Leading Edge
Commentary
Blended Learning Improves Science Education
Brent R. Stockwell,1,2,3,* Melissa S. Stockwell,4,5,6 Michael Cennamo,7 and Elise Jiang1
1Howard Hughes Medical Institute, Department of Biological Sciences, Columbia University, Northwest Corner Building, MC 4846, 550 West
120th Street, New York, NY 10027, USA
2Department of Chemistry, Columbia University, New York, NY 10027, USA
3Department of Systems Biology, Columbia University, New York, NY 10032, USA
4Department of Pediatrics, Columbia University, New York, NY 10032, USA
5Department of Population and Family Health, Mailman School of Public Health, Columbia University, New York, NY 10032, USA
6NewYork-Presbyterian Hospital, New York, NY 10032, USA
7Columbia Center for New Media Teaching and Learning, Columbia University, New York, NY 10027, USA
*Correspondence: bstockwell@columbia.edu
http://dx.doi.org/10.1016/j.cell.2015.08.009
Blended learning is an emerging paradigm for science education but has not been rigorously assessed. We performed a randomized controlled trial of blended learning. We found that in-class
problem solving improved exam performance, and video assignments increased attendance and
satisfaction. This validates a new model for science communication and education.
Blended Learning Is an Emerging
Instructional Pedagogy
At the undergraduate level, science is
most commonly taught using a lecture
and textbook format. In this pedagogical
approach, students are assigned a textbook to read at home before class and
listen to an instructor lecture on the assigned material during class. Although
this approach to learning is effective for
some students, it is suboptimal for many
undergraduate students and may contribute to students leaving the sciences
(Handelsman, 2007).
Recognition of the limitations of the
traditional instructional approach to science using a textbook and lecture format has led to suggestions for examining
alternative methodologies (Handelsman,
2007). In recent years, online learning
options, including massive open online
courses (MOOCs), have become increasingly available as a means to produce
learning in students who cannot attend
classes in person and have been proposed
as alternative learning paradigms (Reich,
2015). However, the low completion
rates of online courses, such as MOOCs,
and the importance of instructor-student
and student-student interactions in classrooms have suggested that online learning
alone is unlikely to be the most effective
strategy for teaching and learning (Glazer,
2012; Reich, 2015).
Nonetheless, the online learning paradigm offers valuable tools that could
supplement or replace aspects of the
traditional
lecture-and-textbook-based
approach to teaching and learning (Glazer,
2012). First, pre-class online video assignments may offer advantages over textbook
assignments, especially for introductory
science courses with complex and dense
material that is unfamiliar to beginning
students (Kagohara, 2010). Reading a
textbook engages mainly visual, language
comprehension, and cognitive neural pathways and requires that the reader is able
to select the most relevant material for
application to the course (Wandell, 2011).
Video instruction, on the other hand, adds
auditory engagement to visual, language
comprehension and cognitive processes,
and allows for more varied emphasis of
the importance of content. Video assignments are typically more engaging for a
large introductory science course and
may stimulate greater engagement with
the course material.
In addition, listening to a lecture in class
involves mostly recording and recalling
information, which are lower levels of
Bloom’s taxonomy of learning (Bloom,
1956). In contrast, solving problems in
real time during class forces students to
synthesize and apply knowledge as they
process it (Amador et al., 2006). A historical comparison has shown that a structured course with in class problem solving
improves performance and reduces the
achievement gap (Haak et al., 2011).
Moreover, a survey of pre/post-test data
in physics courses revealed that interactive teaching improved student learning
(Hake, 1998). Finally, low-stakes formative assessments improved exam performance when compared across students
in different class sections with different
teachers (Roediger et al., 2011). Thus,
active, problem-based learning may improve student performance on exams
(Che et al., 1998; Knight and Wood, 2005).
We wondered if pre-class video assignments and in-class problem solving
would result in increased engagement,
satisfaction and more effective learning.
While it has been suggested that these
techniques in some contexts and in isolation improve learning (Fitzgerald and Li,
2015; Freeman et al., 2014; Glazer, 2012;
Handelsman, 2007), studies involving
undergraduate science education have
generally not been conducted as randomized controlled trials under otherwise identical conditions with the same material and
the same instructor. Rather, they have
been conducted using different instructors
or different course materials, making it
difficult to assess if changes seen were
due specifically to the teaching methods.
We endeavored to rigorously examine
whether video-based preparation and inclass problem solving increase student
engagement, as indicated by class attendance, and ultimately exam performance,
for undergraduate science students.
A Randomized Controlled Trial to
Evaluate Blended Learning
A total of 172 students enrolled in Biochemistry I: Structure and Metabolism,
Cell 162, August 27, 2015 ª2015 Elsevier Inc. 933
Figure 1. Blended Learning Combines
Different Means of Content Delivery
an undergraduate biochemistry course
at Columbia University, were invited to
participate in a randomized controlled trial
in the fall of 2014. A total of 111 students
enrolled in the study. We used a two-bytwo study design, in which we compared
the effects of both video versus textbook
pre-class assignments and lecturing with
instructor-demonstrated problems versus
lecturing with student problem-solving in
class. Students were randomized to one
of four arms: (1) textbook preparation for
lecture, (2) video preparation for lecture,
(3) textbook preparation for problemsolving class, or (4) video preparation for
problem-solving class (Figure 2). The students were stratified for randomization by
gender and prior exam performance (low:
lower third versus high: upper two-thirds)
into each of the four arms to ensure equal
representation of these students in each
study arm; 54% of students in the study
(60/111) were male (Figure 2).
In advance of class, students were
provided either a link to a video or a link
to a textbook reading, covering the
same material, as per their randomization
assignment; students were instructed not
to review other materials and had little
incentive to do so, since the study results
were not included in their grade, as participation was voluntary, and since there
was no assigned textbook for the course
for them to otherwise consult. During
class, students either listened to an
instructor-delivered lecture or listened to
the same lecture material interspersed
with instructions to solve problems related to the presented material, again as
per their randomization arm. To ensure
that the content was the same in each
class, the students who attended the lecture were also provided with the same
problems and their solutions, but these
were explained by the instructor rather
than being solved by the students. Thus,
the difference between the lecture and
problem-solving class formats was not in
the content but rather whether the students actively solved the presented problems or were simply told the answers as
part of the lecture. All lectures were provided by the same instructor.
The first outcome measured was class
attendance. Since participation was voluntary, students were able to discontinue
participation at any point in the study.
We analyzed whether the rate of class
attendance after receiving the preparation
material was different between the students who received a video versus textbook pre-class assignment. Indeed, we
found that more students randomized to
the pre-class video assignment attended
class (84%, 47/56) compared with those
randomized to the textbook assignment
(67%, 37/55) (p = 0.04, Pearson’s chisquare test) (Figure 2). In other words,
twice as many students chose not to
attend class after receiving a textbook
assignment (18/55; 32.7%) compared to
those who received a video assignment
(9/56; 16.1%). This result was consistent
with the hypothesis that a video is a
more engaging way to present new
and complex material to students and
stimulates students to be interested
in learning more about the topic by
attending class.
To test this hypothesis, we examined
the level of satisfaction among the stu-
934 Cell 162, August 27, 2015 ª2015 Elsevier Inc.
dents with the preparation material. There
was significant improvement in satisfaction with the preparation material among
the students who received the video
assignment (4.3/5.0) compared to students who received a textbook assignment (2.9/5.0) (p < 0.0001, Mann-Whitney
test) (Figure 2). This is consistent with the
fact that fewer students who received the
textbook assignment chose to attend
class and suggests that textbook preparation for a science class is less satisfactory and engaging for students compared
to assignments that involve watching a
video.
At the end of each class, students took
a 20 min online, multiple-choice exam,
while still present in class, to test their understanding of the material. First, in order
to test whether the exam measured the
same performance characteristics found
in the rest of the course, we compared
exam performance for students with a
higher prior exam performance versus a
lower one. Indeed, we found that prior
exam performance (high versus low)
correlated with the study exam performance: the median exam scores for students in the high prior exam group (73/
100) was significantly (p = 0.006, Mann
Whitney test) greater than the median
exam score for students in the low prior
exam group (60/100), suggesting that
the students exerted their typical effort
and exhibited similar performance characteristics, despite the fact that participation was voluntary.
We then examined the performance of
the four study arms on the end-of-class
exam to determine the effects of teaching
pedagogy on student performance. We
found that the median exam score
increased within each arm, from arm 1
(textbook preparation, in-class lecture;
61/100) to arm 2 (video preparation, inclass lecture; 67/100) to arm 3 (textbook
preparation, in-class problem solving;
73/100) to arm 4 (video preparation, inclass problem solving; 80/100) (Figure 2).
Indeed, the students in the arm that experienced fully blended learning, with
video preparation and in class problem
solving, had the highest median exam
score (80/100).
We found that the most significant
intervention was the implementation of
student-centered problem solving during
class, as the median score on the exam
Figure 2. Randomized Controlled Trial to Evaluate Teaching and Learning Strategies for Undergraduate Biochemistry
(A) The study followed the indicated scheme, in which students were invited to participate voluntarily in a study and asked to indicate their informed consent. They
were randomized into one of four groups, receiving either a textbook or video assignment before class and either a lecture or a lecture with problem solving during
class. The impact on attendance and exam score was then measured.
(B) Equivalent numbers of male and female students and students with high and low prior exam scores were assigned to each group.
(C) Students who received a video assignment were more likely to attend class than students who received a textbook assignment. The number of students who
attended or did not attend class after receiving a video or textbook assignment is indicated.
(D) Students who solved problems in class performed better than students who only listened to a lecture. The median score of each group of students on the end
of class exam is indicated (out of 100 possible points). The comparison of scores is shown for the students in the lecture and problem-solving groups. Video
assignments were more satisfying to students than textbook assignments; students were asked to rate how satisfied they were with their pre-class assignment on
a scale of 1 (low satisfaction) to 5 (high satisfaction).
was higher for those who were randomized to the lecture-plus-problem-solving
class (74/100) compared to those randomized to the lecture-only class, irrespective of preparation material (63/100)
(p = 0.03, Mann Whitney test) (Figure 2).
We also compared the exam scores of
students in the textbook versus video
preparation groups but found no statistically significant difference in this relatively
modest sample size, despite the trend
toward higher scores in the group that
received the video assignment.
We recognized that the instructor in a
randomized controlled trial may exhibit
unconscious bias. We sought to measure
this objectively by evaluating student
satisfaction with the lecture and problem-solving classes. Student satisfaction
did not differ significantly between students who attended the problem-solving
versus lecture classes (4.1 versus 4.1,
out of 5.0, p = 0.99, Mann-Whitney test)
(Figure 2). This suggests that the traditional lecture was engaging and of high
quality and that the difference in exam
performance was not due to unconscious
instructor bias in the delivery of the lecture material. Moreover, it suggests that
student satisfaction alone is not a reliable indicator of learning gains and
that institutions and instructors should
be cautious in evaluating pedagogical
methods using only student satisfaction
measures, which are the typical metrics
used in course evaluations. Indeed,
objective measures of learning gains
may ultimately be the most accurate
means of assessing courses.
Overall, we conclude that providing students with problems to solve during class
results in significantly improved exam
performance, compared to simply having
the instructor describe the same problems and their solutions during the course
of the lecture.
Implications for Science Education
In recent years, there has been a growing
interest in improving science education to
increase the diversity of individuals who
choose science careers and to broaden
the pool of scientifically literate citizens
globally. The traditional textbook-andlecture approach to teaching undergraduate science, while effective for a
subset of individuals, is not the most
effective means of stimulating learning in
the broadest group of students. We and
others have sought to examine the effectiveness of an alternative pedagogy that
replaces textbooks with video assignments and traditional lecturing with
active, student-centered problem solving.
One result of this trial was that students
who received a video assignment were
both more likely to attend class and to
rate their assignment as providing a
higher degree of satisfaction, compared
to students who received a textbook
assignment. Thus, in this study, video assignments were more effective at stimulating student interest and engagement
with the course material. Therefore,
providing supplemental videos for science
Cell 162, August 27, 2015 ª2015 Elsevier Inc. 935
courses, either in place of, or in addition to,
textbook assignments, may enhance student engagement and motivation.
A second finding was that students who
listened to a traditional instructor-focused
lecture underperformed students who
actively solved problems during the class
period. This was despite the fact that the
students in these two groups rated their
classes as equally satisfying. Thus, while
students enjoyed passively listening to a
lecture as much as being asked to solve
problems during class, they learned the
material better when they actively worked
on problems, rather than simply being
given the problems and answers as part
of the lecture. Therefore, instructors may
wish to provide frequent opportunities
during class for students to apply the content presented to specific problems. This
provides ongoing formative assessments
for students to test their learning and refocuses student attention during an otherwise extended period of lecturing. In addition, asking a student to apply knowledge
to new contexts may assist them in
learning the material in a more effective
way that allows them to extrapolate the
knowledge to new situations in the future;
a related case-based pedagogy has been
used successfully in business, medical,
and law schools.
One of the difficulties faced when trying
to rigorously assess the effectiveness of
teaching and learning strategies is the
challenge of delivering identical content
via the same instructor and same pool of
students such that it is the teaching
method that is being assessed and not
other factors. A variety of strategies have
attempted with limited success to deal
with this challenge, such as historical
comparisons among the students who
take a particular class that has altered its
methodology, comparisons among instructors who use different teaching
methodologies, or comparisons among
the same instructor using different teaching methodologies within one course but
for different topics and at different points
of time in the semester. We suggest that
it is most effective to use the rigor and
methodology of randomized controlled
trials, which are able to measure the
impact of patient interventions in health
and biomedicine, to examine the impact
of teaching approaches on student
learning (Drits-Esser et al., 2014). By using
the same content, the same instructor,
the same point of time, and the same
pool of students randomized to different
conditions, we were able to minimize the
impact of variables beyond the ones we
sought to measure. In addition, the trial
was conducted in a pragmatic fashion
with students within their usual course
with their usual peers and instructor and
is therefore reflective of results in a realworld setting. It also demonstrates the
feasibility of conducting such pragmatic
education trials.
Unfortunately, we were only able to
examine the effects of teaching style in
one class. Additionally, we did not
examine the long-term effect of these interventions on learning; it would be valuable to have students take another exam
several months or even years after participating in the study to examine how well
they retain the information. We propose
that such longitudinal assessments of
learning gains could be organized at the
institutional level by regularly administering follow-up exams to students in
specific majors who have completed the
same required courses early in their
program. Perhaps by coupling such
programs with a compensation system
for participation, the long-term impact
of course structures and pedagogical
methods could be assessed.
Overall, the results from this trial suggest that a blended teaching approach,
which uses video assignments in advance
of each class to stimulate interest in
the topic and provide foundational
knowledge, coupled with lectures having
in-class problem solving, is a more effective strategy for science education
compared with traditional approaches. It
is worth noting that video assignments
did not improve student exam performance on their own but did increase
attendance and satisfaction. By providing
foundational information pre-class, video
assignments can also create time in
class for active learning, such as student
problem solving. Thus, video pre-class
assignments can serve a critical role in
the blended learning paradigm for education.
These results also illustrate the feasibility of using the clinical trial methodology in educational intervention evaluations. Blended learning approaches may
help students learn information in a
936 Cell 162, August 27, 2015 ª2015 Elsevier Inc.
way that they can then translate to
novel situations they will encounter in
their academic and professional careers, which is the hallmark of effective
learning. Institutions and instructors may
wish to consider how to support blended
learning paradigms in their science
curricula.
ACKNOWLEDGMENTS
We thank the Columbia Center for New Media
Technology and Learning (CCNMTL) for assistance in preparing and disseminating videos to
students, Daniel Rabinowitz for advice regarding
statistical analyses, and the students who participated in this study. Brent R. Stockwell is an Early
Career Scientist of the Howard Hughes Medical
Institute. Financial support for this study was
provided by Columbia University’s Office of the
Provost. The authors do not have any financial
conflicts of interest to report.
REFERENCES
Amador, J.A., Miles, L., and Peters, C.B. (2006).
The Practice of Problem-Based Learning: A Guide
to Implementing PBL in the College Classroom
(Anker Pub. Co.).
Bloom B.S., ed. (1956). Taxonomy of Educational
Objectives. Handbook 1: Cognitive Domain (David
McKay Company, Inc.).
Che, S., Wu, W., Nelman-Gonzalez, M., Stukenberg, T., Clark, R., and Kuang, J. (1998). FEBS
Lett. 424, 225–233.
Drits-Esser, D., Bass, K.M., and Stark, L.A. (2014).
CBE Life Sci. Educ. 13, 593–601.
Fitzgerald, N., and Li, L. (2015). J. Chem. Educ.
Published online March 6, 2015. http://dx.doi.org/
10.1021/ed500667.
Freeman, S., Eddy, S.L., McDonough, M., Smith,
M.K., Okoroafor, N., Jordt, H., and Wenderoth,
M.P. (2014). Proc. Natl. Acad. Sci. USA 111,
8410–8415.
Glazer, F.S. (2012). Blended Learning: Across
the Disciplines, across the Academy (Stylus
Publishing).
Haak, D.C., HilleRisLambers, J., Pitre, E., and
Freeman, S. (2011). Science 332, 1213–1216.
Hake, R.R. (1998). Am. J. Phys. 66, 64–74.
Handelsman, J. (2007). Scientific Teaching (W.H.
Freeman and Co.).
Kagohara, D.M. (2010). Dev. Neurorehabil. 13,
129–140.
Knight, J.K., and Wood, W.B. (2005). Cell Biol.
Educ. 4, 298–310.
Reich, J. (2015). Science 347, 34–35.
Roediger, H.L., Agarwal, P.K., McDaniel, M.A., and
McDermott, K.B. (2011). J. Exp. Psychol. Appl. 17,
382–395.
Wandell, B.A. (2011). Ann. N Y Acad. Sci. 1224,
63–80.
Purchase answer to see full
attachment