Seating Arrangement, Group Composition and
Competition-driven Interaction: Effects on Students'
Performance in Physics
R. M. Roxas
1
, S. L. Carreon-Monterola
2
, and C. Monterola
1
1
National Institute of Physics, University of the Philippines, Diliman, Quezon City 1101, Philippines
2
College of Education, University of the Philippines, Diliman, Quezon City 1101, Philippines
Abstract. We probe the effect of seating arrangement, group composition and group-based competition on students’
performance in Physics using a teaching technique adopted from Mazur’s peer instruction method. Ninety eight lectures,
involving 2339 students, were conducted across nine learning institutions from February 2006 to June 2009. All the
lectures were interspersed with student interaction opportunities (SIO), in which students work in groups to discuss and
answer concept tests. Two individual assessments were administered before and after the SIO. The ratio of the post-
assessment score to the pre-assessment score and the Hake factor were calculated to establish the improvement in
student performance. Using actual assessment results and neural network (NN) modeling, an optimal seating
arrangement for a class was determined based on student seating location. The NN model also provided a quantifiable
method for sectioning students. Lastly, the study revealed that competition-driven interactions increase within-group
cooperation and lead to higher improvement on the students’ performance.
Keywords: peer instruction method, seating arrangement, group composition, competition, cooperation
PACS: 84.35.+i, 01.40.Ha, 01.40.gb
INTRODUCTION
The effectiveness of peer instruction (PI) method in
increasing student’s learning of Physics concepts has
already been established [1]. Studies have shown that
PI enhances conceptual understanding, subject
mastery, and problem solving skill [2]. In this teaching
method, the students are given interaction
opportunities (SIOs) to discuss their answers in a given
conceptual question. Research showed that the
percentage of correct answers typically increases after
SIOs [3]. This increase in test scores is not simply due
to the influence of knowledgeable peers. Recently,
peer discussion has been shown to enhance
understanding even when none of the students in a
discussion group originally knows the correct answer
[4].
In this study, we present a strategy to optimize
learning by considering the effect of seating
arrangement, group composition, and motivation if
deliberate student-student interactions are integrated in
a lecture design. We also characterized the information
propagation dynamics in a classroom using neural
network modeling.
METHODOLOGY
Modified Peer Instruction Method
Sixty-eight (68) experimental lectures framed on
Eric Mazur’s Peer Instruction Method [5] were
conducted in 9 learning institutions to investigate the
effect of seating arrangement on students’ learning.
Of these, 20 lectures were done involving college
students and the remaining 48 lectures with high
school students.
For each lecture session, the following format was
used: main lecture (20 min), first assessment (10 min),
student interaction opportunity (SIO) (15 min), and
second assessment (10 min). The main lecture
consisted of Physics demonstrations and conceptual
questions called ConcepTests. After the lecture, a
pencil-and-paper test composed of 10 multiple choice
questions was administered as the first assessment.
During SIO, students were instructed to interact with
their neighbors. Accordingly, they were able to do the
following: review answers to conceptual questions in
the main lecture; discuss prior knowledge on the topic
that was not included in the lecture; list and answer
155
CREDIT LINE (BELOW) TO BE INSERTED ON THE FIRST PAGE OF EACH PAPER
CP1263, International Conference on Physics Education 2009 – ICPE2009
edited by B. Paosawatyanyong and P. Wattanakasiwich
© 2010 American Institute of Physics 978-0-7354-0816-6/10/$30.00