573169

research-article2015
SGOXXX10.1177/2158244015573169SAGE OpenLepp et al.

Article

SAGE Open

The Relationship Between Cell Phone Use

January-March 2015: 1­–9
© The Author(s) 2015
DOI: 10.1177/2158244015573169
and Academic Performance in a Sample sgo.sagepub.com

of U.S. College Students

Andrew Lepp1, Jacob E. Barkley1, and Aryn C. Karpinski1

Abstract
The cell phone is ever-present on college campuses and is frequently used in settings where learning occurs. This study
assessed the relationship between cell phone use and actual college grade point average (GPA) after controlling for known
predictors. As such, 536 undergraduate students from 82 self-reported majors at a large, public university were sampled. A
hierarchical regression (R2 = .449) demonstrated that cell phone use was significantly (p < .001) and negatively (β = −.164)
related to actual college GPA after controlling for demographic variables, self-efficacy for self-regulated learning, self-efficacy
for academic achievement, and actual high school GPA, which were all significant predictors (p < .05). Thus, after controlling
for other established predictors, increased cell phone use was associated with decreased academic performance. Although
more research is needed to identify the underlying mechanisms, findings suggest a need to sensitize students and educators
about the potential academic risks associated with high-frequency cell phone use.

Keywords
education, technology, ICT, mobile phones, smartphones, GPA, self-efficacy

Introduction In support of the “cell phone as disrupter” hypothesis, a

recent study by our group (Lepp et al., 2013) found that cell
Cell phones are an integral part of college life and culture. phone use was negatively associated with an objective mea-
Even a casual observation of today’s college students will sure of cardiorespiratory fitness in a sample of typical U.S.
reveal cell phones being used, both overtly and covertly, in college students. Interview data collected for the study
every possible campus setting, including the classroom. explained the negative relationship by suggesting that cell
Research suggests that college students frequently use the phone use disrupts physical activity and encourages seden-
cell phone during class time despite rules against doing so tary behavior. Unpublished interview data collected as part
(Tindell & Bohlander, 2012). As cell phone technology con- of the same study suggest that cell phone use may also dis-
tinues its rapid development, the device appears capable of rupt behaviors conducive to academic success. For example,
contributing to student learning and improved academic per- when asked to describe cell phone use habits, one participant
formance. For example, modern “smartphones” provide stu- stated, “I usually go on my phone if I’m bored sitting there in
dents with immediate, portable access to many of the same class. Or during homework I’ll take little Twitter breaks.”
education-enhancing capabilities as an Internet-connected Another student said,
computer, such as online information retrieval, file sharing,
and interacting with professors and fellow students (Bull & If I’m in class and I’m bored then I’ll use my phone to look on
McCormick, 2012; Tao & Yeh, 2013). Conversely, recent Facebook. I think it’s just kind of a habit now that I have, which
research suggests that many college students perceive the probably isn’t a good one. But, it’s just that I always have it [the
cell phone primarily as a leisure device, and most commonly phone] on me.
use cell phones for social networking, surfing the Internet,
watching videos, and playing games (Lepp, Li, & Barkley,
2015; Lepp, Barkley, Sanders, Rebold, & Gates, 2013). If
typically utilized for leisure rather than education, then cell 1
Kent State University, OH, USA
phones may disrupt learning within academic settings
Corresponding Author:
(Levine, Waite, & Bowman, 2007). Thus, the potential rela- Andrew Lepp, College of Education, Health and Human Services, Kent
tionship between cell phone use and academic performance State University, Kent, OH, 44242-000, USA.
is not clear. Email: [email protected]

Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 3.0 License
(http://www.creativecommons.org/licenses/by/3.0/) which permits any use, reproduction and distribution of the work without further
permission provided the original work is attributed as specified on the SAGE and Open Access page (http://www.uk.sagepub.com/aboutus/openaccess.htm).
2 SAGE Open

Across the interviews, such statements were more com- Weiss, 2011). In particular, Kirschner and Karpinski (2010)
mon among high-frequency cell phone users than among demonstrated that Facebook users have a lower self-reported
low-frequency users. These statements suggest that some GPA and spend fewer hours per week studying than nonus-
students, particularly high-frequency users, may have diffi- ers. Likewise, Junco (2012a, 2012b) found a strong, negative
culty regulating their cell phone use during academic endeav- relationship between time spent on Facebook and actual
ors such as class participation, homework, and studying. cumulative GPA. These negative relationships have been
Thus, the purpose of the present study was to investigate the found in populations across the world, including North
relationship between cell phone use and academic perfor- America, Europe, and Asia (e.g., Chen & Tzeng, 2010;
mance in a large sample of U.S. college students. Karpinski, Kirschner, Ozer, Mellott, & Ochwo, 2013).
Recently, multitasking has emerged as a possible explana-
tion for the negative relationship between electronic media
Literature Review
use (including cell phone use) and academic performance
Although the cell phone is likely to be on hand while college (Jacobsen & Forste, 2011; Junco & Cotton, 2011; 2012;
students are in class and studying, research investigating its Karpinski et al., 2013; Kirschner & Karpinski, 2010; Rosen
relationship to academic performance is limited. In an early et al.,2013; Wood et al., 2012). Indeed, several studies reveal
study of the phenomenon, Sánchez-Martínez and Otero that students frequently report using a variety of electronic
(2009) used a combination of self-reported monthly cell media including cell phones while in class, studying, and
phone expenses and frequency of use data to identify inten- doing homework (Jacobsen & Forste, 2011; Junco & Cotton,
sive cell phone users in a large sample of Spanish high school 2012; Sánchez-Martínez & Otero, 2009; Tindell & Bohlander,
students. In the study, intensive cell phone use was related to 2012). Several recent studies, using a variety of methods,
school failure as well as other negative behaviors such as identify a negative relationship between multitasking and
smoking and excessive alcohol use. More recent studies academic performance. First, Wood et al. (2012) measured
operationalize cell phone use as calling and texting while uti- the influence of multitasking with an array of electronic
lizing a variety of measures for academic performance. For media on students’ ability to learn from typical, university
example, Jacobsen and Forste (2011) identified a negative classroom lectures. Emailing, MSN messaging, and
relationship between calling, texting, and self-reported grade Facebook use via computer were all investigated as was cell
point average (GPA) among university students in the United phone texting. Results showed that multitasking with any of
States. Similarly, Hong, Chiu, and Hong (2012) found that the technologies was associated with lower scores on follow-
calling and texting were positively correlated with a self- up tests compared with students who did not multitask.
reported measure of academic difficulty among a sample of Second, Junco and Cotton (2012) used a hierarchical regres-
female, Taiwanese university students. While these studies sion to determine the power of multitasking to predict actual
provide a starting point for understanding the relationship cumulative college GPA. Results showed that Facebook-
between cell phone use and academic performance, they nei- multitasking and texting-multitasking were significantly and
ther use objective measures of academic performance nor do negatively related to college GPA after controlling for sex,
they take into account the cell phone’s expanding capabili- actual high school GPA, time preparing for class, and a stu-
ties beyond calling and texting. dent’s Internet skills. Finally, Rosen et al. (2013) observed
Modern cell phones enable users to access a variety of the study behaviors as well as study settings of a sample of
electronic media at almost any time and any place. Popular middle school, high school, and university students.
activities such as playing video games, surfing the Internet, Participants were observed for 15 min with on-task and off-
and monitoring social media sites are now all easily accom- task behavior recorded every minute. Results showed that
plished with most cell phones. Researchers have linked each participants typically became distracted by media such as
of these activities, independent of cell phone use, to aca- Facebook and texting after less than 6 min of studying.
demic performance. For example, heavy video game playing Furthermore, measurements of daily Facebook use and daily
has been associated with lower GPAs (Jackson, von Eye, texting behavior predicted off-task behavior during study
Fitzgerald, Witt, & Zhao, 2011; Jackson, von Eye, Witt, periods as well as self-reported GPA.
Zhao, & Fitzgerald, 2011). Also, low levels of Internet use In review, emerging research suggests that texting,
have been associated with improved academic performance Internet use, email, and social-networking sites such as
(Chen & Peng, 2008). Chen and Tzeng (2010) found that Facebook can potentially increase multitasking and task-
among heavy Internet users information seeking was associ- switching during academic activities and decrease academic
ated with better academic performance, while video game performance. Notably, all of these previously investigated
playing was associated with lower levels of academic perfor- activities can now be accomplished with a single, Internet-
mance. Several recent studies have identified a negative rela- connected cell phone. Therefore, measurements of cell phone
tionship between social-networking site use (e.g., Facebook, use should not be limited to only texting and calling but
MySpace, Twitter) and academic performance (e.g., Rosen, should take this wide array of activities into account.
Carrier, & Cheever, 2013; Stollak, Vandenberg, Burklund, & Furthermore, and in consideration of the ubiquity of the cell
Lepp et al. 3

phone, the relationship between this expanded definition of related construct is SE:AA, which describes an individual’s
cell phone use and academic performance warrants belief in his or her capabilities to learn material from specific
investigation. content areas such as math, science, and history. As origi-
nally conceived and validated by Zimmerman et al. (1992),
SE:SRL influences SE:AA, which in turn influences final
Self-Efficacy Beliefs and Academic Performance
academic achievement. As predicted by the original model
In addition to improving the way cell phone use is measured, and subsequently verified, previous academic performance
a better understanding of the relationship between cell phone can influence both SE:SRL and SE:AA (Caprara, Vecchione,
use and academic performance requires incorporating addi- Alessandri, Gerbino, & Barbaranelli, 2011).
tional, well-established predictors into any statistical models
designed to assess this relationship. An abundance of
Research Question
research suggests that self-efficacy beliefs are among the
strongest predictors of academic performance (for a compre- Considering the existing research, as well as the unpublished
hensive review, see Pajares, 1996). Generally speaking, self- interview data presented in the introduction of this article, it
efficacy describes an individual’s belief in his or her is hypothesized that cell phone use and academic perfor-
capabilities to organize and execute the behaviors necessary mance are related. However, in assessing this relationship,
for success; as such, self-efficacy beliefs are a key mecha- there is a need to consider important statistical controls such
nism in human agency (Bandura, 1982). Self-efficacy beliefs as SE:SRL, SE:AA, and previous academic performance
are domain specific; thus, research has identified self-effi- (i.e., high school GPA). Similarly, research suggests that
cacy beliefs pertinent to academic performance (Pajares, choice in academic major, as well as demographic and
1996). The strength of academic self-efficacy constructs is behavioral factors, may also be predictive of academic per-
their influence over behavior. Students who report high aca- formance and should, therefore, be considered. This study
demic self-efficacy apply greater effort to academic pursuits, considered four such factors: sex, cigarette smoking, class
are more persistent in the face of obstacles, and exhibit a standing, and undergraduate major. Indeed, there are well-
greater interest in learning (Schunk, 1984, 1989). In addi- established sex-related differences in college students’ aca-
tion, research illustrates that academic self-efficacy can demic performance (Peter & Horn, 2005). Likewise, cigarette
mediate the effects of academic ability (Pajares, 1996). As a smoking has been associated with problematic cell phone
result, academic self-efficacy is positively correlated with use and poor academic performance (DeBerard, Spielmans,
virtually all measures of academic performance, including & Julka, 2004; Sánchez-Martínez & Otero, 2009). Class
semester grades, cumulative GPA, homework, test scores, standing and undergraduate major may also be potential pre-
and writing assignments (Multon, Brown, & Lent, 1991; dictors (Kirschner & Karpinski, 2010; Sulaiman & Mohezar,
Pajares, 1996). 2006). In addition, there is a need to operationalize cell
Research has demonstrated that efficacy beliefs are often phone use more broadly (i.e., assess total cell phone use) in
better predictors of academic performance than other com- consideration of the device’s increased functionality. Finally,
monly used social-psychological variables (e.g., Klomegah, there is a need to use objective measures of academic perfor-
2007; Paulsen & Gentry, 1995; Pintrich & Schunk, 2002). mance such as students’ official cumulative GPA. This study
For example, self-efficacy proved to be the strongest predic- fulfills these many needs by answering the following ques-
tor of college student’s academic performance in a model tion: What is the relationship between total cell phone use
including task value, goal orientations, metacognitive self- (i.e., calling, texting, video games, social networking, surf-
regulation, self-regulation, and learning strategies (Al-Harthy ing the Internet, software-based applications, etc.) and aca-
& Was, 2010). Two self-efficacy constructs in particular demic performance (i.e., actual college GPA) after controlling
have received much attention for their ability to predict aca- for previously identified predictors of academic performance
demic performance (Pajares, 1996). These are self-efficacy (i.e., actual high school GPA, SE:SRL, SE:AA, sex, cigarette
for self-regulated learning (SE:SRL) and self-efficacy for use, class standing, and academic major)?
academic achievement (SE:AA; Zimmerman, Bandura, &
Martinez-Pons, 1992). SE:SRL concerns an individual’s
Method
belief in his or her capabilities to proactively regulate his or
her learning on the path to academic achievement. This The dependent variable for this study, academic perfor-
includes belief in one’s ability to resist distractions while mance, was objectively assessed using participants’ actual
learning and to create study environments conducive to cumulative college GPA. In addition, actual high school GPA
learning. As such, it is an important variable to consider was used as a statistical control. Because these are sensitive
when exploring the relationship between potential distractors data, and collecting them involves accessing participants’
such as cell phones or other new media and academic perfor- official academic records, participants were assured that data
mance (LaRose & Eastin, 2004; LaRose, Lin, & Eastin, collection, storage, and reporting would guarantee confiden-
2003; LaRose, Mastro, & Eastin, 2001; Odaci, 2011). A tiality and anonymity. Participants were recruited during
4 SAGE Open

class time from courses that typically attract students from a As accurately as possible, please estimate the total amount of
diversity of undergraduate majors. Representative courses time you spend using your mobile phone each day. Please
include introduction to sociology, general biology, American consider all uses except listening to music. For example:
politics, human nutrition, and world history. During class consider calling, texting, sending photos, gaming, surfing the
Internet, watching videos, Facebook, email, and all other uses
time, the principal investigators explained the methods to all
driven by “apps” and software.
students present, answered questions, addressed concerns,
and ensured that the informed consent document was read,
Participants provided best estimates for hours of cell
understood, and signed. After this, a survey was distributed
phone use per day and minutes per day. Total use in minutes
and completed during class by all students who consented to
was calculated for each participant as hours × 60 + minutes.
participate in the study. On the survey, students provided
In developing this measure of total cell phone use, two focus
their university email address, which was later used to access
groups of undergraduate students reviewed the question for
their academic records. If students did not consent to have
content validity criteria, including (a) clarity in wording, (b)
their GPA retrieved, they did not participate in the study. This
relevance of the items, (c) use of standard English, (d)
method produced an initial sample size of 536 undergraduate
absence of biased words and phrases, (e) formatting of items,
students from 82 self-reported majors.
and (f) clarity of the instructions (Fowler, 2002). Most stu-
dents provided feedback from the criteria categories of (a),
(b), (c), and (f). Appropriate alterations were made to the sur-
Measures
vey based upon the responses and suggestions. In consider-
The survey took approximately 10 min to complete. Students ation of this measure’s construct validity, participants’ daily
first provided basic demographic and lifestyle information. text messaging and daily calling were assessed as this is how
Students completed the validated SE:SRL (Zimmerman et cell phone use has been operationalized in previous research
al., 1992) and SE:AA scales (Zimmerman et al., 1992). (e.g., Jacobsen & Forste, 2011). Total daily cell phone use
Participants also provided information regarding their cell (calling, texting plus all other uses such as Internet browsing
phone use as operationalized by Lepp et al. (2013) and, and games) was positively correlated with daily texting (r =
finally, their email addresses. Email addresses were used to .430, p < .001) and daily calling (r = .210, p < .001), suggest-
access each student’s official academic records from which ing that the measures are related but not identical. In addi-
college and high school GPAs were collected. tion, we assessed construct validity in a small group (N = 21)
SE:SRL is an 11-item scale that measures how well stu- of undergraduate college students at the same university
dents believe that they can use a variety of self-regulated from which the present sample was culled. Self-reported
learning strategies such as finish homework assignments by total cell phone use (minutes) as assessed by this measure
deadlines, study when there are other interesting things to do, had a large, significant correlation (r = .510, p = .018) to
concentrate on school subjects, and arrange a place to study objectively measured cell phone use (minutes) obtained by
without distractions (Zimmerman et al., 1992, p. 668). accessing students’ actual cell phone records (unpublished
SE:AA is a nine-item scale that measures how well students data). Thus, this self-report measure was carefully developed
believe that they can achieve success in important academic to assure content validity, while subsequent testing provided
domains such as reading, writing, English grammar, mathe- evidence of construct and criterion validity.
matics, science, social studies, and computer use. For the
items in both self-efficacy measures, students used a seven-
point Likert-type scale to rate their perceived capability to do
Data Analysis
well (i.e., 1 = not too well to 7 = very well). Responses for the All analyses were performed using SPSS for Windows
items in each scale were summed, thereby producing a total (Version 18.0, SPSS Inc, Evenston, Illinois). First, indepen-
score. Higher scores indicate greater self-efficacy. Both dent samples t tests were used to examine differences in GPA
scales have been previously validated and found to have between males and females and smokers and nonsmokers.
strong internal consistency (coefficient α = .87 and .70, Likewise, ANOVA was used to examine differences in GPA
respectively; Zimmerman et al., 1992). Since their develop- between class (i.e., freshman, sophomore, junior, senior) and
ment, both have been consistently shown to be reliable pre- a categorization of students based on the college that houses
dictors of academic performance in variety settings (Pajares, their major (i.e., education, health, and human services; arts
1996). Likewise, the SE:SRL and SE:AA scales demon- and sciences; business and communications). Second,
strated strong internal consistency with this study’s sample Pearson’s correlations were performed to examine the rela-
of undergraduate students (coefficient α = .84 and .73, tionships between the following variables: college GPA,
respectively; N = 536). SE:SRL, SE:AA, high school GPA, and total cell phone use.
Total daily cell phone use was measured using the follow- Third, hierarchical regression was used to answer this study’s
ing item: central research question:
Lepp et al. 5

Research Question 1: What is the relationship between Table 1.  Descriptive Statistics.
total cell phone use and academic performance after con-
N M SD
trolling for known predictors? Toward this end, the fol-
lowing model was initially proposed: College GPA 518 3.03 0.601
High school GPA 483 3.22 0.473
College GPA = sex , smoker, SE:SRL 518 56.42 8.96
class standing, college major SE:AA 518 44.44 7.07
Cell phone use 518 300.55 243.52
( Block 1)
+ SE : AA ( Block 2 ) + Note. GPA = grade point average; SE:SRL = self-efficacy for self-regulated
learning; SE:AA = self-efficacy for academic achievement.
SE : SRL ( Block 3) +
high school GPA ( Block 4 ) 69% of the data set (n = 360), which is greater than the per-
+ total daily cell phone use ( Block 5) . centage of females (59%) in the overall undergraduate stu-
dent body of the University.
The categorical variables of interest were assessed in the From this data set, the assumptions of regression were
first block of this model: sex, cigarette smoking, class, and examined, and a preliminary analysis was performed to
college. Blocks 2 to 4 in this model are identical to the model assess the linearity of the relationship between the study’s
developed by Zimmerman et al. (1992) and supported by independent continuous variables (SE:SRL, SE:AA, high
others (e.g., Caprara et al., 2011) to predict academic perfor- school GPA, total cell phone use) and college GPA. Using a
mance. Block 5 added cell phone use to the model and Lack of Fit Test, the assumption of linearity was upheld (p =
thereby tested whether or not daily cell phone use uniquely .906). The assumptions of normality and homoskedasticity
predicted college academic performance (GPA) after con- were also met using residual scatterplots.
trolling for these other, previously established variables. On average, students reported spending 300 min per day
Finally, to further illustrate the relationship between cell using their cell phones (SD = 243). The sample’s mean GPA
phone use and GPA, a tertile split for cell phone use was was 3.03 (SD = 0.60). Independent sample t tests demon-
performed. Students in this final sample (N = 518) were strated significant differences between males and females
divided into the following groups: low cell phone use group (p < .001) and smokers and nonsmokers (p < .001). Females’
(M = 94.6 min per day, SD = 41.0, n = 180), moderate use GPA (M = 3.09, SD = 0.63) was significantly higher than
group (M = 235.1 min per day, SD = 45.2, n = 173), and high males’ (M = 2.88, SD = 0.62), and nonsmokers’ GPA (M =
use group (M = 601.3 min per day, SD = 226.8, n = 164). An 3.07, SD = 0.64, n = 432) was significantly higher than
ANOVA was then utilized to compare mean GPA across the smokers’ (M = 2.80, SD = 0.58, n = 85). An ANOVA demon-
three cell phone use groups (high, moderate, low). Post hoc t strated significant differences in mean GPA between the four
tests were performed for any significant main effect. classes (p < .001). Freshmen had a mean GPA of 3.21 (SD =
0.67), sophomores had a mean GPA of 2.93 (SD = 0.64),
juniors had a mean GPA of 3.02 (SD = 0.55), and seniors had
Results a mean GPA of 2.94 (SD = 0.48). Finally, the 82 self-reported
majors were categorized into three groups based on the col-
Assumption Checking, Descriptive Statistics, and lege housing the major (education, health, and human ser-
Preliminary Analyses vices; arts and sciences; business and communications). An
Before conducting any descriptive or inferential statistics, an ANOVA found no significant difference in mean GPA
examination of outliers (i.e., cell phone use, GPA, age, between these three groups (p = .081). Thus, this variable
SE:SRL, SE:AA) was conducted. Following the method of was not included in further analysis.
Rosen et al. (2013), total cell phone use values that were Table 1 provides descriptive statistics for the continuous
more than 3 standard deviations from the mean were trun- variables used in this model. Table 2 illustrates the results of
cated to exactly 3 standard deviations from the mean. This Pearson’s correlations. There are several significant correla-
procedure was applied to measures of total cell phone use for tions worth noting. There was a significant, negative correla-
seven participants. Outliers on any of the remaining vari- tion between cell phone use and college GPA (p < .001). There
ables were removed from the study. This procedure resulted was a significant, positive correlation between both measures
in 18 cases being removed and yielded a final analysis sam- of self-efficacy (SE:SRL, SE:AA) and college GPA (p < .001).
ple of 518 students. The age range of the data set was 18 to There was a significant, negative correlation between both
28, with a mean of 20.28 (SD = 1.78). The data set was measures of self-efficacy (SE:SRL, SE:AA) and cell phone
evenly distributed by class (freshmen = 132, sophomores = use (p ≤ .041). Finally, high school GPA was significantly and
139, juniors = 134, and seniors = 113). Females comprised positively correlated with college GPA (p < .001).
6 SAGE Open

Table 2.  Pearson Correlation Coefficients (r). Table 3.  Hierarchical Regression Predicting College GPA: Model
Summary.
College High school
Sex/class/smoke SE:AA SE:SRL HS GPA CP use
  GPA GPA SE:SRL SE:AA
  Block 1 Block 2 Block 3 Block 4 Block 5
High school GPA .611***  
2
SE:SRL .341*** .242***   R .058 .101 .165 .425 .449
SE:AA .200*** .275*** .456***   ΔR2 .058 .043 .064 .259 .024
Cell phone use −.234*** −.168*** −.090* −.239*** ΔF 9.755 22.922 36.580 213.86 20.454
p .000 .000 .000 .000 .000
Note. GPA = grade point average; SE:SRL = self-efficacy for self-regulated
learning; SE:AA = self-efficacy for academic achievement. Note. GPA = grade point average; SE:SRL = self-efficacy for self-regulated
*p < .05. ***p < .001. learning; SE:AA = self-efficacy for academic achievement; HS = high
school; CP = cell phone.

Hierarchical Regression
The study’s aim was to assess the relationship between cell
As described above, the preliminary analysis supported test- phone use and academic performance after controlling for
ing the following hierarchical regression model: known predictors of academic performance. A hierarchical
College GPA = sex , smoker, regression was used for this purpose allowing for the devel-
opment of a model which used sex, cigarette smoking behav-
class standing ( Block 1) +
ior, class standing, SE:AA, SE:SRL, and high school GPA to
SE : AA ( Block 2 ) predict college GPA. Each of these variables were significant
predictors of college GPA. Females, as has been the recent
+ SE : SRL ( Block 3) +
trend, had higher GPAs than males (Peter & Horn, 2005).
high school GPA ( Block 4 ) + Smokers, as suggested in previous research, had lower GPAs
total daily cell phone use ( Block 5 ) . than nonsmokers (DeBerard et al., 2004; Sánchez-Martínez
& Otero, 2009). Class was a significant predictor as well,
Table 3 provides the model summary results for the hierar- with freshmen and juniors doing slightly better academically
chical regression predicting college GPA with total cell phone than sophomores and seniors in this sample. As expected,
use as the final block in the model. Each block significantly SE:SRL, SE:AA, and high school GPA were all positively
added to the prediction of the criterion variable. In Block 1, associated with GPA (Zimmerman et al., 1992). Finally, total
females had a significantly greater GPA than males (β = .120, cell phone use (min/day) was added to the end of this regres-
p = .007), nonsmokers had a significantly higher GPA than sion model. After controlling for the previously established
nonsmokers (β = .155, p = .001), and class standing proved predictors of academic performance, total cell phone use was
significant as well (β = −.111, p = .013). In Block 2, there was found to be a significant negative predictor of GPA. These
a significant, positive relationship between college GPA and results suggest that given two college students from the same
SE:AA (β = .210, p < .001). In Block 3, there was a signifi- university with the same class standing, same sex, same
cant, positive relationship between college GPA and SE:SRL smoking habits, same belief in their ability to self-regulate
(β = .289, p < .001). In Block 4, there was a significant, posi- their learning and do well academically, and same high
tive relationship between college GPA and high school GPA school GPA—the student who uses the cell phone more on a
(β = .553, p < .001). Finally, there was a significant, negative daily basis is likely to have a lower GPA than the student who
relationship between total daily cell phone use and college uses the cell phone less.
GPA (β = −.164, p < .001). This total model explained 44.9% Previous research suggests that college students’ cell
of the variance in college GPA (R2 = .449). phone use may be a distraction in academic settings (Levine
Finally, the ANOVA comparing GPA across the three cell et al., 2007). Two previous studies using large random sam-
phone use groups (low, moderate, high) revealed a significant ples of college students found that 89% (N = 302) and 83%
main effect (F = 11.70, df = 2, p < .001). Specifically, the high (N = 251) of the students surveyed perceived the cell phone
cell phone use group had a GPA (M = 2.84, SD = 0.61) that primarily as a leisure device rather than as an educational
was significantly lower (p < .001) than both the moderate use tool (Barkley & Lepp, 2013; Lepp et al., 2013). Because the
group (M = 3.06, SD = .61) and the low use group (M = 3.15, cell phone is ever-present and commonly used for leisure, it
SD = 0.45). There was not a statistically significant difference is likely that it occasionally distracts from learning in class,
between the low use and moderate use groups (p = .175). in the library, in the dormitories, and in any other setting utilized
by students for academic purposes. In addition, there is a grow-
ing amount of research that suggests electronic media in any
Discussion
form encourages multitasking (Jacobsen & Forste, 2011; Junco
This study was exploratory in nature. Therefore, the findings & Cotton, 2011, 2012; Karpinski et al., 2013; Kirschner &
are best understood as initial steps into a new line of inquiry. Karpinski, 2010; Wood et al., 2012) and task-switching
Lepp et al. 7

(Rosen et al., 2013), both of which are negatively related to at risk of over reporting use. In lieu of objective data, future
academic performance. studies may seek to further validate this measure.
Considering these explanations, it is likely that the mod- Furthermore, future studies should assess the time devoted to
ern cell phone creates a temptation to surf the Internet, check common specific uses such as social networking, gaming,
social media (e.g., Facebook), play video games, contact and information search, in addition to measuring overall use
friends, explore new applications, or engage with any num- as was done here. Second, the sample consisted of under-
ber of cell-phone-based leisure activities, which some stu- graduate college students from a single, large, public univer-
dents fail to resist when they should otherwise be focused on sity in the Midwestern United States. Although the behavioral
academics. As such, the negative relationship between cell norms governing cell phone use appear to be consistent
phone use and academic performance identified here could among today’s college students (Anderson & Rainie, 2011;
be attributed to students’ decreased attention while studying Tindell & Bohlander, 2012), attempts to generalize these
or a diminished amount of time dedicated to uninterrupted results to other populations should be made with caution.
studying. Indeed, a similar argument has been proffered to Therefore, future research should include college students
explain the negative relationship between general social- from different types of universities and from different geo-
networking site use or Facebook use and academic perfor- graphic regions. In addition, high school and junior high
mance (Karpinski et al., 2013; Kirschner & Karpinski, 2010). school students should be studied as recent research suggests
Future research should examine the many potential underly- that the relationships identified here may be evident in
ing reasons for the negative relationship identified here, younger students as well (Rosen et al., 2013).
including time spent studying and multitasking. Of course,
this line of research has demonstrated only relationships and Conclusion
not causality. Thus, there is a need to explore these relation-
ships over time and with experimental designs. This research utilized a more holistic measure of cell phone
There is also a need to better understand how specific cell use than previous studies. The measure accounts for the cell
phone uses are related to academic performance. While this phone’s expanded capabilities in the realm of social network-
study found that cell phone use as a whole was negatively ing, gaming, and Internet use. After controlling for SE:SRL,
associated with academic performance, the relationship may SE:AA, and other important predictors such as actual high
vary with particular uses. In other words, contrary to the school GPA, this measure of cell phone use was a significant
findings presented here, there may be specific uses that are and negative predictor of college students’ academic perfor-
positively related to academic performance. For example, mance, objectively measured as cumulative GPA. Presently,
Norris (1996) found that while TV watching as a whole was cell phone use is a dominant and defining characteristic of
negatively associated with political participation, watching this generation of college students and often occurs during
TV news and public affairs programming was positively class time, while completing homework, and while studying
associated with political participation. Likewise, Chen and (Smith, Raine, & Zickuhr, 2011; Tindell & Bohlander, 2012).
Tzeng (2010) found that using the Internet for information Therefore, more research is needed to better understand the
seeking was associated with better academic performance, mechanisms underlying this relationship. Even so, educators
while using the Internet for video game playing was associ- and administrators in higher education may wish to carefully
ated with lower levels of academic performance. Finally, consider policies regarding cell phone use in the classroom,
Junco (2012a) found that the total amount of time college laboratories, and other settings where learning occurs.
students spend on Facebook, as well as the total number of Undoubtedly, the capability of the cell phone to entertain,
times students check Facebook, were negatively associated connect, and inform will continue its rapid development. As
with campus engagement. However, some Facebook activi- such, cell phones and related devices will only increase in
ties such as creating events and RSVPing for events were popularity and use. Therefore, there is a need to better under-
positively associated with campus engagement. Thus, assess- stand how this technology can be harnessed to make a genu-
ing cell phone use as a whole is likely to provide only a par- ine contribution to student learning. We may discover
tial understanding of an undoubtedly complex relationship. conditions where learning is enhanced by having the cell
Additional research assessing time devoted to specific cell phone on; likewise, we may discover conditions where learn-
phone uses such as gaming, social networking, information ing is enhanced by having it off.
search, and the use of educational software (apps) is needed.
While these findings build upon and extend previous Declaration of Conflicting Interests
research in this area, there are limitations. First, cell phone The author(s) declared no potential conflicts of interest with respect
use was self-reported. Although the self-report measure used to the research, authorship, and/or publication of this article.
in this study was carefully developed to assure content valid-
ity and a subsequent test provided evidence of criterion Funding
validity, research by Boase and Ling (2013) illustrates that The author(s) received no financial support for the research and/or
continuous, open-ended self-report cell phone measures are authorship of this article.
8 SAGE Open

References Junco, R. (2012a). The relationship between frequency of Facebook

use, participation in Facebook activities, and student engage-
Al-Harthy, I. S., & Was, C. A. (2010). Goals, efficacy and metacog-
ment. Computers & Education, 58, 162-171.
nitive self-regulation: A path analysis. International Journal of
Junco, R. (2012b). Too much face and not enough books: The
Education, 2(1), E2.
relationship between multiple indices of Facebook use and
Anderson, J., & Rainie, L. (2011). Millennials will benefit and
academic performance. Computers in Human Behavior, 28,
suffer due to their hyperconnected lives (The Pew Research
187-198.
Center’s Internet and American Life Project). Retrieved from
Junco, R., & Cotton, S. R. (2011). Perceived academic effects of
http://www.pewinternet.org/Reports/2012/Hyperconnected-
instant messaging use. Computers & Education, 56, 370-378.
lives/Overview.aspx
Junco, R., & Cotton, S. R. (2012). No A 4 U: The relationship
Bandura, A. (1982). Self-efficacy mechanism in human agency.
between multitasking and academic performance. Computers
American Psychologist, 37(2), 122-147.
& Education, 59, 505-514.
Barkley, J. E., & Lepp, A. (2013). Cellular telephone use is associ-
Karpinski, A. C., Kirschner, P. A., Ozer, I., Mellott, J. A., & Ochwo,
ated with greater sedentary behavior independent of leisure-
P. (2013). An exploration of social networking site use, mul-
time physical activity. Applied Physiology, Nutrition, and
titasking, and academic performance among United States and
Metabolism, 38(S1), 1023.
European university students. Computers in Human Behavior,
Boase, J. A., & Ling, R. (2013). Measuring mobile phone use:
29, 1182-1192.
Self-report versus log data. Journal of Computer Mediated
Kirschner, P. A., & Karpinski, A. C. (2010). Facebook and aca-
Communication, 18, 508-519.
demic performance. Computers in Human Behavior, 26, 1237-
Bull, P., & McCormick, C. (2012). Mobile learning: Integrating
1245.
text messaging into a community college pre-algebra course.
Klomegah, R. Y. (2007). Predictors of academic performance of
International Journal on E-Learning, 11, 233-245.
university students: An application of the goal efficacy model.
Caprara, G. V., Vecchione, M., Alessandri, G., Gerbino, M., &
College Student Journal, 41, 407-415.
Barbaranelli, C. (2011). The contribution of personality traits
LaRose, R., & Eastin, M. S. (2004). A social cognitive theory of
and self-efficacy beliefs to academic achievement: A longitu-
internet uses and gratifications: Towards a new model of media
dinal study. British Journal of Educational Psychology, 81,
attendance. Journal of Broadcasting & Electronic Media, 48,
78-96.
358-377.
Chen, S. Y., & Tzeng, J. Y. (2010). College female and male heavy
LaRose, R., Lin, C. A., & Eastin, M. S. (2003). Unregulated
Internet users’ profiles of practices and their academic grades
Internet usage: Addiction, habit, or deficient self-regulation?
and psychosocial adjustment. Cyberpsychology, Behavior, and
Media Psychology, 5, 225-253.
Social Networking, 13, 257-262. LaRose, R., Mastro, D., & Eastin, M. S. (2001). Understanding
Chen, Y. F., & Peng, S. S. (2008). University students’ Internet use Internet usage: A social-cognitive approach to uses and gratifi-
and its relationships with academic performance, interpersonal cations. Social Science Computer Review, 19, 395-413.
relationships, psychosocial adjustment, and self-evaluation. Lepp, A., Li, J., & Barkley, J. (2015). Exploring the relationships
CyberPsychology & Behavior., 11, 467-469. between college students’ cell phone use, personality and lei-
DeBerard, M. S., Spielmans, G. I., & Julka, D. C. (2004). Predictors sure. Computers in Human Behavior, 43: 210–219.
of academic achievement and retention among college fresh- Lepp, A., Barkley, J. E., Sanders, G. J., Rebold, M., & Gates, P.
men: A longitudinal study. College Student Journal, 38(1), (2013). The relationship between cell phone use, physical and
66-80. sedentary activity, and cardiorespiratory fitness in a sample
Fowler, F. J. Jr., (2002). Survey research methods (3rd ed.). of U.S. college students. International Journal of Behavioral
Thousand Oaks, CA: Sage Publications. Nutrition and Physical Activity, 10, Article 79. Retrieved from
Hong, F. Y., Chiu, S. I., & Hong, D. H. (2012). A model of the http://www.ijbnpa.org/content/10/1/79
relationship between psychological characteristics, mobile Levine, L. E., Waite, B. M., & Bowman, L. L. (2007). Electronic
phone addiction and use of mobile phones by Taiwanese uni- media use, reading, and academic distractibility in college
versity female students. Computers in Human Behavior, 28, youth. Cyber Psychology and Behavior, 10, 560-566.
2152-2159. Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of
Jackson, L. A., von Eye, A., Fitzgerald, H. E., Witt, E. A., & Zhao, self-efficacy beliefs to academic outcomes: A meta-analytic
Y. (2011). Internet use, videogame playing and cell phone investigation. Journal of Counseling Psychology, 1, 30-38.
use as predictors of children’s body mass index (BMI), body Norris, P. (1996). Does Television erode social capital? A reply to
weight, academic performance, and social and overall self- Putnam. Political Science & Politics, 29, 474-480.
esteem. Computers in Human Behavior, 27, 599-604. Odaci, H. (2011). Academic self-efficacy and academic procrasti-
Jackson, L. A., von Eye, A., Witt, E. A., Zhao, Y., & Fitzgerald, H. nation as predictors of problematic Internet use in university
E. (2011). A longitudinal study of the effects of Internet use students. Computers & Education, 57, 1109-1113.
and videogame playing on academic performance and the roles Pajares, F. (1996). Self-efficacy in academic settings. Review of
of gender, race and income in these relationships. Computers in Educational Research, 66, 543-578.
Human Behavior, 27, 228-239. Paulsen, M., & Gentry, J. (1995). Motivation, learning strategies,
Jacobsen, W. C., & Forste, R. (2011). The wired generation: and academic performance: A study of the college finance
Academic and social outcomes of electronic media use among classroom. Financial Practice & Education, 5(1), 78-89.
university students. Cyberpsychology, Behavior, and Social Peter, K., & Horn, L. (2005). Gender differences in participation
Networking, 14, 275-280. and completion of undergraduate education and how they have
Lepp et al. 9

changed over time (NCES 2005-169). Washington, DC: U.S. Tindell, D. R., & Bohlander, R. W. (2012). The use and abuse of
Government Printing Office, U.S. Department of Education, cell phones and text messaging in the classroom: A survey of
National Center for Education Statistics. college students. College Teaching, 60, 1-9.
Pintrich, P., & Schunk, D. (2002). Motivation in education theory: Wood, E., Zivcakova, L., Gentile, P., Archer, K., De Pasquale, D.,
Theory, research, and application (2nd ed.). Upper Saddle & Nosko, A. (2012). Examining the impact of off-task multi-
River, NJ: Prentice Hall. tasking with technology on real-time classroom learning.
Rosen, L. D., Carrier, M., & Cheever, N. A. (2013). Facebook and Computers & Education, 58, 365-374.
texting made me do it: Media-induced task-switching while Zimmerman, B. J., Bandura, A., & Martinez-Pons, M. (1992).
studying. Computers in Human Behavior, 29, 948-958. Self-motivation for academic attainment: The role of self-effi-
Sánchez-Martínez, M., & Otero, A. (2009). Factors associated with cacy beliefs and personal goal setting. American Educational
cell phone use in adolescents in the community of Madrid Research Journal, 29, 663-676.
(Spain). CyberPsychology and Behavior, 12, 131-137.
Schunk, D. H. (1984). The self-efficacy perspective on achieve-
ment behavior. Education Psychologist, 19, 119-218.
Author Biographies
Schunk, D. H. (1989). Social cognitive theory and self-regulated Andrew Lepp, PhD is an associate professor at Kent State
learning. In B. J. Zimmerman & D. H. Schunk (Eds.), Self- University in the College of Education, Health and Human
regulated learning and academic achievement: Theory, research Services (EHHS). He teaches in EHHS’s Recreation, Park and
and practice (pp. 83-110). New York, NY: Springer Verlag. Tourism Management program. His research interests include
Smith, A., Raine, L., & Zickuhr, K. (2011). College students “new media” (e.g. smartphone, internet) and young people’s
and technology (The Pew Research Center’s Internet and lives.
American Life Project). Retrieved from http://pewinternet.org/
Jacob E. Barkley, PhD is an associate professor of Exercise
Reports/2011/College-students-and-technology.aspx
Science in the College of Education Health and Human Services
Stollak, M. J., Vandenberg, A., Burklund, A., & Weiss, S. (2011).
at Kent State University. Dr. Barkley’s research focuses on
Getting social: The impact of social networking usage on
examining factors related to physical activity and sedentary
grades among college students. In Proceedings from ASBBS
behaviors.
Annual Conference (pp. 859-865). Retrieved from: http://
asbbs.org/files/2011/ASBBS2011v1/PDF/S/StollakM.pdf Aryn C. Karpinski, PhD is an assistant professor at Kent State
Sulaiman, A., & Mohezar, S. (2006). Student success factors: University in the College of Education, Health and Human Services
Identifying key predictors. Journal of Education for Business, (EHHS) in the Department of Foundations, Leadership, and
81, 328-333. Administration (FLA), in the Evaluation and Measurement pro-
Tao, Y., & Yeh, C. R. (2013). Transforming the personal response gram. She teaches research methods, measurement, and advanced
system to a cloud voting service. In S. Uesugi (Ed.), IT enabled statistics courses. Her research interests involve technology and
services (pp. 139-156). Verlag, Austria: Springer. social media in education.