” to measure the degree
to which teachers are developing and improv-
ing their skills as professional educators. Social
consequences implies that teachers use C&I to
communicate and collaborate with colleagues,
develop social networks, and that they use tech-
nology to improve their communication with
students. A social consequence of computer
access would be the level of formal and infor-
mal networks that form to support teacher inte-
gration of technology. Our inclusion of this con-
cept follows the work of March (2001) and
Warschauer (2003a; 2003b; 2003c) who stressed
the importance of teachers being able to use
technology to actively participate in communi-
cation exchange to improve their practice. We
also define the concept of social consequences
to include teachers’ perceptions of how com-
puters and the Internet engage students in high-
er order learning (i.e., increased motivation and
engagement, learning in new ways, deeper
understanding).
In summary we agree that the “digital divide”
correctly characterizes a technology gap
between rich and poor, however the term is
much too simplistic to encompass the vast dif-
ferences in opportunity, experiences, and prac-
tices that exist between high and low SES stu-
dents. A more accurately defined “digital
divide” does not simply describe the division
between technology “haves” and “have nots”
but addresses inequalities in technology and
learning. A broader framework that includes a
deeper understanding of the “digital divide”
involves questioning local, state and federal
policies that attempted to solve complex socie-
tal issues with simplistic solutions. Providing
resources for schools to purchase computers
did not address the more important issues
regarding poverty, inequality, and differential
opportunities made available to low and high
SES students.
Our framework complicates the concept “digi-
tal divide” by pursuing questions that address
the importance of the social aspects of the
divide. For instance, we not only ask questions
(2003a; 2003b) stressed that C&I access signifies
social practices leading to the use of technology
for building collaboration and cooperation
among users. This reconceptualization of the
digital divide urges consideration for how peo-
ple use the Internet for furthering the process of
social inclusion including wider participation
in democratic societies (Warschauer, 2003a;
2003b).
For this study we redefine the notion of the
“digital divide” based on the work of Martin
(2003), DiMaggio et al. (2001), and Warschauer
(2003a; 2003b; 2003c) to understand more fully
C&I access in educational settings. Our re-con-
ceptualization of the “divide” is based on four
elements re-defining access to C&I: (1) the phys-
ical access (DiMaggio et al., 2001; Martin,
2003); (2) C&I use in the classroom (Martin,
2003); (3) availability of support for C&I use
(Martin, 2003); and (4) social consequences of
the use of IT (March, 2001; Warschauer, 2003).
Having a computer and being connected to the
Internet define the central characteristics of
Physical Access. Physical Access to C&I is the
critical factor for encouraging teachers to inte-
grate C&I in their classroom instruction.
Physical Access however, is not a binary con-
cept. Physical access can be described as a spec-
trum in which certain teachers may have com-
puters connected to the Internet in their own
classrooms, while other may have less conven-
ient connections or even more remote access in
libraries or computer labs.
C&I use expands our definition of “access” by
including the amount of time teachers spend
using C&I for instructional purposes in school
and at home. It includes frequency of email use,
creating instructional material, and keeping stu-
dent records. C&I use also implies the frequen-
cy teachers use higher order instructional strate-
gies in their classrooms.
Availability of support is defined as the degree
of support for teachers to acquire the skills
needed to integrate C&I in their classroom
teaching. Support may take the form of training
made available through state or district funds,
or local initiatives to support teacher integra-
tion of technology in the classroom. Support is
also characterized by teacher perceptions of the
The High School Journal – Feb/ Mar 2007
35
concerning teacher access and use; teachers’
instructional practices; and administrative sup-
port, but we also seek to understand the social
consequences that emerge when teachers have
access to computers and the Internet.
Specifically our questions are: (1) how does
access to C&I differ between high and low
resource schools? (2) How does access influ-
ence the work that teachers do, including how
they teach and what they teach; and (3) What
type of support, funding, and administrative
support exist in high and low-resource schools
and (4) What are the social consequences,
including development of social networks, that
develop from the use of C&I ?
Method
Procedures
In 1999 State U. formed a partnership with five
area high schools with purpose of improving
the schools’ learning opportunities for their pre-
dominantly Latino student bodies. The partner-
ship centered on providing teacher professional
development, curricular reform, and technolo-
gy initiatives. Technology initiatives, supported
by the above referenced Digital High School
Grants Program and State U. expertise, empha-
sized professional development for helping
teachers to adopt the use of C&I in their class-
rooms. Our purpose in approaching these
schools for this study was to investigate the
impact of the technology initiatives.
The schools. Each of the five partner high
schools, which we designated low-resource
schools, shared similar characteristics2. All
were located within 80 miles of the State U. and
were classified by the California Department of
Education (CDE) as “low performing schools.”
This designation was based on the schools’ low
API3rankings that ranged from 2 to 4. Aside
from low API rankings the partner schools
exhibited other symptoms of troubled schools
including extremely low college-going rates,
disciplinary problems, high truancy, and low
teacher morale (Casas & Fenstermacher, 2000).
As a result of their status as “low performing
schools,” the partner schools were under a
directive to take significant steps to improve
their academic standing to avoid sanctions,
including possible takeover from CDE.
We selected an additional school (S6) to pro-
vide a contrasting reference point for the five
partner high schools. S6, a high resource
school, was also located within 80 miles of
State U. S6 received an API score of 10 (the
highest possible) and was located in an upper
middle class (median family income = $77,000)
southern California neighborhood. In contrast,
the five low resource schools were located in
low-income southern California neighborhoods
with median family incomes ranging from
$37,000 to $39,000 (U.S. Census Bureau, n.d.).
Questionnaire development. We began our data
collection by interviewing administrators (prin-
cipals and vice principals) and one teacher from
each of the six schools. A simple protocol was
used (how do teachers us computers in your
school? How do students use computers? What
type of support is available for the use of com-
puters in the classroom?). These interviews pro-
vided us with basic phenomenological data
regarding C&I use, and informed the develop-
ment of a 52 item questionnaire to assess teach-
ers’ access and use of computers; perceptions of
how students use C&I; and level of support for
use of C&I and the degree to which teacher
approaches to education are being transformed.
In addition we reviewed computer access ques-
tionnaire items being considered for the
Educational Longitudinal Survey of 2002.
Those items were based on previous national
surveys, including the National Assessment of
Educational Progress (2000); National
Educational Longitudinal Study of 1988 and
the Fast Response Staff Survey (NCES, 2001).
Specifically we queried teachers about their (1)
access to C&I; (2) frequency of use of computers
in school and at home; (3) degree of training and
professional development to support classroom
use of C&I; and (4) how computers influence
teacher conceptions of their role in the class-
room (i.e. use of technology to communicate
with colleagues and students outside of the
classroom)
Sample. We distributed the questionnaire to the
teachers at the six schools. Our sample includ-
ed teachers who taught courses in the State U.’s
“a-g categories” that determine eligibility for
admission to the university. Students must take
courses from each of the six categories: (a) his-
Redefining the Digital Divide
36
posite variables, along with corresponding fac-
tor loadings are included in the appendix. We
also included estimates of reliability
(Cronbach’s) for the researcher created compos-
ites. In addition to the above composites we
selected two variables as proxies for profession-
al activities (use C&I for on-line conversations
with colleagues) and communication (use C&I
for email communication with students).
We used these factors to describe the elements
of our definition for a newly conceptualized
view of the “digital divide.” Physical access is
defined by one factor, physical access to C&I.
This factor described the number of computers
available for teacher use, connections to the
Internet, and access to local area networks. C&I
use was defined by three factors: (1) school use
of C&Is; (2) home use of C&I; and (3) instruc-
tional practices. Instructional practices includ-
ed the frequency in which teachers engaged in
higher order instructional strategies, including
problem solving, data analysis and word pro-
cessing. Social Consequences includes three
measures: (1) professional activities is a variable
that measures the frequency teachers engaged
in professional on-line communication with
colleagues regarding instructional issues, (2)
communication with students measures the
amount of email communication teachers have
with students about homework, and the fre-
quency that they post assignments on the
Internet, and (3) the degree of student engage-
ment in their learning.
Availability of Support is defined by two fac-
tors: (1) teachers’ perceptions of the amount of
training they have received to use C&I, and (2)
their views on the existence of administrative
and economic barriers to integration of C&I into
their classrooms.
Results
Physical Access
Our findings show (Table 1) that teachers from
the six schools reported differing degrees of
physical access to C&I, F_ (5, 178.6) = 9.08,
p < .01. Effect sizes, “eta squared” (_2), describe
the proportion of variability in the dependent
measure that is attributable to a factor. For
example “eta squared” for access shows that
14% of the total variance in the dependent
tory/social sciences (two years); (b) English
(four years); (c) mathematics (minimum of three
years); (d) laboratory sciences (two years); (e)
foreign language (two years); (f) visual and per-
forming arts (one year); and (g) additional col-
lege preparatory electives (one year). We
excluded teachers from non a-g subject areas
such as physical education, vocational educa-
tion, or special education. Our rationale for
focusing on a-g subject areas was to survey
teachers who taught the required courses for
entry into State University. We sent 398 ques-
tionnaires to the teachers who met these criteria
at the six schools. We received 285 usable ques-
tionnaires for a return rate of 72%.
Analytic Plan
We obtained multiple measures to construct
several dimensions that described how C&I was
used at the six schools. We used principal com-
ponents factor analysis to develop the dimen-
sions and an analysis of variance (ANOVA) to
distinguish the six schools along the prescribed
factors. Because of the unequal sample sizes in
this study, we followed Cohen’s (2000) sugges-
tion for using the Brown-Forsythe formula to
adjust the F statistic (F_). The Brown-Forsythe
formula takes into account the unbalanced
ANOVA design and heterogeneity of the sample
variances. Additionally, the Brown-Forsythe
procedure is appropriate because the samples
exhibited a consistent pattern with larger
groups having smaller variances than the small-
er groups. We followed this analysis with
Games-Howell (GH) post hoc comparisons to
determine how the schools differed from each
other along each dimension. Games & Howell
(1976) recommended this approach when the
study includes unequal sample sizes and vari-
ances.4
Measures
We used principal components factor analysis
with a promax rotation to analyze C&I use by
teachers for the entire sample and extracted the
weighted factor scores5. This analysis yielded
six distinct factors: physical access; barriers to
computer use; instructional practices; student
engagement; C&I school use; and training. In
addition to these factors there was one
researcher created composite: school use of C&I.
Information on the construction of these com-
The High School Journal – Feb/ Mar 2007
37
measure can be explained by the independent
variable (school).
Table 2 displays results from the post hoc analy-
sis showing how teachers from the six schools
reported their degree of physical access to C&I.
S6 teachers indicated they had more computers
available for instruction, more computers con-
nected to the Internet, newer computers, and
more computers connected to local area net-
works (LANs) than any of the five low-resource
schools. Further analysis shows that there were
no significant differences in physical access
among the low-resource schools.
C&I Use
Teachers reported significant differences, F_ (5,
194.9) = 6.12, p < .01, in the amount of C&I use
in school (Table 1). This highly significant dif-
ference indicates a disparity among the schools
in the degree that teachers used computers for
keeping administrative records, creating
instructional materials, and general use of com-
puters in their classroom. Results of the post
hoc analysis (Table 3) illustrate that significant
mean differences can be detected between S6
and three low-resource schools (S1, S3, S5).
From this finding we inferred that S6 teachers
used their in-school computers to support
instruction more frequently than S1, S3, and S5
teachers. No other differences using the GH
approach were detected between S6 and the
other schools, nor were there differences among
the five low-resource schools. The findings sup-
port a relationship between frequent use of in-
school computers and the availability of com-
puters in teachers’ classrooms. As Becker and
his colleagues (1999) reported, when teachers
have access to computers and Internet connec-
tions, it facilitates their use and their propensi-
ty for assigning student work on computers.
Teachers reported significantly different C&I
home use to support their work, F_ (5, 165.7) =
3.16, p < .01 (Table 1). This variable, describing
frequency that teachers use computer at home,
and use email at home to support instruction,
produced a somewhat unexpected finding
(Table 4). S3 teachers showed more home use of
computers than either S4 or S5 teachers, also
from low resource schools. A closer examina-
tion of these schools showed that S3 (M = 1.70,
s = 1.17) had far fewer computers per classroom
connected to the Internet than either S4 (M =
2.29, s = .98) or S5 teachers (M = 2.64, s = 1.21).
A possible explanation is that because S3 teach-
ers had fewer in-school computers connected to
the Internet than any other school in the study,
they relied on their home computers to support
their classroom work.
How often teachers used various instructional
practices (Table 1), including problem solving,
data analysis, and word processing by students
was shown to be significant, F_ (5, 156.8) = 1.9,
p < .10. Table 5 displays our post hoc analysis
for the variable instructional practices. We did
not detect any significant mean differences
between any of the pairs using the GH post hoc
analysis. This led us to perform pariwise com-
parisons for S6 and the five low-resource
schools using the Least Significant Difference
(LSD) approach. LSD is roughly equivalent to
performing all pairwise t-tests, comparing each
pair to a critical value based on p = .05 (except
that the error term is based on data from all con-
ditions rather than being computed for each
pair). This procedure emphasizes sensitivity
minimizing Type II errors, but increases the
possibility of Type I error. The results of the
LSD analysis, therefore, must be interpreted
with caution. Nevertheless, Table 5 shows that
there were significant mean differences
between S6 and the five low resources schools,
adding support to the findings of previous
researchers (Bruce, 1999; Healy, 1998; Natriello,
2001; Warschauer, 2003a; 2003b; 2003c) who
claim that high resource schools use C&I to sup-
port higher order activities, while low-resource
schools use C&I for much simpler activities
such as drill and practice routines.
Social Consequences
Our analysis of Professional activities indicated
clearly significant differences among the
schools F_ (5, 114.1) = 9.08, p < .01 (Table 1).
We selected this variable to describe how teach-
ers used C&I to engage in professional activity
(i.e. how teachers use technology to improve
their teaching). We found that S6 teachers
engaged more frequently in professional type
behavior than teachers from four of the five low-
resource schools (Table 6). The largest signifi-
cant mean differences were between S6 and S1;
S6 and S3; and S6 and S5 (Table 6). A smaller
Redefining the Digital Divide
38
puters in the classroom, it was not surprising to
find relatively small mean differences (not dis-
played here) between high and low resource
school teachers and their views regarding barri-
ers and training.
Conclusions
Our goal was to redefine the notion of the “dig-
ital divide” to provide a more accurate frame-
work for analyzing the technology gap between
high and low resource schools. We derived a
framework from the work of Warschauer
(2003a; 2003b; 2003c), March (2001), Martin
(2003), and DiMaggio, et al. (2001) to introduce
a multidimensional view of the “divide” that
broadens the concept of “access” to include not
only whether teachers have physical access to
C&I , but what they do when they are on-line.
This view of C&I access describes how teachers
use computers to support instruction, and
social consequences of Internet use, including
skill development, communication, and build-
ing social networks.
Evidence from our study re-affirms a “divide”
between high and low resource schools. We
showed that S6, a high resource school, had
more computers per classroom, more connec-
tions to the Internet, and more access to local
area networks than any of the low-resource
schools. Further examination however begins to
unveil a more complex view of the “divide.” S6
with its greater access to C&I than low-resource
schools, had more teachers using C&I to support
instructional activities. In addition to more fre-
quent use, we presented modest findings that
S6 teachers were more likely to engage in C&I
practices that encouraged creative and critical
thinking in their students. These results comply
with Becker and his colleague’s (1999) report
that teachers with more access to C&I used com-
puters more frequently and assigned computer
work to students more often than teachers with
less access. Additionally, we added cautious
support to Natriello (2001) and Wenglinsky’s
(1998) assertions that high resource schools are
more likely to involve students in higher order
learning processes such as problem solving and
data analysis.
Our expanded view of the “divide” addresses
the social consequences attributed to the range
mean difference, yet still significant was detect-
ed between S6 and S4 (Table 6). Interestingly,
we also found that S2 and S4 teachers engaged
more frequently in professional type behavior
than S3 teachers (Table 6). S3 teachers, again
fettered by fewer in-school computers connect-
ed to the Internet, engaged less frequently in
professional type activity.
Our analysis of the variable communication
with students showed modest differences in
teacher use of email for communicating with
students outside of class, F (5, 142.8) = 2.69, p <
.05 (Table 1). Although the results are not sub-
stantial (_2= .059), they are indications that
teachers from high resource schools may be tak-
ing advantage of available technology to inte-
grate innovative approaches to teaching and
learning in their classrooms. Table 7 shows that
S6 teachers were more likely than S1 and S4
teachers to use email with students to discuss
school work than teachers from the two low
resource schools (using the GH approach).
Table 7 also shows mean differences (using the
LSD approach) between S6 and three remaining
low-resource schools (S2, S3, S5) indicating
more frequent use of C&I by S6 teachers for
communicating with students. Finding signifi-
cance using the LSD approach (because of
increased chance of Type I error) tempers our
findings, but our results lend cautious support
to previous research attesting that high resource
schools use innovative teaching strategies more
frequently than low-resource schools (Natriello,
2001; Wenglinsky, 1998).
In a somewhat unexpected finding, teachers
reported no significant differences in their per-
ceptions of increased student engagement due
to use of C&I (Table 1). The lack of significance
regarding student engagement shows more
agreement than disagreement among teachers
and their beliefs concerning how computers
affect student learning.
Available Support
Two factors that we identified as proxies for
available support were found not to be signifi-
cant: Barriers to implementation; and presence
of training to support C&I for classroom instruc-
tion (Table 1). Because of the amount of
statewide funding and local administrative
commitment for implementing the use of com-
The High School Journal – Feb/ Mar 2007
39
of access to C&I. We agree with DiMaggio et al.
(2001), that differential access to C&I found in
schools points to inequality in teachers’ oppor-
tunities to develop knowledge and technical
skills in ways that enhance their professional
practice and social life. To illustrate, S6 teachers
consulted more frequently with on-line col-
leagues regarding instructional issues than their
lower resource counterparts. A consequence of
this communication disparity is that high
resource teachers are more likely than low-
resource teachers to be exposed to opportuni-
ties leading to the skill development and
knowledge acquisition through the construc-
tion of social networks. As Hargittai (2001)
explained, social networks expose teachers to
technological innovation and a more expansive
instructional repertoire than teachers with less
exposure.
Our redefinition of the “divide” accounts for a
more sweeping condition than previously
described in the popular literature. A newly
defined “divide” encompassing the social con-
sequences attributed to C&I use addresses the
vast differences in teachers’ skills, knowledge,
and professional practices characterizing high
and low resource schools. In addition, a new
“divide” explains further how our stratified
educational system provides more opportuni-
ties for innovation, experimentation, and cre-
ativity for society’s more privileged socioeco-
nomic groups.
From a policy perspective it remains a high pri-
ority to close the “hardware gap” between high
and low resource schools. Although as
Warschauer (2003a) reminds us, the divide is
not “binary” and there is no single overriding
factor for determining or closing such a divide.
Overcoming this divide means furthering the
process of social inclusion including policies
directing state and local districts to address the
knowledge, skill, and social gap characterizing
teachers from high and low resource schools.
Low resource school teachers are in need of pro-
fessional development to acquire skills to inte-
grate C&I into their instructional practices. In
addition, structures need to be developed to
assemble teachers into school, district, and
national networks to support C&I use in their
classroom.
Finally there are social justice concerns regard-
ing the “divide” between high and low resource
schools. In addition to plugging the hardware
gap, education policy makers must address
issues related to the impoverished communities
in which the schools are located. It is one thing
to provide in-school computers, but it is also
essential for students to have C&I access at
home, including up-to-date computers, soft-
ware, and high speed Internet connections. It is
only through social policies, grant initiatives,
and programs to provide C&I connections that
students from low-resource schools will
approach the technology standards existing in
more privileged communities.
References
Attewell, P. (2001). The first and second digital divides.
Sociology of Education, 74(3), 252-259.
Becker, H., Ravitz, J. & Wang, Y.T. (1999). Teachers and
teacher directed use of computers and software
(Report #3). Irvine, CA: University of California
Irvine and University of Minnesota, Center for
Research on Information Technology and
Organizations.
Bruce, B. (1999). Speaking the unspeakable about 21st
century technologies. In G. E. Hawisher and C. L.
Selfe (Eds.). Passions, Pedagogies, and 21st Century
Technologies (p. 221-228). Logan, UT and Urbana, IL:
Utah State University Press and NCTE.
Coley, R., Cradler, J., Engel, P.K. (1997). Computers and
classrooms: The status of technology in U.S. schools
policy and information report. Princeton, NJ:
Educational Testing Service.
Compaigne, B. 2001. Declare the war won. In
Compaigne, Benjamin M. (ed.) The digital divide:
Facing a crisis or creating a myth? (p. 315-336).
Cambridge, MA: MIT Press Sourcebooks.
Casas, M. & Fenstermacher, S. (2000). 1999-2000 end-of-
year report on campus outreach initiatives. Santa
Barbara, CA: University of California at Santa Barbara
Chancellor’s Outreach Advisory Board.
Cohen, B.H. (2000). Explaining Psychological Statistics.
2nd edition. New York: John Wiley & Sons.
Coley, R.J., Cradler, J., & Engle, P.K. (1997). Computers
and classrooms: The status of technology. Princeton,
NJ: Educational Testing Service.
Cuban, L. (1986). How teachers taught: The classroom
use of technology since 1920. New York, NY:
Teachers College Press.
Cuban, L., Kirkpatrick, H., & Peck, C.(2001). High access
and low use of technologies in high school class-
rooms: Explaining an apparent paradox. American
Educational Research Journal, 38(4), 813-834.
DiMaggio P., Hargittai, E., Neuman, W.R. & Robinson,
J.P. (2001). Social Implications of the Internet.
Annual Review of Sociology. 27, 307–36
Duran, R.P. (2001). Technology, education, and at risk
students. Santa Barbara, CA: University of California
at Santa Barbara, Gevirtz Graduate School of
Education.
Redefining the Digital Divide
40
Appendix: Measures
Variables from Teacher Technology Survey and
factor scores in parentheses.
Physical access to C&I is the degree to which
teachers have access to computers and the
Internet in the classroom. Variables include:
number of computers available for instruction,
.887; number of computers with access to the
internet, .947; number of computers less than 3
years old, .775; number of computers connected
to local area network, .904. Variables coded 1 =
one; 2 = two; 3 = three; and 4 = four or more.
Barriers are the structural or administrative bar-
riers to instructors using computers and the
Internet in their classrooms. Variable include:
outdated computers, .485; lack of good soft-
ware, .638; inadequate training; .880; lack of
release time, .795; lack of administrative sup-
port (.850); lack of funding (.704). Variables
coded 1 = not a barrier; 2 = small barrier; 3=
moderate barrier; and 4 = great barrier.
Instructional practices describes variables
showing how teachers used the computers and
the Internet in their classroom: use computers
to solve problems (.874); use computers to ana-
lyze data (.670); instruct word processing (.437).
Variables coded 1 = never/very rarely; 2 = 1-2
times per month; 3 = 1-2 times per week; 4 =
almost every day; 5 = every day.
C&I school use shows how teachers use com-
puter and the Internet in their classrooms. The
variables are: how often do you use a computer
in school (.777); how often do you create
instructional materials (.629); how often do you
use the computer for administrative records
(.750). Variables coded 1 = every day; 2 = almost
every day; 3 = a few times a week; 4 = between
once a week and once a month; 5 = never.
Student engagement is the degree to which stu-
dents are engaged in the classroom. Variable
included are: students are more engaged and
motivated when computers are used (.655);
computers allow students to learn in ways not
possible with traditional techniques (.793);
computers help students reach deeper under-
standing (.885). Variables coded 1 = strongly
agree; 2 = agree; 3 = disagree; 4 = strongly dis-
agree.
Games, P. A. & Howell, J. F. (1976). Pairwise multiple
comparison procedures with unequal n’s and/or vari-
ances. Journal of Educational Statistics, 1, 113-125
Hargittai, E. (2003). Digital divide and what to do about
it. In D.Jones (Ed.), New Economy Handbook.
Retrieved November 8, 2005, from
http://www.eszter.com/research/pubs/hargittai-digi-
taldivide.pdf
Healy, J. (1998). Failure to connect: How computers
affect our children’s minds for better and worse. New
York, NY: Simon and Schuster.
Kling, R. (1998). Technological and Social Access to
Computing, Information, and Communication
Technologies. Bloomington, IN: Indiana University,
Center for Social Informatics. Retrieved October 27,
2005 http://www.slis.indiana.edu/kling/pubs/
NGI.htm
March, T. (1999). Ten stages of working the web for edu-
cation. Multimedia Schools. 6(3), 26-30.
Martin, S. (2003). Is the Digital Divide Really Closing?
A Critique of Inequality Measurement in A Nation
Online. IT& Society, 14, 1-13.
National Center for Education Statistics. (2003a).
Internet Access in U.S. Public Schools and
Classrooms: 1994–2002, NCES 2004-01. U.S.
Department of Education: Washington, DC.
National Center for Education Statistics (2003b).
Computer and Internet Use by Children and
Adolescents in 2001, NCES 2004–014, Washington,
DC: U.S. Department of Education.
(NTIA) National Telecommunication and Information
Administration (2002). A nation online: How
American are expanding their use of the Internet.
Washington, DC: U.S. Department of Commerce.
Natriello, G. (2001). Bridging the second digital divide:
What can sociologists of education contribute?
Sociology of Education, 74(3), 260-265.
Schiff, T. & Solmon, L.C. (1999). California Digital High
School: Process evaluation. Santa Monica, CA:
Milken Family Foundation
Shavelson, R. (1996). Statistical reasoning for the
behavioral sciences. 3rd Edition. Boston, MA: Allyn
and Bacon.
Simons, J. (2000, January 27). Cheap computers bridge
digital divide. Wall Street Journal, p. A22.
Thierer, A. (2002). How free computers are filling the
digital divide. Heritage Foundation Backgrounder
(1361). Retrieved October 15, 2005 from www.her-
itage.org/Research/InternetandTechnology/BG1361.cfm
United States Census Bureau (n.d.). Quickfacts.
Retrieved November 10, 2005, http://quickfacts.cen-
sus.gov/qfd/states/06/0642524.html
Warschauer, M. (2003a). Technology and equity: A com-
parative study. Paper presented at the annual meeting
of the American Educational Research Association,
Chicago, IL.
Warschauer, M. (2003b). Demystifying the digital
divide. Scientific American. 289(2), 43-47.
Warschauer, M. (2003c). Technology and social inclu-
sion: Rethinking the digital divide. Cambridge, MA:
MIT Press
Wenglinsky, H. (1998). Does it compute? The relation-
ship between educational technology and student
achievement in mathematics. Princeton, NJ:
Educational Testing Service.
The High School Journal – Feb/ Mar 2007
41
Training. Variables include: how helpful was
college preparing you to use computers (.521);
how helpful are formal classes to use computers
(.758); Variables coded 1 = very unprepared; 2 =
somewhat prepared; 3 = adequately prepared; 4
= well prepared; 5 = very well prepared; how
many hours of professional development for
using computers and the Internet have you had
(.791). Variables coded 1 = 1-4 hours; 2 = 5-8
hours; 3 = 9-12 hours; 4 = 13-16 hours; 5 = 17-
24 hours; 6 = 25 hours or more.
Researcher created composite variable with
Cronbach’s alpha in parentheses.
C&I home use (_ = .872). Use computer at home;
Use email at home. Variables coded 1 = every
day; 2 = almost every day; 3 = a few times a
week; 4 = between once a week and once a
month; 5 = never.
Individual variables selected
Professional practices: Use C&I for on-line dis-
cussion with collegues; Variable coded 1 = every
day; 2 = almost every day; 3 = a few times a
week; 4 = between once a week and once a
month; 5 = never.
Communication with students: Use C&I for
email communication with students; Variable
coded 1 = every day; 2 = almost every day; 3 =
a few times a week; 4 = between once a week
and once a month; 5 = never.
Endnotes
1 NCES defined high poverty schools as schools in
which at least 71% of the students received a free or
reduced price lunch. Low poverty schools are those
schools with less than 35% of the students receiving
free or reduced price lunch).
2 We used several indicators to distinguish low (S1-S5)
from high resource (S6) schools: (1) the percentage of
students receiving a free or reduced price lunch
(FRPL), S1’s FRPL was 42%; S2 = 42%; S3 = 42%; S4
= 42%; S5 = 42 %; and S6 = 8%; (2) Percentage of
parents with high school diplomas: S1 = 60%, S2 =
64%, S3 = 76%, S4 = 57%, S5 = 58%, S6 = 97%; and
Percentage of parents with college degrees: S1 = 11%,
S2 = 10%, S3 = 18%, S4 = 19%, S5 = 15%, and S6 =
38%.
3 The API rankings receive by the schools were based
on the results of various indicators (i.e., statewide
assessments) including the Standardized Testing and
Reporting (STAR) program and California Standards
Tests. To rank the schools, the California Department
of Education divides API scores into deciles. Schools
are numbered from 1 to 10 (highest rank). Schools
with rankings of 4 or less are designated underper-
forming schools. In this study, S1 had a ranking of 4;
S2 = 3; S3 = 3; S4 = 3; S5= 4; and S6 = 10.
4 To test for equality of variances we use Levene’s test
for homogeneity of variances (SPSS 13.0 for
Windows).
5 A promax rotation is used rather than a varimax rota-
tion because of the theoretical interconnectedness of
the concepts. Varimax rotation would have forced no
correlation between the various factors when in fact
the theoretical operationalization would indicate sys-
tematic covariance between the factors.
Redefining the Digital Divide
42
Source df1 df2 F_ _2
Physical Access
Physical Access to C&I 5 178.6 9.08*** .140
C&I Use
School use of C&I 5 194.9 6.12*** .096
Home use of C&I 5 165.7 3.16*** .055
Instructional practices 5 156.8 1.87* .032
Social consequences
Professional activities 5 114.1 9.08*** .167
Communication with students 5 142.8 2.69** .047
Student Engagement 5 191.4 1.68 .027
Training
Barriers 5 180.2 1.71 .029
Training 5 133.1 .610 .010
*p < .10
**p < .05
***p < .01
Table 1. Analysis of Variance for Teacher C&I Use (F’= Brown-Forsythe Statistic)
S1 S2 S3 S4 S5 S6
S1 —- -.026 .401 .267 .053 -1.148***
S2 —- —- .427 .293 .079 -1.122***
S3 —- -.134 -.348 -1.549***
S4 —- -.214 -1.415***
S5 —- -1.201***
S6 —-
***p < .01
Table 2. Post Hoc Comparisons (Games-Howell) of Mean Differences for Physical Access to C&I by
School (S1-S6)
S1 S2 S3 S4 S5 S6
S1 —- .803 .239 .688 .251 1.310***
S2 —- -.563 -.115 -.551 .508
S3 —- .449 -.013 1.072***
S4 —- -.437 .622
S5 —- 1.059***
S6 —-
***p < .01
Table 3. Post Hoc Comparisons (Games-Howell) of Mean Differences for School use of C&I by
School (S1-S6)
The High School Journal – Feb/ Mar 2007
43
Redefining the Digital Divide
S1 S2 S3 S4 S5 S6
S1 —- .379 .653 -.109 .011 .683
S2 —- .273 -.489 -.368 .304
S3 —- -762** -.641** .030
S4 —- .121 .794
S5 —- .672
S6 —-
** p < .05
Table 4. Post Hoc Comparisons(Games-Howell) of Mean Differences for Home Use of C&I by School
(S1-S6)
S1 S2 S3 S4 S5 S6
S1 —- -.092 -.160 -.027 -215 -.536**
S2 —- -.068 .065 -.123 -.444**
S3 —- .133 -.054 -.375**
S4 —- -.188 -.509***
S5 —- -.321*
S6 —-
*p < .10
** p < .05
***p < .01
Table 5. Post Hoc Comparisons (LSD) of Mean Differences for Instructional Practices by School
(S1-S6)
S1 S2 S3 S4 S5 S6
S1 —- .402 -.209 .354 -.040 1.305***
S2 —- -.611** -.047 -.441 .903
S3 —- .563*** .167 1.514***
S4 —- -.115 .950**
S5 —- 1.345**
S6 —-
** p < .05
***p < .01
Table 6. Post Hoc Comparisons (Games-Howell) of Mean Differences for Professional Activities by
School (S1-S6)
44
The High School Journal – Feb/ Mar 2007
S1 S2 S3 S4 S5 S6
S1 —- .411 .277 .122 .277 .938**
S2 —- -.134 -.288 -.133 .527†††
S3 —- -.154 .001 .662††
S4 —- .155 .817**
S5 —- .661†††
S6 —-
** p < .05
†† p < .05
††† p < .05
Table 7. Post Hoc Comparisons(Games-Howell* and LSDÜ) of Mean Differences for Communication
with Students by School (S1-S6)
