Empirical Evidence That Text Messaging During Class Disrupts Comprehension
Topical Article
OMG! Texting in Class ¼ U Fail 🙁 Empirical Evidence That Text Messaging During Class Disrupts Comprehension
Amanda C. Gingerich1 and Tara T. Lineweaver1
Abstract In two experiments, we examined the effects of text messaging during lecture on comprehension of lecture material. Students (in Experiment 1) and randomly assigned participants (in Experiment 2) in a text message condition texted a prescribed con- versation while listening to a brief lecture. Students and participants in the no-text condition refrained from texting during the same lecture. Postlecture quiz scores confirmed the hypothesis that texting during lecture would disrupt comprehension and retention of lecture material. In both experiments, the no-text group significantly outscored the text group on the quiz and felt more confident about their performance. The classroom demonstration described in Experiment 1 provides preliminary empiri- cal evidence that texting during class disrupts comprehension in an actual classroom environment. Experiment 2 addressed the selection bias and demand characteristic issues present in Experiment 1 and replicated the main findings. Together, these two experiments clearly illustrate the detrimental effects of texting during class, which could discourage such behavior in students.
Keywords texting, distraction, comprehension, cognitive overload
According to Nielsen, U.S. wireless subscribers between the ages
of 18 and 24 sent an average of 790 text messages per month
between January 2006 and June 2008 (The Nielsen Company,
2008). That equates to more than one text message on average
sent each hour of every day over the entire month. Survey
responses of college students are even more impressive; 95% of students report bringing their phones with them to class every day
and 91% report using their phones to text message during class time (Mayk, 2010). Given that text messaging during class is so
common, the question of how dividing attention between lecture
and text messaging affects students’ comprehension and retention
of classroom material warrants investigation.
Although there is a relative abundance of research showing
the dangers of dividing one’s attention through text messaging
while driving (e.g., Drews, Yazdani, Godfrey, Cooper, &
Strayer, 2009; Hosking, Young, & Regan, 2009; Lee, 2007;
Strayer & Johnston, 2001), to our knowledge, no published
research has addressed the effects of text messaging in the
classroom on comprehension of lecture material. Several
researchers have demonstrated that intrusive noises such as a
cell phone ringing during cognitive tasks impair academic per-
formance (e.g., End, Worthman, Mathews, & Wetterau, 2010;
Hughes & Jones, 2001), but no one has published research
investigating the potentially detrimental effects of text messa-
ging during class on learning. This issue is especially pertinent
given the high prevalence of text messaging among college
students (Mayk, 2010).
Copious research has demonstrated the detrimental cogni-
tive effects of divided attention, although not conducted in a
classroom setting. Rubinstein, Meyer, and Evans (2001), for
example, found that people lost time as they switched from one
cognitive task to another; the amount of time they lost
increased as the task became more complex or unfamiliar.
Given that students rarely (if ever) focus on a lecture while text
messaging, task switching may be a more accurate description
of what texting students are doing in class. Other researchers
have demonstrated that divided attention impairs memory
particularly when attention is divided during the initial learning
and encoding of new information (Fernandes & Moscovitch,
2000). Thus, students trying to learn the typically new, com-
plex, and unfamiliar material introduced during lecture may
be particularly vulnerable to the divided attention associated
with text messaging.
In addition to the scarcity of empirical research addressing
the effects of text messaging on learning in the classroom,
we were motivated to conduct these experiments by the results
of a previous in-class demonstration in which we paired
1 Department of Psychology, Butler University, Indianapolis, IN, USA
Corresponding Author:
Amanda C. Gingerich, Department of Psychology, Butler University, 4600
Sunset Avenue Indianapolis, IN 46208, USA.
Email: mgingeri@butler.edu
Teaching of Psychology 2014, Vol 41(1) 44-51 ª The Author(s) 2013 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0098628313514177 top.sagepub.com
at ROCHESTER INST OF TECHNOLOGY on March 20, 2015top.sagepub.comDownloaded from
students with someone sitting near them. Half of the pairs com-
prised a ‘‘text-first’’ group, who completed an entire prescribed
text-messaging conversation before reading a research article
describing neurological evidence that divided attention modu-
lates the extent to which declarative memory or habit learning
contributes to solving a complex problem (Foerde, Knowlton,
Poldrack, & Smith, 2006). The remaining pairs comprised a
‘‘text-while-reading’’ group, who began reading the article
immediately and completed the same text-messaging conversa-
tion while they read. As such, the ‘‘text-first’’ group focused on
only one task at a time, while the ‘‘text-while-reading’’ group
engaged in frequent task switching. All students took a quiz
over the article as soon as they finished reading it.
After 15 min, 70% of the ‘‘text-first’’ group had started the quiz compared to only 40% in the ‘‘text-while-reading’’ group, despite the time delay that occurred before beginning the read-
ing in the ‘‘text-first’’ group. After another 10 min, 40% in the ‘‘text-first’’ group had finished the quiz, compared to only 20% in the ‘‘text-while-reading’’ group. Finally, scores on the quiz
revealed that the students who texted while simultaneously
reading the article earned an average of 3.6 points, which was
notably lower than the 5.6 points earned by the ‘‘text-first’’
group. Based on these findings, we designed a pair of experi-
ments to more rigorously test the deleterious effects of text
messaging on learning in a classroom environment.
Experiment 1
Method
Participants
A total of 67 students across three consecutive semesters of an
upper-level cognitive processes class at Butler University
participated in this experiment. Participation was voluntary,
and we did not compensate students, as the experiment was
conducted as part of an in-class demonstration.
Materials and Procedure
During a lecture on attention and time management, we asked
for volunteers who had unlimited text-messaging plans on their
mobile devices and who had their device with them to partici-
pate in a demonstration. We selected a subset of these volun-
teers equal to approximately half of the students in the class
to be in the text condition of the experiment. Those in the text
condition (n ¼ 35) submitted their mobile phone numbers, which we shuffled and redistributed to one other text-condition
student. During the lecture, students in the text condition began
and sustained a prescribed conversation via text message with
both the person whose phone number they received and the
person who received their phone number. The remaining stu-
dents in the class (those who did not volunteer or whom we did
not select for the text condition) were in the no-text control
condition (n ¼ 32).
All participants heard a brief lecture on time management
strategies and took a quiz (announced before the lecture) on the
material from the lecture. We projected the text conversation
(see Appendix) on the screen at the front of the room and began
the lecture, which lasted approximately 12 min. After a 5- to
7-min delay, participants completed a multiple-choice quiz
on the time management material. After taking the quiz but
before seeing their grade, participants also assessed their per-
formance by indicating what percentage of the questions they
believed they had answered correctly (1 ¼ 0–49%, 2 ¼ 50– 59%, 3 ¼ 60–69%, 4 ¼ 70–79%, 5 ¼ 80–89%, 6 ¼ 90–99%, 7 ¼ 100%). We then collected the experimental materials and resumed the class lecture on divided attention.
Results
Figure 1 displays the mean percentage of correctly answered
quiz items as a function of text-messaging condition. A two-
way analysis of variance (ANOVA) with condition (text vs.
no-text) and semester (Spring 2010 vs. Fall 2010 vs. Spring
2011) as between-participants independent variables revealed
a main effect of condition, F (1, 61)¼ 14.24, mean square error (MSE) ¼ 427.51, p < .001, Z2 ¼ .189. Participants in the text condition (M ¼ 60.14%, SD ¼ 23.81%) answered significantly fewer quiz items correctly than did participants in the no-text
condition (M ¼ 79.22%, SD ¼ 15.56%). Neither the main effect of semester nor the condition by semester interaction
reached significance, both Fs (2, 61) < 1, suggesting that the
effect of condition was consistent across time.
Figure 2 displays the mean performance self-assessments of
participants in the text condition and participants in the no-text
condition. A two-way ANOVA with condition (text vs. no-text)
and semester (Spring 2010 vs. Fall 2010 vs. Spring 2011) as
between-participants factors revealed a main effect of text con-
dition, F(1, 61) ¼ 31.35, MSE ¼ 1.32, p < .001, Z2 ¼ .339.
Figure 1. Mean percentage of correctly answered quiz items as a function of text-messaging condition in Experiment 1.
Gingerich and Lineweaver 45
at ROCHESTER INST OF TECHNOLOGY on March 20, 2015top.sagepub.comDownloaded from
Participants in the text condition (M¼ 3.83, SD¼ 1.38) did not believe they answered as many items correctly on the quiz as
did those in the no-text condition (M ¼ 5.53, SD ¼ .84). This indicates that participants who texted during the lecture were
aware that their performance on the quiz was compromised.
The main effect of semester, F(2, 61) ¼ 1.56, MSE ¼ 1.32, p ¼ .218, Z2 ¼ .049, and the condition by semester interaction, F(2, 61) ¼ 0.72, MSE ¼ 1.32, p ¼ .493, Z2 ¼ .023, failed to reach significance. Thus, the lower confidence that the text
group had in their quiz scores relative to the no-text group was
consistent across all three semesters.
Discussion
The demonstration described in Experiment 1 effectively illus-
trates how texting during class impairs comprehension and
retention of lecture material as measured by quiz scores. The
results of Experiment 1 also provide preliminary empirical
evidence that text messaging during lecture interferes with
mastery of lecture material. However, our findings are limited
by a number of factors. Our study did not utilize a true experi-
mental design. We did not randomly assign our students to the
text or the no-text condition. Instead, students who had mobile
phone plans with unlimited text messaging volunteered for the
text condition so that students would not incur any monetary
expense as a result of their participation. As such, a selection
bias may have influenced our results. Specifically, students
who are heavy media multitaskers may be more likely to have
phone plans with unlimited text messaging. Previous research
has shown that, compared to light media multitaskers, heavy
media multitaskers are more likely to experience interference
from irrelevant stimuli in the environment and from irrelevant
memory representations (Ophir, Nass, & Wagner, 2009). Thus,
if heavy media multitaskers were overrepresented in our text
group and light media multitaskers were overrepresented in our
no-text group, the group differences we documented in quiz
scores may be attributable to a differential vulnerability to
interference from irrelevant stimuli.
Another potential scientific limitation of Experiment 1 is
that we conducted it as part of an in-class demonstration of the
effects of divided attention. Although students were aware that
they would take a quiz on the information, which, presumably,
motivated their best effort, they were also aware of the intended
outcome of the demonstration, which may have introduced
demand characteristics.
To control both of these potential confounds, we designed
Experiment 2 to replicate the results of Experiment 1 in a con-
trolled laboratory setting. We excluded students who did not
have unlimited texting as part of their mobile phone plans and
randomly assigned students who did have unlimited texting to
the text or the no-text condition. We lectured over material
unrelated to divided attention and included participants who
were not enrolled in a cognitive processes class in order to
reduce these possible biases.
We also used Experiment 2 to follow up on our somewhat
surprising results that demonstrated that students in the text
group accurately judged their performance on the quiz to be
lower than that of the no-text group. Our students made this
judgment after they took the quiz, but before they received their
score. One possibility is that students do not realize that text
messaging during class is distracting to their learning until after
they are faced with a quiz on which they do not know several of
the answers. In a typical classroom setting, quizzes may follow
lectures by several days to several weeks. In this case, a number
of intervening events occur between the lecture and the quiz,
and students who are seeking explanations for a poor quiz score
may be unlikely to attribute their poor retention of lecture
material to texting during class. In addition, students who
recognize their learning deficit only after taking a quiz may
be too late to adjust their study habits in order to overcome
it. In other words, students may not feel that their learning is
impaired after text messaging during a lecture, but they may
realize this impairment after attempting to retrieve the material
later. If this is the case, text messaging during class may have
an even greater detrimental effect on students’ grades than we
were able to demonstrate in our study. Students may not recog-
nize their lower mastery of the lecture material covered while
they were texting until they take a quiz or exam, when it is too
late for them to compensate for their decreased initial learning
by increasing the time and effort they dedicate to studying that
information in preparation for the graded assessment.
Because the design of Experiment 1 did not allow us to
determine at what point in time students became aware of how
texting during lecture affected their learning, we gathered
students’ predictions about their performance at several time
points in Experiment 2 to investigate whether students who text
message during lecture recognize that their comprehension has
been compromised even before they take a quiz. Including
additional measures of students’ learning confidence also
allowed us to investigate whether the accuracy of students’
metacognitive judgments are affected by their being distracted
Figure 2. Mean performance self-assessment as a function of text- messaging condition in Experiment 1.
46 Teaching of Psychology 41(1)
at ROCHESTER INST OF TECHNOLOGY on March 20, 2015top.sagepub.comDownloaded from
by text messaging while learning new material in class. Some
researchers have found that dividing participants’ attention
while they encode a word list results in less accurate judgments
of their own learning than allowing participants to focus their
attention (e.g., Barnes & Dougherty, 2010). Experiment 2
allowed us to investigate whether this is also true when applied
to learning lecture material in a classroom-like setting.
Experiment 2
In order to address the selection bias and demand characteristics
present in Experiment 1, we designed and conducted a more con-
trolled experiment, one that did not take place in the context of a
class on divided attention and one in which participants were
randomly assigned to text or not to text during the ‘‘lecture.’’
Method
Participants
Fifty-six undergraduate students (40 women and 16 men) at
Butler University participated in this experiment. Participants
either received extra credit in a psychology course or a
US$10 gift card to a fast-food restaurant in return for their par-
ticipation. Participants were recruited from psychology courses
through an online participant management program.
Materials and Procedure
Participation occurred in group testing sessions that ranged in
size from 1 person to 15 people. After obtaining informed
consent, we told participants that the purpose of the study was
to investigate how classmates communicate about lecture
material and how their communication affects learning. We
randomly assigned each participant to the ‘‘text’’ or ‘‘no-text’’
condition. However, in order to minimize participants’ expec-
tations regarding their performance in this experiment, we told
participants that they were assigned to one of three conditions:
a no-text group, a text about lecture content group, or a text
about unrelated content group. In reality, no participants texted
about lecture content. Participants were told that they would
listen to a lecture about the effectiveness of various study
strategies before taking a quiz on the lecture content. All
participants were told to take notes on the lecture material as
though they were in class. Participants then received a packet
containing their assigned condition and a text conversation
between two people. Those assigned to the text condition
received the phone number of one other participant and were
instructed to skim the conversation and to begin texting that
conversation when the ‘‘lecture’’ began. Those assigned to the
no-text condition were instructed to read the text conversation,
which had occurred between two students recently. After learn-
ing what the quiz would involve but before the lecture began,
participants predicted how well they would perform on the quiz
by indicating how many of the nine quiz questions they
expected to answer correctly. After they made their prediction,
the experimenter began the lecture on how people learn.
Participants in the text condition sustained the prescribed con-
versation via text message with another student in the room
during the lecture, whereas those in the no-text condition did
not. At the end of the approximately 30-min lecture, partici-
pants again predicted how well they would perform on the quiz.
In order to prevent participants from rehearsing lecture
content, we instituted a distractor phase, in which participants
