Teaching Sociology http://tso.sagepub.

com/

Effects of Web-Based Instruction on Math Anxiety, the Sense of Mastery, and Global Self-Esteem: A
Quasi-Experimental Study of Undergraduate Statistics Students
Gundy Karen Van, Beth A. Morton, Hope Q. Liu and Jennifer Kline
Teaching Sociology 2006 34: 370
DOI: 10.1177/0092055X0603400404

The online version of this article can be found at:

http://tso.sagepub.com/content/34/4/370

Published by:

http://www.sagepublications.com

On behalf of:

American Sociological Association

Additional services and information for Teaching Sociology can be found at:

Email Alerts: http://tso.sagepub.com/cgi/alerts

Subscriptions: http://tso.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://tso.sagepub.com/content/34/4/370.refs.html

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

 EFFECTS OF WEB-BASED INSTRUCTION ON
MATH ANXIETY, THE SENSE OF MASTERY, AND
GLOBAL SELF-ESTEEM: A QUASI-EXPERIMENTAL
STUDY OF UNDERGRADUATE STATISTICS STUDENTS*
To explore the effects of web-based instruction (WBI) on math anxiety, the
sense of mastery, and global self-esteem, we use quasi-experimental data
from undergraduate statistics students in classes assigned to three study con-
ditions, each with varied access to, and incentive for, the use of online tech-
nologies. Results suggest that when statistics course requirements included
the use of WBI techniques, such as Blackboard’’s (Blackboard Inc. 2001) digi-
tal drop box and online student discussion board, class levels of math anxiety
were reduced from the beginning (Time I) to the end (Time II) of the course
instruction periods. In classes that required student participation in online dis-
cussion forums, self-esteem levels appear to have been enhanced. Perceived
mastery levels, however, were not influenced significantly by use of the WBI
tools we consider here. The findings indicate that the incorporation of WBI
techniques into statistics courses may benefit college students; yet, the
mechanisms by which WBI tools affect student outcomes require elucidation.
We recommend that widespread implementation of WBI follow only from sys-
tematic evaluation of its efficacy across various educational settings, student
populations, and social conditions.

KAREN VAN GUNDY HOPE Q. LIU

University of New Hampshire Children’’s Health Education Center

BETH A. MORTON JENNIFER KLINE

American Institutes for Research University of New Hampshire

MOST UNDERGRADUATE SOCIOLOGY PRO- sonal and academic outcomes are related to
GRAMS require students to take classes in math anxiety, the sense of mastery, and
statistical and methodological reasoning as self-esteem (Ashcraft 2002; Conners,
part of their curricula (Wagenaar 2004). Mccown, and Roskos-Ewoldsen 1998; Fitz-
Such classes tend to draw students with high gerald and Jurs 1996; Lane and Lane 2001;
““math anxiety”” levels (Helmericks 1993; Mone, Baker, and Jeffries 1995; Ross and
Royse and Rompf 1992; Schacht and Stew- Broh 2000; Wang et al. 1999; Wilhite
art 1990), which can erode the personal 1990). Thus, college student success may
resources that predict student success and be enhanced by the implementation of inno-
well-being (Ashcraft 2002). Indeed, per- vative course instruction strategies that ease
student anxieties and promote feelings of
*This study was supported, in part, by the
American Sociological Association’’s Teaching ers for their valuable comments on this paper.
Endowment Fund and a Liberal Arts Faculty Please address all correspondence to Karen Van
Summer Fellowship from the University of New Gundy, Department of Sociology, Horton Social
Hampshire to Karen Van Gundy. An earlier Science Center, University of New Hampshire,
version of this paper was presented at the 2003 Durham, NH 03824-3586; email:
American Sociological Association meetings in karen.vangundy@unh.edu.
Atlanta, GA. We are grateful to the editor of Editor’’s note: The reviewers were, in alpha-
Teaching Sociology and two anonymous review- betical order, Wava Haney and Tim Owens.

Teaching Sociology, Vol. 34, 2006 (October:370-388) 370

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 371
personal confidence, control, and worth. tion that is available strictly to their stu-
One teaching approach with the potential dents.
to improve student outcomes is supplement- The digital drop box in Blackboard
ing conventional classroom instruction with ““allows student to exchange files with the
online technologies that connect students instructor”” and provides students with an
with their classmates and instructors outside electronic space to save copies of their work
class (Khan 1997; Monteith and Smith (Blackboard Inc. 2001:96). That is, students
2001; Owston 1997). While recent studies can turn in assignments electronically or
of college students have examined web- upload files to their own account for later
based discussion board use (Benson et al. use. Students can access the saved files
2002; Persell 2004) and the anxiety- from anywhere they have Internet access.
reducing effects of interactive learning Thus, students using multiple web-
(Townsend et al. 1998), no studies that we connected computers to complete assign-
could identify have considered the extent to ments (e.g., first on-campus and then at
which online communication technologies home) need not save their work on disk
influence changes in math anxiety, the sense since they can edit files saved in the digital
of mastery, and self-esteem. Here we ex- drop box. This can be particularly useful in
plore such processes using quasi- a statistics course, where students may not
experimental data from undergraduate sta- always remember to take a disk to class to
tistics students in classes assigned to three work on computer assignments.
study conditions, each with varied access Finally, Blackboard’’s discussion board is
to, and incentive for, the use of online tech- an online communication tool ““designed for
nologies. Before turning to our analyses, we asynchronous use, meaning that students do
review the literature on web-based instruc- not have to be available at the same time to
tion technologies and examine their hy- have a conversation”” (Blackboard Inc.
pothesized effects on math anxiety, the 2001:76). Discussion forums are generally
sense of mastery, and self-esteem. organized around a topic provided by the
instructor and a series of student-created
WEB-BASED INSTRUCTION discussion ““threads”” that relate to that topic
and to classmate postings. These online
Web-based instruction (WBI) involves the conversations are logged and grouped
use of ““distance learning”” tools that connect around a student’’s initial posting and class-
students, classmates, and instructors via mates’’ related replies. For example, in a
technological media (Greene and Meek statistics course, the instructor might pro-
1998; Keegan 1986). Although such con- vide a forum related to learning statistics
nections can take a variety of forms, we software:
focus on the three features of Blackboard 5
considered in the present study: electronic Discussion forum: ““Are you having problems
with StataQuest? Here you can ask and answer
access to class materials; the digital drop
questions about using the software.””
box; and the student discussion board
(Blackboard Inc. 2001). Electronic avail-
Then students can post questions and an-
ability of documents in Blackboard permits
swers:
students to access course materials outside
of class. Therefore, students who miss an Student 1: ““Does anyone know the syntax for
on-campus class or misplace copies of calculating a frequency distribution?””
course materials can download the materials
on their own time via the web. Unlike with Student 2: ““Yes, just type ‘‘tab’’ then the name
a common Internet webpage, however, of the variable.””
Blackboard access is restricted to enrolled
students and course instructors. Thus, in- Student 1: ““It worked! Thanks!””
structors may provide electronic informa-

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

372 TEACHING SOCIOLOGY
All course participants can view the discus- variations in computer attitudes also are
sion board postings, even if they do not post unclear. A study of 351 undergraduates
messages. Thus, regardless of their own found older students to be prone to com-
direct participation, students may benefit puter anxiety and low computer-related con-
from online exchanges, such as when their fidence (Jennings and Onwuegbuzie 2001).
own questions are asked and answered by However, a study of 384 participants ages
classmates. Moreover, instructors can be 20-75 found a negative association between
informed in advance about potential student age and computer anxiety (Czaja and Sharit
problems with the course material. 1998), while a study of 60 undergraduates
Proponents of WBI suggest that it can found no age effects (Koohang 1986).
enhance the college-learning environment in Though Internet access is on the rise, im-
a number of ways. First, some argue that portant variations in access by age, socio-
because web-based tools are asynchronous, economic, rac ial/eth nic, location
instruction can be more student-centered (rural/urban), and disability statuses remain
(Liu 2003). For instance, students can ac- (Department of Commerce 2000). That
cess materials on their own time and thus some students may be burdened dispropor-
reflect more on the material if needed (Khan tionately by WBI use, therefore, is impor-
1997; Owston 1997). In addition, the use of tant for college educators to recognize and
online communication technologies arguably address.
fosters collaborative learning (Khan 1997; Of additional concern are the limits of
Owston 1997) and offers students who tend electronic ““patterns of interaction and dia-
to avoid in-class discussion a less inhibiting logue”” (Benson et al. 2002; Nicol, Minty,
discussion forum (Chester and Gwynne and Sinclair 2003:278). That is, since
1998; Jaffee 1997). By engaging these and online communication lacks the social cues
other students to post and respond to course present in face-to-face interactions, its ef-
readings, some argue, the quantity and fectiveness as a learning tool may be hin-
quality of student involvement can be im- dered (Benson et al. 2002; Nicol et al.
proved (Persell 2004). 2003; Winiecki 1999). However, we sus-
However, in a recent review of the litera- pect that the advantages of supplemental
ture Benson et al. (2002) found ““little sys- online discussion and other WBI tools
tematic evidence”” regarding the beneficial probably outweigh the disadvantages. Spe-
effects of computer-mediated interaction cifically, we suggest that some WBI fea-
among undergraduate students (p. 142). tures may reduce math anxiety and increase
Moreover, researchers have identified po- levels of perceived mastery and self-esteem
tential problems with the use of WBI in among undergraduate statistics students.
college courses. Of particular concern are
social status inequalities in computer atti- MATH ANXIETY
tudes (Jennings and Onwuegbuzie 2001) and
access (Persell 2004), but existing studies Sharing conceptual ground with general and
that examine such social inequities of the exam anxieties (Fariña et al. 1991; Helmer-
““digital divide”” (Benson et al. 2002:145) icks 1993; Horney 1937; Suinn 1969; Tate
provide inconsistent results. Green 1990), ““math anxiety”” is defined as
Regarding sex, for example, a study of ““feelings of anxiety, dread, nervousness,
316 undergraduate students suggests that and associated bodily symptoms related to
males exhibit more positive computer atti- doing mathematics”” (Fennema and Sherman
tudes (Smith and Necessary 1996), while 1976:324). It involves an ““emotional state””
studies across age and school level (junior of unease or fear regarding ““future math-
high, high school, and college) have found related activities”” (Conners et al. 1998:40;
no evidence for such sex differences Hembree 1990). Perceived math deficits
(Gressard and Lloyd 1987; Jennings and may stem from a range of sources, includ-
Onwuegbuzie 2001; Koohang 1986). Age ing prior aversive math encounters, a sense

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 373
of powerlessness in a technology-driven that help reduce the tediousness of calcula-
society (Helmericks 1993), and/or gender tion and ease computational anxieties
and age-related patterns of socialization or (Alkhateeb 2002).
experience. Indeed, math anxiety tends to Although an early study found that com-
be higher among females (Benson 1989; puter use in statistics courses did not influ-
Betz 1978; Hembree 1990; Llabre and ence math anxiety (Stevens 1982), recent
Suarez 1985; Reyes 1984), and this ten- exploratory work suggests that computer
dency may intensify during the university use in statistics courses may help reduce
years (Sax 1994). math anxiety by increasing familiarity with
Much research documents that math anxi- everyday technology use, encouraging criti-
ety thwarts academic success (Ashcraft cal thinking (Alkhateeb 2002), and promot-
2002; Conners et al. 1998; Fitzgerald and ing perceptions that statistics are useful
Jurs 1996; Schact and Stewart 1990; Tate (Zanakis and Valenzi 1997). It is not clear,
Green 1990; Zanakis and Valenzi 1997). however, whether such effects are consis-
Thus, teaching approaches that offset math tent across sex and age. Perhaps because
anxiety offer clear educational benefits. gender socialization patterns promote male,
Moreover, because of the growth of statis- but not female, development of instrumental
tics software use in college courses1 and the skills (Rosenfield 1999), female students
reliable correlation between computer and may be disadvantaged by computer use re-
math anxieties (Campbell 1988; Collis quirements in their statistics courses. Age-
1987; Dambrot et al. 1985; Fariña et al. related effects of computer exposure, cou-
1991; Glass and Knight 1988; Howard pled with changing cultural expectations,
1986; Jennings and Onwuegbuzie 2001; also may influence the effects of computer
Krendl, Broihier, and Fleetwood 1989; use in statistics courses. For instance, a
Levin and Gordon 1989), promoting com- study of 87 statistics students found younger
puter skills acquisition is a promising ap- female students to be disproportionately
proach for statistics instruction (Alkhateeb troubled by computer use in their courses
2002; Shashaani 1995). (Rendulic and Terrell 2000).
Experience with computers appears to Yet because new computer technologies,
reduce computer anxiety (Glass and Knight like online discussion forums, can encour-
1988; Howard 1986; Jones and Wall 1989- age interactive and collaborative learning,
1990), and perhaps via similar processes, their use in statistics courses can enrich
computer experience may influence math students’’ learning and reduce their anxieties
anxiety and interest (Shashaani 1995). Be- (Benson et al. 2002; Khan 1997; Owston
cause computers increasingly permeate chil- 1997; Rendulic and Terrell 2000; Zanakis
dren’’s environments, many of today’’s col- and Valenzi 1997). Online discussion fo-
lege students have gained computer skills rums offer students opportunities for peer
early in life and feel quite comfortable with teaching and learning, which can have posi-
computers (Alkhateeb 2002; Liu 2003). In tive benefits both for student tutors and the
addition, students tend to be more ““visual”” tutored (Conners et al. 1998; Ward 1984).
learners than past generations (Khan 1997; According to Conners et al., ““By explaining
Owston 1997). Thus, web-based tools may the material to others, [tutors] can identify
reduce math anxiety, in part, by allowing weaknesses in their own knowledge and
““math phobes”” to draw on familiar skills remediate those deficiencies. The [tutored]
benefit from extra individual attention be-
1
Although we could locate no studies that yond what they get from their instructor””
have examined the growth of statistics software (p. 41). Likely due to its promotion of stu-
use in college courses, a search of current intro- dent participation (Tobias 1980, 1995),
ductory college statistics textbooks suggested group or cooperative work appears to in-
that the majority of texts are bundled with some
crease positive math self-concept
kind of statistics software package.
(Townsend et al. 1998) and reduce math

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

374 TEACHING SOCIOLOGY
and other anxieties (Johnson and Johnson very occurrence of some stressors”” (Turner,
1989; Slavin 1990; Zanakis and Valenzi Taylor, and Van Gundy 2004:37). Even
1997). In fact, an exploratory study sug- though math anxieties may be unavoidable,
gests that statistics test scores may be higher self-efficacious students may not be as ad-
among online students in contrast with tra- versely affected by them. Indeed, it appears
ditional students, a result attributed to in- that academic achievement increases a sense
creased opportunities for, and the higher of personal control, which in turn, fosters
quality of, collaborative learning among subsequent achievement (Ross and Broh
online students (Schutte 1996). Such tech- 2000). Moreover, a sense of mastery tends
niques may also contribute to students’’ per- to promote problem solving and effective
ceived sense of mastery, which can further coping (Mirowsky and Ross 1989; Ross and
enhance their learning. Mirowsky 1989; Ross and Sastry 1999;
Wheaton 1983), which cultivate general
THE SENSE OF MASTERY well-being and academic accomplishment.
Given the academic and emotional harm
Related to notions of self-efficacy (Bandura associated with perceived fatalism, attention
1977), locus of control (Lefcourt 1976), and to instruction techniques that promote stu-
personal control (Gurin, Gurin, and Morri- dents’’ perceived mastery is needed. WBI is
son 1978), the sense of mastery is defined a potentially useful tool in this regard be-
as ““the extent to which one regards one’’s cause it is both asynchronous and interac-
life-chances as being under one’’s own con- tive. That is, asynchronous approaches can
trol in contrast to being fatalistically ruled”” promote self-regulated learning, which ap-
(Pearlin and Schooler 1978:5). Its converse, pears to increase students’’ self-confidence
fatalism, involves ““a tendency to believe in in Internet use (Joo et al. 2000). Female
the efficacy of environmental rather than students, in particular, seem to be more
personal forces in understanding the causes self-efficacious in their self-regulated learn-
of life outcomes”” (Wheaton 1983:211). In ing (Joo et al. 2000). Moreover, a sense of
the face of math anxieties or related emo- academic competence may be enhanced by
tional duress, student achievement in col- cooperative learning strategies (Johnson and
lege statistics courses is likely enhanced by Johnson 1989; Slavin 1990), like those
a sense of personal mastery and/or under- available via the online discussion board.
mined by a sense of fatalism. That is, stu- Online forums that encourage students to
dents who feel that they cannot control their explain the course material to their class-
own academic or personal fate may, in turn, mates may enhance mastery of the material
be less likely to succeed. Indeed, it appears for both peer tutors and the tutored
that a sense of mastery is reliably linked to (Conners et al. 1998). Similarly, self-
educational outcomes (Chemers, Hu, and esteem may be enhanced by such interac-
Garcia 2001; D’’Amico and Cardaci 2003; tions.
Joo, Bong, and Choi 2000; Mone et al.
1995; Randhawa, Beamer, and Lundberg SELF-ESTEEM
1993; Wang et al. 1999; Wilhite 1990).
There are a number of potential mecha- According to Rosenberg (1965:5), global
nisms by which a sense of mastery may self-esteem involves ““an attitude of ap-
improve academic and personal outcomes. proval or disapproval toward oneself”” that
To the extent that one’’s sense of control develops from processes of reflected ap-
reflects a personal history of success or praisal (Cooley 1912; Mead 1934), social
failure, it may imply a level of instrumental comparison (Festinger 1954; Pettigrew
ability or effectiveness (White 1959). High 1967), and self-attribution (Rosenberg
levels of perceived mastery may promote 1979). That is, ““in the absence of objective
the ““recognition of one’’s ability to alter or information about themselves people judge
adapt to circumstances and to minimize the themselves on the basis of comparison with

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 375
others”” (Rosenberg, Schooler, and Schoen- since online forums furnish students with
bach 1989:1006). Related to the formation the knowledge that their classmates may be
of personal mastery, self-attributions may struggling with difficult coursework, stu-
additionally derive, in part, from one’’s per- dent self-comparisons are probably more
sonal history of success or failure (Gecas favorable than they would be in more tradi-
and Schwalbe 1983; Pugliesi 1989). Thus, tional classroom settings. There is evidence
in the context of the college statistics that interactive learning techniques help
course, student self-esteem may be impor- reduce emotional duress, in part, by allevi-
tantly influenced both by individual aca- ating stress ““among people with similar
demic achievements and by classmate inter- fears facing the same problems”” (Johnson
actions. and Johnson 1989; Slavin 1990; Townsend
Evidence regarding the effect of self- et al. 1998:43). By extension, such tech-
esteem on academic outcomes is mixed. niques similarly may serve to promote a
Some studies have found that self-esteem positive sense of self-worth among students
and ““positive self-worth”” increase educa-
tional achievement and attainment (Owens RESEARCH QUESTION
1994:391; Wang et al. 1999). Others sug-
gest that such a link is the result of an effect In sum, we ask: Does the supplemental use
of academic accomplishment on subsequent of WBI tools in undergraduate statistics
self-esteem levels (Rosenberg et al. 1989; courses influence students’’ math anxiety,
Ross and Broh 2000). Other research notes perceived mastery, and self-esteem? In par-
a reciprocal relationship such that self- ticular, we test whether computer experi-
esteem ““both influences and is influenced ence (e.g., uploading and downloading files
by academic achievement”” (Liu, Kaplan, from Blackboard) and online interaction
and Risser 1992:141). Yet a growing body (e.g., discussion forums) affect these out-
of work finds that global self-esteem does comes. Due to the inconsistencies noted in
little to foster academic capabilities the literature, we leave the direction and
(Covington 1992, 2001; D’’Amico and Car- nature of their effects open for discovery.
daci 2003; Moeller 1994; Shaw 1994), at
least among students who do not value METHODS
school success (Rosenberg et al. 1995).2
Notwithstanding the ambiguous effects of Sample and Data Collection
self-esteem on academic outcomes, how- We collected data in fall 2002 and spring
ever, studies document robust links between 2003 in a quasi-experiment conducted in
global self-esteem and psychological well- four undergraduate statistics classes at a
being (Rosenberg et al. 1995; Thoits 1995). predominantly white university in the north-
Thus, the promotion of student self-esteem east United States. Enrolled students (N =
is a potentially useful goal for concerned 175) received extra credit for voluntary
educators. participation in the study. We administered
Of the WBI tools we examine here, the pre-test and post-test surveys in all classes
online discussion board is the most relevant at the beginning (Time I) and end (Time II)
for self-esteem enhancement. Prior research of the course instruction periods, respec-
suggests that cooperative learning strategies tively. We informed students about their
increase student self-esteem (Johnson and rights as human subjects and about the con-
Johnson 1989; Slavin 1990). Moreover, fidentiality and anonymity of the study be-
fore they answered the surveys. We asked
2
Rosenberg et al. (1995:141) suggest that students to write their mother’’s maiden
specific (academic) self-esteem is a ““much bet- name (or some other word) on pre- and
ter predictor of school performance”” than is post-test surveys so that they could be
global self-esteem. Yet global self-esteem still
matched. The first author taught all courses.
appears to exert statistically significant effects
on school marks net of specific self-esteem.
In all sections lectures were delivered using

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

376 TEACHING SOCIOLOGY
PowerPoint presentations; students were percent; the final class sample was 30. Stu-
required to learn and use StataQuest statis- dents who dropped this class scored higher
tics software (Anagnoson and DeLeon on math anxiety compared to those who did
1997); and at least four class meetings were not (p < .10).
held in a computer lab to assist students Analyses (available upon request) sug-
with learning and using StataQuest and gested that attrition rates were consistent
web-based materials. across the classes, with one exception: stu-
Class 1 (n = 41) met on Tuesdays and dents were more likely to drop Class 1 com-
Thursdays from 9:40 a.m. to 11:00 a.m. pared to Class 2 (p < .05). We suspect that
during fall semester 2002. The attrition rate this difference may be attributable, in part,
for this class was 41.4 percent, resulting in to variations in technology-related expecta-
a final class sample of 24 students. Analy- tions and attitudes. While Class 2 students
ses (not shown) suggested few differences were not required to download course as-
between students who remained in the class signments from the web, Class 1 students
compared to those who dropped it, with one were required to do so (see Conditions A
notable exception: students who dropped and B below). Thus, students may have
Class 1 tended to have less favorable tech- been more likely to drop Class 1 because its
nology-related attitudes3 than those who requirements were more challenging. That
remained in the class (p < .01). Class 2 (n view is supported by the finding that Class
= 46) met on Tuesdays and Thursdays 1 dropouts held less favorable technology-
from 11:10 a.m. to 12:30 p.m. in fall 2002. related attitudes3 than those who remained
The attrition rate was 19.5 percent, result- in the class. Though class dropouts may
ing in a final class sample of 37. We ob- have enrolled in other statistics courses, no
served no differences between respondents students who dropped a class in this study
who dropped and those who remained in the enrolled in another class examined here.
class. Class 3 (n = 46) met on Tuesdays Computer anxiety levels (Heinssen, Glass,
and Thursdays from 9:40 a.m. to 11:00 and Knight 1987)4 did not vary by class at
a.m. during spring semester 2003. The at- Time I or Time II and were similar for stu-
trition rate for Class 3 was 26.0 percent; the dents who dropped or remained in all four
final class sample was 34. Students who classes. Each class participated in one of
dropped the class were more likely to be three study conditions, the characteristics of
male than those who remained (p < .10). which appear in Table 1. Class 1 partici-
Class 4 (n = 42) met on Tuesdays and pated in Condition A, Class 2 participated
Thursdays from 2:10 p.m. to 3:30 p.m. in in Condition B, and classes 3 and 4 partici-
spring 2003. The attrition rate was 28.5 pated in Condition C.
3 4
At Time I and Time II, we assessed the ex- We used a shortened and modified version of
tent to which students perceived that the empha- the Computer Anxiety Rating Scale (CARS),
sis on and/or use of technology was useful, developed by Heinssen, Glass, and Knight
particularly in classroom settings. The first and (1987), to measure computer anxiety. At both
third authors developed scale items for use in Time I and Time II, we asked students to what
the larger study. The measurement scale con- extent they agreed with ten statements about
sisted of the sum of seven items, which ranged their perceptions of computer-related worry,
from 1 to 4 (strongly disagree to strongly uneasiness, and confusion. We summed these
agree). Higher scores reflect more favorable ten items, which ranged from 1 to 4 (strongly
technology-related attitudes. At Time I, scores disagree to strongly agree), to create the scale.
range from 8 to 27; at Time II, scores range Higher scores reflect higher computer anxiety
from 8 to 28. Means (and standard deviations) levels. At both Time I and Time II, scores range
are 18.90 (3.54) and 19.21 (4.15) for Time I from 10 to 25. Mean (and standard deviation)
and Time II, respectively. Respective Cron- scores are 20.26 (5.85) and 19.45 (5.57), re-
bach’’s (1951) alphas at Time I and Time II are spectively. Time I and Time II Cronbach’’s
.763 and .861. (1951) alpha scores are .875 and .860.

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 377
Study Conditions expected students to use at least one of nine
Condition A. We required students in Con- online discussion forums to post questions
dition A to use Blackboard 5 (Blackboard or answers to classmates’’ questions. Fo-
Inc. 2001) and we made assignments and rums addressed topics like learning and
data available only via Blackboard. Thus, using StataQuest, completing assignments,
we required students to use the web to ob- and reviewing for exams. To promote col-
tain such materials. In addition, we made laborative work outside of class, we encour-
class e-mail and an online discussion board aged students to use the forums prior to
available to students via Blackboard. We homework deadlines and exams. Since we
hoped that such web-based communication required students to post questions and/or
venues would facilitate peer teaching and answers to questions, we presume they en-
learning online. However, we did not re- gaged in more interactive work than those
quire students to use these technologies. in conditions without this requirement.
Finally, we required students to upload Sta- Moreover, we held students accountable for
taQuest data and log files via Blackboard’’s the merit of their forum postings. Unlike in
digital drop box. the other conditions, then, students were
Condition B. Students in Condition B did likely more motivated to examine course
not have access to Blackboard. Instead, we materials carefully in order to articulate
limited web-based resources to general in- online questions and answers clearly and
formation (syllabi, assignments, and data) correctly. Finally, the discussion board
posted on a one-page class website. With served as an alternative source of informa-
the exception of the class data, however, we tion and/or tutoring unavailable to students
also distributed all information posted on in the other conditions.
the website on paper in class. Initially, we
limited e-mail contact only to the instructor Measures
and teaching assistant, but students volun- Math anxiety. We used a ten-item modified
tarily composed an e-mail list for those in- version of one of Fennema and Sherman’’s
terested in having electronic contact with (1976) nine Mathematics Attitudes Scales
classmates. Finally, we required students to (MAS) to measure math anxiety at Time I
turn in StataQuest data and log files on and Time II. We adapted the items to also
disks. include concerns about statistics. Response
Condition C. As in Condition A, we re- choices for the items were: (1) strongly
quired students in Condition C to use Black- disagree, (2) somewhat disagree, (3) some-
board to upload data and log files to the what agree, and (4) strongly agree. We cre-
digital drop box. Unlike Condition A, how- ated the scale by summing the ten items
ever, we made assignments available via such that higher scores reflect higher math
Blackboard and on paper in class. In addi- anxiety levels. Scores range from 11 to 40
tion, we required students to use the online at Time I and from 10 to 40 at Time II.
discussion board in Blackboard. That is, at Mean (and standard deviation) scores are
least four times during the semester, we 26.64 (7.16) and 24.79 (7.39) for Time I

Table 1. Characteristics of the Three Study Conditions

Condition A Condition B Condition C
Access to paper copies of course materials in class * *
Access to course materials via a common webpage *
Access to course materials via Blackboard * *
Access to an online student discussion board * *
Required turning in files via floppy disk *
Required downloading assignments from Blackboard *
Required uploading files to the digital drop box * *
Required using the online student discussion board *

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

378 TEACHING SOCIOLOGY
and Time II, respectively. Respective Cron- spective mean (and standard deviation)
bach’’s (1951) alpha scores are .935 and scores are 22.70 (3.14) and 23.36 (3.56) at
.930 at Time I and Time II. Table 2 shows Time I and Time II. Time I and Time II
items and principal components factor load- alpha scores (Cronbach 1951) are .731 and
ings for this scale and the other scales. .788, respectively.
Sense of mastery. We used Pearlin and Self-esteem. A six-item subset from
Schooler’’s (1978) seven-item scale to meas- Rosenberg’’s (1979) measure assessed global
ure perceived mastery at both Time I and self-esteem, measured at Time I and Time
Time II. Respondents indicated the extent to II. Respondents indicated the extent to
which they felt a sense of personal control which they agreed with items about per-
over life circumstances. Response choices ceived self value or worth. Responses
ranged from 1 (strongly disagree) to 4 ranged from 1 (strongly disagree) to 4
(strongly agree). We summed items such (strongly agree). We summed items such
that higher scores indicate a greater sense of that higher scores reflect higher self-esteem
mastery. Scores range from 15 to 28 at levels. Scores range from 10 to 24 and 13
Time I and from 14 to 28 at Time II. Re- to 24 at Time I and Time II, respectively.

Table 2. Factor Loadings for Measurement Scales (N = 125)

Factor Loadings
Time I Time II
Math Anxiety
1. It wouldn’’t bother me at all to take more math or statistics courses * .642 .654
2. I have usually been at ease during math or statistics tests * .836 .823
3. I have usually been at ease during math or statistics courses * .786 .820
4. I usually don’’t worry about my ability to solve math or statistics problems * .773 .809
5. I almost never get uptight when taking math or statistics tests * .798 .794
6. I get really uptight during math or statistics tests .843 .767
7. I get a sinking feeling when I think of trying hard math or statistics problems .806 .763
8. My mind goes blank when I think of trying hard math or statistics problems .744 .771
9. Mathematics and statistics make me feel uncomfortable and nervous .851 .807
10. Mathematics and statistics make me feel uneasy and confused .894 .837
Cronbach’’s (1951) alpha .935 .930
Sense of Mastery
1. There is little I can do to change many of the important things in my life * .570 .676
2. What happens to me in the future mostly depends on me .455 .575
3. I have little control over the things that happen to me * .619 .662
4. There is really no way I can solve some of the problems I have * .686 .624
5. I often feel helpless in dealing with problems of life * .720 .747
6. Sometimes I feel I’’m being pushed around in life * .602 .722
7. I can do anything I really set my mind to .671 .660
Cronbach’’s (1951) alpha .731 .788
Self-Esteem
1. I am able to do things as well as others .539 .531
2. I am satisfied with myself .775 .788
3. I take a positive attitude towards my self .669 .850
4. I am inclined to feel that I am a failure * .598 .678
5. I feel I have a number of good qualities .759 .790
6. I am a person of worth at least equal to others .558 .638
Cronbach’’s (1951) alpha .722 .808
Note: Presented are principal components factor loadings.
* Reverse coded.

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 379
Respective mean (and standard deviation) (1) for students who lived on campus and
scores at Time I and Time II are 21.19 (0) for those who did not. Fifty-nine percent
(2.50) and 21.39 (2.64). Internal reliability of the sample reported living on campus.
scores (Cronbach 1951) are .722 and .808 We coded gender (1) for females and (0)
at Time I and Time II, respectively. for males. Sixty-eight percent of the sample
Sociodemographic characteristics. Table was female. We assessed age by a four-
3 presents sociodemographic characteristics category item as follows: (0) 18 years, (1)
for the total sample and by study condition. 19 years, (2) 20 years, (3) 21 or more
Students’’ expected course grade, assessed at years. Eighteen percent of the total sample
Time II, is based on an item that asked, was 18 years old, 37.6 percent was 19,
““Based on your progress in this course so 24.0 percent was 20, and 20.8 percent was
far, what do you expect to receive as your age 21 or older. As Table 3 indicates, so-
final course grade?”” Twenty percent of stu- ciodemographic characteristics are consis-
dents reported expecting a grade of A–– or tent across the study conditions with one
better; 54.4 percent expected a B–– to B+; exception: relative to students in Condition
and 25.6 percent expected a C+ or lower C, Condition A students were more likely
grade. Major requirement is a dichotomous to expect a grade of B–– to B+ (p < .10).
variable coded (1) for students who re-
ported taking the class to fulfill a require- RESULTS
ment for their major and (0) for those who
did not. Forty-seven percent of the sample Table 4 presents mean levels of Time I and
reported taking the course for a major re- Time II math anxiety, sense of mastery, and
quirement. We coded on-campus resident self-esteem by sociodemographic character-

Table 3. Sociodemographic Characteristics for the Sample and by Study Condition

Total Sample Condition A Condition B Condition C
(N = 125) (n= 24) (n = 37) (n = 64)
Expected Course Grade††
A–– or above 20.0 29.1 24.3 14.0
B–– to B+ 54.4 37.5 * 56.7 59.3 *
C+ or below 25.6 33.3 18.9 26.5
Major Requirement
Yes 47.2 37.5 43.2 53.1
No 52.8 62.5 56.8 46.9
On-Campus Resident
Yes 59.2 62.5 48.6 64.0
No 40.8 37.5 51.4 36.0
Gender
Female 68.0 54.1 70.2 71.8
Male 32.0 45.8 29.8 28.2
Age
18 years 17.6 25.0 16.2 15.6
19 years 37.6 45.8 40.5 32.8
20 years 24.0 16.6 24.3 26.5
21+ years 20.8 12.5 18.9 25.0
Note: Presented are percentage scores. Scores may not sum to 100 due to rounding.
††
Missing data: expected course grade (N = 123)
* p < .10 (Condition A vs. Condition C; two-tailed test)

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

380 TEACHING SOCIOLOGY
Table 4. Means Levels of Math Anxiety, Sense of Mastery, and Self-Esteem by Sociodemographic
Characteristic N = 125)
Math Anxiety Sense of Mastery Self-Esteem
Time I Time II Time I Time II Time I Time II
††
Expected Course Grade
A–– or above 23.04 19.69 23.76 23.76 21.20 21.52
B–– to B+ 27.35 25.22 22.33 23.26 21.07 21.39
C+ or below 27.25 27.06 22.65 23.25 21.43 21.28
Mean Difference‡‡ p = .032 p = .000 ns ns ns ns
Major Requirement
Yes 25.25 22.77 22.54 23.20 20.86 21.23
No 27.87 26.59 22.84 23.50 21.48 21.53
Mean Difference‡‡ p = .040 p = .003 ns ns ns ns
On-Campus Resident
Yes 27.00 25.54 22.55 23.28 20.98 21.02
No 26.11 23.71 22.92 23.47 21.49 21.92
Mean Difference‡‡ ns ns ns ns ns ns
Gender
Female 26.98 24.64 22.55 23.11 21.24 21.43
Male 25.90 25.12 23.02 23.87 21.07 21.30
Mean Difference‡‡ ns ns ns ns ns ns
Age
18 years 27.45 24.40 22.13 22.95 20.31 20.40
19 years 25.31 24.12 22.48 23.14 21.02 21.23
20 years 27.23 25.60 22.83 23.20 21.83 21.43
21+ years 27.65 25.40 23.42 24.26 21.50 22.46
Mean Difference‡‡ ns ns ns ns ns p = .056
†† Missing data: expected course grade (N = 123)
‡‡ Probability values based on ANOVA results; ns = nonsignificant at p < .10

istic. It appears that at both Time I and and II) and whether or not the course was a
Time II, math anxiety levels were lowest major requirement (Time I only) among
among students who expected better course students in conditions A and B relative to
grades and were taking the class to fulfill a those in condition C. Among Condition A
major requirement. At Time II, older stu- students, Time II math anxiety levels were
dents had greater self-esteem levels than greater among males relative to females.
younger students. There are no residential, Finally, Time I self-esteem levels were as-
sex, or age differences in math anxiety or sociated with higher grade expectations
the sense of mastery, nor are there residen- among condition A students and lower
tial or sex differences in self-esteem. grade expectations among condition B stu-
In separate analyses (not shown), we ex- dents.
amined the extent to which the bivariate Table 5 presents mean scores for Time I
patterns noted in Table 4 were generally and Time II math anxiety, sense of mastery,
consistent across the three study conditions. and self-esteem for the total sample and by
Although power limitations preclude mak- study condition. According to the table,
ing reliable statistical comparisons, similar Time II math anxiety scores are signifi-
patterns seemed to emerge across the study cantly lower (p = .000) than Time I scores
conditions, with a few exceptions. First, for the total sample. Thus, over the course
math anxiety levels appeared to be less instruction period, participation in the statis-
clearly linked to grade expectations (Time I tics course appears to have reduced student

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 381
12
Table 5. Mean Differences in Math Anxiety, Sense of Master, and Self-Esteem at Time I and Time
II for the Sample and by Study Condition (N = 125)
Math Anxiety Sense of Mastery Self-Esteem
Total Sample
Time I 26.64 22.70 21.19
Time II 24.79 23.36 21.39
Mean Difference†† p = .000 p = .006 ns
Condition A
Time I 24.95 a 22.66 21.25
Time II 21.95 b 23.16 20.91
Mean Difference†† p = .010 ns ns
Condition B
Time I 25.54 22.72 21.10
Time II 24.62 23.35 20.83 c
Mean Difference†† ns ns ns
Condition C
Time I 27.90 a 22.70 21.21
Time II 25.96 b 23.43 21.89 c
Mean Difference†† p = .000 p = .038 p = .028

Note: Number of cases for Condition A, B, and C are 24, 37, and 64, respectively.
††
Differences in time-lagged class-level mean scores; ns = nonsignificant at p < .10 (two-tailed tests)
a
p < .10 (Condition A vs. Condition C at Time I; two-tailed test)
b
p < .10 (Condition A vs. Condition C at Time II; two-tailed test)
c
p < .10 (Condition B vs. Condition C at Time II; two-tailed test)

math anxiety. Interestingly, this effect may limitations preclude drawing such a conclu-
be more pronounced for students enrolled in sion here since Conditions A and B have
the Blackboard courses. That is, time- smaller sample sizes. In fact, mean differ-
lagged mean scores are significantly differ- ences in Time I and Time II mastery levels
ent among Condition A (p = .010) and appear generally consistent across the study
Condition C (p = .000) students, but not conditions. Apparently, online course com-
among students in Condition B. It appears, ponents do not notably affect the sense of
then, that the required use of web-based mastery; at least they did not do so among
instructional materials may ease student students in our sample.
math anxieties. It is noteworthy, moreover, Based on results from the total sample, it
that such effects are observed even among appears that self-esteem was not influenced
Condition A students, whose baseline levels by participation in the statistics course. That
of math anxiety were already relatively low. is, according to Table 5, Time I and Time
Table 5 also suggests that a general sense II self-esteem levels are similar for the en-
of mastery may have been enhanced by stu- tire sample. However, within-condition
dent participation in the statistics course. results suggest that, relative to Time I,
That is, Time II mastery levels are signifi- Time II self-esteem levels are significantly
cantly higher than Time I levels for the total higher among condition C students only (p
sample (p = .006). Within-condition analy- = .028). Moreover, self-esteem levels are
ses suggest that the time-lagged mean dif- slightly (albeit not significantly) reduced for
ferences in mastery are statistically signifi- students in Conditions A and B. Thus our
cant only among Condition C students (p = results support the notion that the use of an
.038). While this result supports the idea online class discussion board enhances self-
that online class discussion board use in- esteem among undergraduate statistics stu-
creases perceived mastery, statistical power dents.

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

382 TEACHING SOCIOLOGY
12
DISCUSSION pear to contribute to student self-esteem.
We found significant elevations in class
Undergraduate sociology majors often dread self-esteem levels when student use of the
fulfilling their statistics requirement——a cir- online discussion board was required. For
cumstance that can undermine their aca- classes with no discussion board require-
demic and personal well-being. Thus in- ment, self-esteem levels were similar (if not
structors need strategies that alleviate stu- slightly lower) at Time II relative to Time I.
dents’’ math anxieties and promote a sense Thus, despite the noted interpersonal limits
of mastery and positive self-esteem. Web- of electronic patterns of communication
based instruction (WBI) techniques, such as (Benson et al. 2002; Nicol et al. 2003;
Blackboard’’s digital drop box and online Winiecki 1999), the self-esteem benefits of
student discussion board, are potentially face-to-face cooperative learning (Johnson
useful tools in this regard. Indeed, our find- and Johnson 1989; Slavin 1990) seem to
ings suggest that supplementing conven- extend to web-based approaches that are
tional course instruction with WBI methods interactive or collaborative. Moreover, in
is a promising new approach for statistics response to an item asking students to
students and instructors. ““share any additional observations or com-
We observed reductions in math anxiety ments regarding [their] experiences with the
in courses that required student use of WBI online components of [the] course,”” many
technologies via Blackboard (Blackboard students in Condition C indicated that the
Inc. 2001). That is, classes that required discussion board was extremely helpful.
students to download course assignments, Nearly all students in that condition used
upload files to the digital drop box, and/or the discussion board well beyond the four
participate in online student discussion fo- required times. This suggests that online
rums showed significantly lower levels of discussion forums not only enhance self-
math anxiety at Time II relative to Time I. esteem, but are valued by students.
The class that did not use these Blackboard In contrast, some students in conditions
tools showed no noticeable change in math without the online discussion requirement
anxiety over the course of the semester. voiced concern about a sense of alienation
Interestingly, observed changes in math that may develop from the reliance upon
anxiety apparently are not attributable to technology. These students indicated a pref-
student participation in web-based discus- erence for face-to-face interaction with in-
sions. We found that math anxiety was re- structors and classmates. As one student
duced both for classes that used and did not noted, introverted students who are uncom-
use the online discussion forums.5 Common fortable asking questions or participating in
among these classes was their access to, and class discussions may be less likely to de-
required use of, the digital drop box. We velop communication skills in courses that
speculate, therefore, that reductions in math emphasize electronic interaction. In addi-
anxiety may derive from an effect of in- tion, while responses suggested that stu-
creased computer familiarity or experience, dents differed in their opinions about web-
which tends to relieve computer anxieties based technologies and how they should be
(Glass and Knight 1988; Howard 1986; used, a common theme was that technology
Jones and Wall 1990), and may, via parallel should complement traditional course ap-
processes, alleviate math anxieties proaches rather than replace them.
(Alkhateeb 2002; Shashaani 1995; Zanakis Surprisingly, WBI does not appear to
and Valenzi 1997). influence students’’ perceived mastery.
Yet web-based discussion forums do ap- Though a sense of mastery tended to im-
5
Although students in Condition A had access prove over the course of the instruction
to an online discussion board, they were not period in all classes, such improvement did
required to use it. In fact, we observed only one not vary significantly by class access to, or
student posting over the instruction period. use of, the WBI technologies considered

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 383
here. Moreover, while math anxiety was dents can access postings of classmate an-
linked to lower expected course grades, swers to common questions via the online
neither perceived mastery nor self-esteem discussion board, instructor time spent ex-
seemed to affect grade expectations. This plaining the same material to multiple stu-
suggests that such self-concept dimensions dents can be minimized. Finally, the discus-
do not affect course success. Arguably, sion board also gives the instructor a sense
then, the enhancement of students’’ self- of the concepts and skills with which stu-
confidence and self-worth is not a practical dents struggle outside of class. This knowl-
objective for statistics instructors. Yet, as edge can be used to specifically address
Beane (1982) observes, ““improving self- such problems in class.
concept is an important goal for its own
sake”” (p. 504). Indeed, regardless of their LIMITATIONS
effects on course achievement, the sense of
mastery and self-esteem are crucial re- Our results, while suggestive, should be
sources for the maintenance of mental interpreted in view of the study’’s limita-
health and well-being (see Thoits 1995). tions. First, the study relies on a quasi-
Although our quantitative analyses experimental design for which assignment
showed no bivariate effect of on-campus of class conditions was nonrandom. While
residential status on math anxiety, the sense such a design may introduce selection bi-
of mastery, or self-esteem, students’’ open- ases, our analyses suggest that class condi-
ended observations suggest that commuters tions did not vary significantly by relevant
may feel disadvantaged with respect to tech- characteristics like age, sex, on-campus
nology access. Commuters in Condition A, residency, computer anxiety, and whether
for example, expressed that often they did or not the class was a major requirement. In
not have Internet access from home, and as addition, a small sample size, such as ours,
a result, had to make special trips to campus increases the probability of a Type II error
to download assignments from campus com- decision. Thus the nonsignificant statistical
puter clusters or from friends who had on- associations observed here might be due to
campus capabilities. One student said that a lack of statistical power. Also unclear is
he could not ““access Blackboard for about whether results can be generalized across
the first month of school,”” an experience semester or academic context. Thus, future
that he found ““nerve wracking to say the studies that use larger representative sam-
least.”” Thus, the potential for residential ples of students at various institutions are
and other access-related obstacles are im- needed.
portant for instructors to acknowledge and In addition, our study considers only a
confront. few WBI tools and their uses. For instance,
In addition, WBI use may both advantage the use of private discussion forums (e.g.,
and disadvantage course instructors. Setting an instructor and one student) or structured
up the web-based portion of the course and online assignments (e.g., quizzes) might
learning to use its tools can require a size- yield different results. In fact, future re-
able time investment initially. Once instruc- search should consider a broad range of
tors are familiar with the online course available techniques and their uses. More-
components, however, there are potential over, it is not clear if the effects we observe
benefits. For instance, requiring students to extend to other course types. For compara-
upload files to the digital drop box allows tive purposes, therefore, we examined time-
instructors to collect and store copies of lagged mean differences in math anxiety,
student coursework, which provides extra mastery, and self-esteem in two additional
copies of misplaced student papers, reduces undergraduate sociology courses, Drugs and
a need to collect computer work on disks, Society (n = 29) and Sociological Analysis
and helps detect (and perhaps limit) student (n = 23), which the first author taught dur-
plagiarism. Also, because all enrolled stu- ing fall 2002 and fall 2003, respectively

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

384 TEACHING SOCIOLOGY
(analyses not shown). Blackboard was avail- Esteem as Curriculum Issues.”” Educational
able for both classes, although its use was Leadership 39:504-6.
not required for Drugs and Society, and Benson, Denzel E., Wava Haney, Tracy E. Ore,
neither class required student use of online Caroline Hodges Persell, Aileen Schulte,
James Steele, and Idee Winfield. 2002.
class discussion forums. Students in Socio-
““Digital Technologies and the Scholarship of
logical Analysis were required to upload Teaching and Learning in Sociology.”” Teach-
assignments to the digital drop box. Analy- ing Sociology 30:140-57.
ses suggested that neither class showed sig- Benson, Jeri. 1989. ““Structural Components of
nificant changes in math anxiety, mastery, Statistical Test Anxiety in Adults: An Explora-
or self-esteem. Such results imply that the tory Model.”” Journal of Experimental Educa-
WBI effects observed may be specific to tion 57:247-61.
statistics courses. Betz, Nancy E. 1978. ““Prevalence, Distribution
The mechanisms by which WBI tech- and Correlates of Math Anxiety in College
Students.”” Journal of Counseling Psychology
niques influence educational and psycho-
25:441-8.
logical outcomes remain elusive. Our re- Blackboard Inc. 2001. Blackboard 5 Level 1
sults suggest that student use of some Black- Student Manual. Washington DC: Blackboard
board techniques may ease math anxiety and Inc.
promote self-esteem in college-level statis- Campbell, N. Jo. 1988. ““Correlates of Computer
tics courses, but those effects may derive Anxiety of Adolescent Students.”” Journal of
from various sources. For example, it may Adolescent Research 3:107-17.
not be the effects of WBI per se, but result- Chemers, Marin M., Li-tze Hu, and Ben F.
ing student engagement levels, that produce Garcia. 2001. ““Academic Self-Efficacy and
the results we observe. To maximize the First-Year College Student Performance and
Adjustment.”” Journal of Educational Psychol-
effectiveness of WBI tools, future work
ogy 93:55-64.
should attempt to clarify how they affect Chester, Andrea and Gillian Gwynne. 1998.
student outcomes. We recommend that the ““Online Teaching: Encouraging Collaboration
widespread implementation of innovative through Anonymity.”” Journal of Computer-
instruction methods, like WBI, should fol- Mediated Communication. Retrieved Septem-
low only from systematic evaluation of their ber 1, 2006 (http://jcmc.huji.ac.il/vol14/issue
efficacy across educational settings, diverse 2/chester.html).
student populations, and various social con- Collis, Betty. 1987. ““Sex Differences in the
ditions and situations. Association between Secondary School Stu-
dents’’ Attitudes toward Mathematics and to-
REFERENCES ward Computers.”” Journal for Research in
Mathematics Education 18:394-402.
Alkhateeb, Haitham M. 2002. ““A Preliminary Conners, Frances A., Steven M. Mccown, and
Study of Achievement, Attitudes toward Suc- Beverly Roskos-Ewoldsen. 1998. ““Unique
cess in Mathematics, and Mathematics Anxiety Challenges in Teaching Undergraduate Statis-
with Technology Based Instruction in Brief tics.”” Teaching of Psychology 25:40-2.
Calculus.”” Psychological Reports 90:47-57. Cooley, Charles H. 1912. Human Nature and
Anagnoson, Theodore J. and Richard E. the Social Order. New York: Scribner’’s.
DeLeon. 1997. StataQuest 4. Belmont, CA: Covington, Martin V. 1992. Making the Grade:
Duxbury Press. A Self-Worth Perspective on Motivation and
Ashcraft, Mark H. 2002. ““Math Anxiety: Per- School Reform. New York: Cambridge Uni-
sonal, Educational, and Cognitive Conse- versity Press.
quences.”” Current Directions in Psychological _____. 2001. ““The Science and Politics of Self-
Science: The Journal of the American Psycho- Esteem: Schools Caught in the Middle.”” Pp.
logical Association 11:181-5. 351-74 in Extending Self-Esteem Theory and
Bandura, Albert. 1977. ““Self-Efficacy: Toward a Research: Sociological and Psychological
Unified Theory of Behavioral Change.”” Psy- Currents, edited by Timothy J. Owens, Shel-
chological Review 84:191-215. don Stryker, and Norman Goodman. New
Beane, James A. 1982. ““Self-Concept and Self- York: Cambridge University Press.
Cronbach, Lee J. 1951. "Coefficient Alpha and

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 385
the Internal Structure of Tests." Psycho- 87:125-35.
metrika. 16:297-334. Gurin, Patricia, Gerald Gurin, and Betty M.
Czaja, Sara J. and Joseph Sharit. 1998. ““Age Morrison. 1978. ““Personal and Ideological
Differences in Attitudes toward Computers.”” Aspects of Internal and External Control.””
The Journals of Gerontology 53:329-40. Social Psychology Quarterly 41:275-96.
Dambrot, Faye H, Michelle A. Watkins-Malek, Heinssen, Robert K. Jr., Carol R. Glass, and
S. Mark Silling, Rodeney S. Marshall, and Jo Luanne A. Knight. 1987. ““Assessing Com-
Ann Garver. 1985. ““Correlates of Sex Differ- puter Anxiety: Development and Validation of
ences in Attitudes toward and Involvement the Computer Anxiety Rating Scale.”” Com-
with Computers.”” Journal of Vocational Be- puters in Human Behavior 3:49-59.
havior. 27:71-86. Helmericks, Steven G. 1993. ““Collaborative
D’’Amico, Antonella and Maurizio Cardaci. Testing in Social Statistics: Toward Gemein-
2003. ““Relations among Perceived Self- stat.”” Teaching Sociology 21:287-97.
Efficacy, Self-Esteem, and School Achieve- Hembree, Ray 1990. ““The Nature, Effects, and
ment.”” Psychological Reports 92:745-54. Relief of Math Anxiety.”” Journal for Research
Department of Commerce. 2000. ““Falling in Mathematics Education 21:33-46.
through the Net: Toward Digital Inclusion.”” Horney, Karen. 1937. The Neurotic Personality
National Telecommunications and Information of Our Time. New York: Norton.
Administration. Retrieved September 1, 2006 Howard, Geoffry S. 1986. Computer Anxiety
(http://www.ntia.doc.gov/ntiahome/fttn00/con and the Use of Microcomputers in Manage-
tents00.html). ment. Ann Arbor, MI: UMI Research Press.
Fariña, Francisca, R. Arce, J. Sobral, and R. Jaffee, David. 1997. ““Asynchronous Learning:
Carames. 1991. ““Predictors of Anxiety to- Technology and Pedagogical Strategy in a
wards Computers.”” Computers in Human Distance Learning Course.”” Teaching Sociol-
Behavior 7:263-7. ogy 25:262-77.
Fennema, Elizabeth and Julia A. Sherman. 1976. Jennings, Susan Evans and Anthony J. Onwueg-
““Fennema-Sherman Mathematics Attitudes buzie. 2001. ““Computer Attitudes as a Func-
Scales: Instruments Designed to Measure Atti- tion of Age, Gender, Math Attitude, and De-
tudes toward the Learning of Mathematics by velopmental Status.”” Journal of Educational
Females and Males.”” Journal for Research in Computing Research 25:367-84.
Mathematics Education 7:324-6. Johnson, David W. and Roger T. Johnson. 1989.
Festinger, Leon. 1954. ““A Theory of Social Cooperation and Competition: Theory and
Comparison Processes.”” Human Relations Research. Edina, MN: Interaction Book Com-
7:117-40. pany.
Fitzgerald, Shawn M. and Stephen J. Jurs. 1996. Jones, Paul E. and Robert E. Wall. 1990.
““A Model Predicting Statistics Achievement ““Components of Computer Anxiety.”” Journal
among Graduate Students.”” College Student of Educational Technology Systems 18:161-8.
Journal 30:361-6. Joo, Young-Ju, Mimi Bong, and Ha-Jeen Choi.
Gecas, Viktor and Michael L. Schwalbe. 1983. 2000. ““Self-Efficacy for Self-Regulated Learn-
““Beyond the Looking-Glass Self: Social Struc- ing, Academic Self-Efficacy, and Internet
ture and Efficacy-Based Self Esteem.”” Social Self-Efficacy in Web-Based Instruction.”” Edu-
Psychology Quarterly 46:77-88. cational Technology Research and Develop-
Glass, Carol R. and Luanne A. Knight. 1988. ment 48:5-17.
““Cognitive Factors in Computer Anxiety.”” Keegan, Desmond. 1986. The Foundations of
Cognitive Therapy and Research 12:351-66. Distance Education. London, England: Croom
Greene, Bernard and Anne Meek. 1998. Helm.
““Distance Education in Higher Education In- Khan, Badrul H. 1997. ““Web-Based Instruction:
stitutions: Incidence, Audiences, and Plans to What Is It and Why Is It?”” Pp. 5-18 in Web-
Expand.”” Issue Brief, Institute of Education Based Instruction, edited by B. H. Khan.
Sciences National Center for Education Statis- Englewood Cliffs, NJ: Educational Technol-
tics, U.S. Department of Education, Washing- ogy Publications.
ton, DC. Koohang, Alex A. 1986. ““Effects of Age, Gen-
Gressard, Clarice P. and Brenda H. Lloyd. der, College Status, and Computer Experience
1987. ““An Investigation of the Effects of on Attitudes toward Library Computer Sys-
Mathematics Anxiety and Sex on Computer tems (LCS).”” Library and Information Science
Attitudes.”” School Science and Mathematics Research 8:349-53.

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

386 TEACHING SOCIOLOGY
Krendl, Kathy A., Mary C. Broihier, and Cyn- Owston, Ronald D. 1997. ““The World-Wide-
thia Fleetwood. 1989. ““Children and Com- Web: A Technology to Enhance Teaching and
puters: Do Sex-Related Differences Persist?”” Learning.”” Educational Researcher 26:27-33.
Journal of Communication 39:85-93. Pearlin, Leonard I. and Carmi Schooler. 1978.
Lane, John and Andrew Lane. 2001. ““Self- ““The Structure of Coping.”” Journal of Health
Efficacy and Academic Performance.”” Social and Social Behavior 19:2-21.
Behavior and Personality 29:687-94. Persell, Caroline Hodges. 2004. ““Using Focused
Lefcourt, Herbert M. 1976. Locus of Control: Web-Based Discussions to Enhance Student
Current Trends in Theory and Research. New Engagement and Deep Understanding.””
York: Wiley. Teaching Sociology 32:61-78.
Levin, Tomar and Claire Gordon. 1989. ““Effects Pettigrew, Thomas F. 1967. ““Social Evaluation
of Gender and Computer Experience on Atti- Theory: Convergences and Applications.”” Pp.
tudes toward Computers.”” Journal of Educa- 241-311 in Nebraska Symposium on Motiva-
tional Computing Research 5:69-88. tion 1967, edited by David Levine. Lincoln,
Liu, Hope Q. 2003. "Development of an Online NE: University of Nebraska Press.
Course Using a Modified Version of Keller's Pugliesi, Karen. 1989. ““Social Support and Self-
Personalized System of Instruction." Ph.D. Esteem as Intervening Variables in the Rela-
dissertation, Teaching and Learning Depart- tionship between Social Roles and Women’’s
ment, Virginia Tech, Blacksburg, VA. Well-Being.”” Community Mental Health Jour-
Liu, Xiaoru, Howard B. Kaplan, and Will Ris- nal 25:87-100.
ser. 1992. ““Decomposing the Reciprocal Rela- Randhawa, Bikkar S., James E. Beamer, and
tionship between Academic Achievement and Ingvar Lundberg. 1993. ““Role of Mathematics
General Self-Esteem.”” Youth and Society Self-Efficacy in the Structural Model of
24:123-48. Mathematics Achievement.”” Journal of Edu-
Llabre, Maria M. and Eduardo Suarez. 1985. cational Psychology 1:41-8.
““Predicting Math Anxiety and Course Per- Rendulic, Paul and Steve Terrell. 2000.
formance in College Women and Men.”” Jour- ““Anxiety toward Statistics and the Use of
nal of Counseling Psychology 32:283-7. Computers: A Study of Graduate Level Educa-
Mead, George H. 1934. Mind, Self, and Society. tion Majors.”” Journal of Computing in Higher
Chicago, IL: University of Chicago Press. Education 11:104-20.
Mirowsky, John and Catherine E. Ross. 1989. Reyes, Laurie H. 1984. ““Affective Variables and
Social Causes of Psychological Distress. New Mathematics Education.”” The Elementary
York: Aldine de Gruyter. School Journal 84:558-81.
Moeller, Thomas G. 1994. ““What Research Says Rosenberg, Morris. 1965. Society and Adoles-
about Self-Esteem and Academic Perform- cent Self-Image. Middletown, CT: Wesleyan
ance.”” Education Digest 59:34-7. University Press.
Mone, Mark, Douglas D. Baker, and Frank _____. 1979. Conceiving the Self. New York:
Jeffries. 1995. ““Predictive Validity and Time Basic Books.
Dependency of Self-Efficacy, Self-Esteem, Rosenberg, Morris, Carmi Schooler, and Carrie
Personal Goals, and Academic Performance.”” Schoenbach. 1989. ““Self-Esteem and Adoles-
Educational and Psychological Measurement cent Problems: Modeling Reciprocal Effects.””
55:716-27. American Sociological Review 54:1004-18.
Monteith, Moria and Jan Smith. 2001. Rosenberg, Morris, Carmi Schooler, Carrie
““Learning in a Virtual Campus: The Peda- Schoenbach, and Florence Rosenberg. 1995.
gogical Implications of Students’’ Experi- ““Global Self-Esteem and Specific Self-Esteem:
ences.”” Innovations in Education and Teach- Different Concepts, Different Outcomes.””
ing International 38:119-32. American Sociological Review 60:141-56.
Nicol, David J., Ian Minty, and Christine Sin- Rosenfield, Sarah. 1999. ““Splitting the Differ-
clair. 2003. ““The Social Dimensions of Online ence: Gender, the Self, and Mental Health.””
Learning.”” Innovations in Education and Pp. 209-24 in Handbook of the Sociology of
Teaching International 40:270-80. Mental Health, edited by Carol S. Aneshensel
Owens, Timothy J. 1994. ““Two Dimensions of and Jo C. Phelan. New York: Plenum.
Self-Esteem: Reciprocal Effects of Positive Ross, Catherine E. and Beckett A. Broh. 2000.
Self-Worth and Self-Deprecation on Adoles- ““The Roles of Self-Esteem and the Sense of
cent Problems.”” American Sociological Re- Personal Control in the Academic Achieve-
view 59:391-407. ment Process.”” Sociology of Education

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

EFFECTS OF WEB-BASED INSTRUCTION 387
73:270-84. dial Mathematics Classes.”” Journal of Negro
Ross, Catherine E. and John Mirowsky. 1989. Education 59:320-35.
““Explaining Social Patterns of Depression: Thoits, Peggy. 1995. ““Stress, Coping, and So-
Control and Problem-Solving——or Support and cial Support: Where Are We? What Next?””
Talking.”” Journal of Health and Social Behav- Journal of Health and Social Behavior Extra
ior 30:206-19. Issue:53-79.
Ross, Catherine E. and Jaya Sastry. 1999. ““The Tobias, Sheila. 1980. ““Math Anxiety: What
Sense of Personal Control: Social-Structural You Can Do About It.”” Today’’s Education
Causes and Emotional Consequences.”” Pp. 69:26-9.
369-94 in Handbook of the Sociology of Men- _____. 1995. Overcoming Math Anxiety. New
tal Health, edited by Carol S. Aneshensel and York: Norton.
Jo C. Phelan. New York: Plenum. Townsend, Michael A., Dennis W. Moore,
Royse, David and Elizabeth Lewis Rompf. 1992. Bryan F. Tuck, and Keri M. Wilton. 1998.
““Math Anxiety: A Comparison of Social Work ““Self-Concept and Anxiety in University Stu-
and Non-Social Work Students.”” Journal of dents Studying Social Science Statistics Within
Social Work Education 28: 270-7. a Co-operative Learning Structure.”” Educa-
Sax, Linda J. 1994. ““Mathematical Self- tional Psychology 18:41-54.
Concept: How College Reinforces the Gender Turner, R. Jay, John Taylor, and Karen Van
Gap.”” Research in Higher Education 35:141- Gundy. 2004. ““Personal Resources and De-
66. pression in the Transition to Adulthood: Ra-
Schacht, Steven and Brad J. Stewart. 1990. cial/Ethnic Comparisons.”” Journal of Health
““What’’s So Funny about Statistics? A Tech- and Social Behavior 45:34-52.
nique for Reducing Student Anxiety.”” Teach- Wagenaar, Theodore C. 2004. ““Is There a Core
ing Sociology 18:52-6. in Sociology? Results from a Survey.”” Teach-
Schutte, Jerald G. 1996. ““Virtual Teaching in ing Sociology 32:1-18.
Higher Education: The New Intellectual Su- Wang, Li-ya, Edward Kick, James Fraser, and
per-Highway or Just Another Traffic Jam?”” Thomas Jerome Burns. 1999. ““Status Attain-
California State University Northridge Depart- ment in America: The Roles of Locus of Con-
ment of Sociology. Retrieved September 1, trol and Self-Esteem in Educational and Occu-
2006 (http://www.csun.edu/sociology/virexp.h pational Outcomes.”” Sociological Spectrum
tm). 19:281-98.
Shashaani, Lily 1995. ““Gender Differences in Ward, Eric 1984. ““Statistics Mastery: A Novel
Mathematics Experience and Attitude and Approach.”” Teaching of Psychology 11:223-5.
Their Relation to Computer Attitude.”” Educa- Wheaton, Blair. 1983. ““Stress, Personal Coping
tional Technology 35:32-8. Resources, and Psychiatric Symptoms.”” Jour-
Shaw, Peter. 1994. ““Education Bulletin: Self- nal of Health and Social Behavior 27:78-89.
Esteem Rises to All-time High; Test Scores White, Robert W. 1959. ““Motivation Reconsid-
Hit New Lows.”” Antioch Review 52:467-74. ered: The Concept of Competence.”” Psycho-
Slavin, Robert E. 1990. Cooperative Learning: logical Review 66:297-333.
Theory, Research, and Practice. Boston, MA: Wilhite, Stephen C. 1990. ““Self-Efficacy, Locus
Allyn and Bacon. of Control, Self-Assessment of Memory Abil-
Smith, Brien N. and James R. Necessary. 1996. ity, and Study Activities as Predictors of Col-
““Assessing the Computer Literacy of Under- lege Course Achievement.”” Journal of Educa-
graduate College Students.”” Education tional Psychology 82: 696-700.
117:188-93. Winiecki, Donald J. 1999. ““Keeping the Thread:
Stevens, Phyllis. 1982. ““Computer Programs to Adapting Conversational Practice to Help
Reduce Math Anxiety.”” ERIC No. ED222328, Distance Students and Instructors Manage
ERIC Science, Mathematics, and Environ- Discussions in an Asynchronous Learning
mental Education Clearinghouse, U.S. Depart- Network.”” The America Center for the Study
ment of Education, Washington, DC. of Distance Education. Retrieved September
Suinn, Richard. M. 1969. ““The STABS, a 1, 2006 (http://www.ed.psu.edu/acsde/deos/de
Measure of Test Anxiety for Behavioral Ther- osnews/deosnews9_2.asp).
apy: Normative Data.”” Behavior Research Zanakis, Stelios H. and Enzo R. Valenzi. 1997.
Therapy 1:303-11. ““Student Anxiety and Attitudes in Business
Tate Green, LuEthel. 1990. ““Test Anxiety, Statistics.”” Journal of Education for Business
Mathematics Anxiety, and Teacher Com- 73:10-6.
ments: Relationships to Achievement in Reme-

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011

388 TEACHING SOCIOLOGY
Karen Van Gundy is assistant professor of sociol- research associate in the Federal Statistics Program at
ogy at the University of New Hampshire. Her research the American Institutes for Research in Washington,
and teaching specialties include the sociology of mental D.C.
health, substance use, and statistics. Currently, she is
Co-Principal Investigator for a grant funded by the Hope Q. Liu earned her doctorate in curriculum
National Science Foundation. Recent publications and instruction from Virginia Tech. Her areas of ex-
appear in Social Science & Medicine, the Journal of pertise include instructional design, distance learning,
Health and Social Behavior, and Social Psychology and evaluation. Currently, she is the distance learning
Quarterly. project manager for the Children’’s Health Education
Center in Milwaukee, Wisconsin.
Beth A. Morton earned her M.A. in sociology from
the University of New Hampshire. Her areas of inter-
est include education policy, at-risk students, and ele- Jennifer Kline earned her M.A. in sociology from
mentary/secondary teacher quality. Currently, she is a the University of New Hampshire.

Downloaded from tso.sagepub.com at UNIV OF NEW HAMPSHIRE on January 24, 2011