Chiu2015 PDF

Accepted Manuscript
The Effects of Augmented Virtual Science Laboratories on Middle School Students’

Understanding of Gas Properties
Jennifer L. Chiu, Crystal J. DeJaegher, Jie Chao
PII: S0360-1315(15)00051-2
DOI: 10.1016/j.compedu.2015.02.007
Reference: CAE 2798
To appear in: Computers & Education
Received Date: 7 August 2014

Revised Date: 29 December 2014
Accepted Date: 7 February 2015
Please cite this article as: Chiu J.L., DeJaegher C.J. & Chao J., The Effects of Augmented Virtual
Science Laboratories on Middle School Students’ Understanding of Gas Properties, Computers &
Education (2015), doi: 10.1016/j.compedu.2015.02.007.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT

Jennifer L. Chiu, Crystal J. DeJaegher, Jie Chao
PT
University of Virginia, 405 Emmet Street South, Charlottesville, VA 22904 USA
Corresponding author: Jennifer L. Chiu
RI
ADDRESS: 313 Bavaro Hall, Curry School of Education, University of Virginia, Charlottesville,
VA, 22904, USA. PHONE: 650 5759848, FAX: 434 9247461, EMAIL: jlchiu@virginia.edu
SC
Abstract
The Next Generation Science Standards (NGSS) emphasize authentic scientific practices such as
developing models and constructing explanations of phenomena. However, research documents
how students struggle to explain observable phenomena with molecular-level behaviors with
U
current classroom experiences. For example, physical laboratory experiences in science enable
students to interact with observable scientific phenomena, but students often fail to make
AN
connections to underlying molecular-level behaviors. Virtual laboratory experiences and
computer-based visualizations enable students to interact with unobservable scientific concepts,
but students can have difficulties connecting to actual instantiations of the observed phenomenon.
This paper investigates how combining physical and virtual experiences into augmented virtual
M
science laboratories can help students build upon intuitive ideas and develop molecular-level
explanations of macroscopic phenomena. Specifically, this study uses the Frame, a sensor-
augmented virtual lab that uses sensors as physical inputs to control scientific simulations.
D
Eighth-grade students (N=45) engaged in a Frame lab focused on the properties of gas. Results
demonstrate that students using the Frame lab made progress developing molecular-level
TE
explanations of gas behavior and refining alternative and partial ideas into normative ideas about
gases. This study offers insights for how augmented virtual labs can be designed to enhance
science learning and encourage scientific practices as called for in the NGSS.
EP
Highlights
• Combining virtual and physical labs has potential to promote science learning.
• Minimal research exists on augmented virtual technologies in authentic classrooms.
C
• Students use physical controls of the Frame to manipulate rich, virtual simulations
• Pilot tests show student improvement on explanations about gas properties.
AC
Keywords
Evaluation of CAL systems; Improving classroom teaching; Interactive learning environments;
Multimedia/hypermedia systems
ACCEPTED MANUSCRIPT
Running head: EFFECTS OF AUGMENTED VIRTUAL LABS ON SCIENCE 1
UNDERSTANDING

PT
1. Introduction
The Next Generation Science Standards (NGSS) stress authentic scientific practices such
as developing models and constructing explanations of phenomena (NGSS Lead States, 2013).
However, research documents that students have difficulty developing molecular-level
RI
explanations of observable phenomena, critical to complex science understanding (Ben-Zvi,
Eylon, & Silberstein, 1986; Gabel, 1999). Moreover, existing approaches to science classes can
leave students with isolated or superficial ideas (Linn & Eylon, 2011). For example, hands-on
SC
laboratory experiences give students direct experience with phenomena and scientific practices
(National Research Council [NRC], 2006), but are not always successful in getting students to
understand underlying scientific concepts (Finkelstein et al., 2005). Virtual technology tools,
U
software, and simulations, have been successfully implemented in science classrooms to help
students develop explanations of complex science topics (Bell & Trundle, 2008; Carlsen &
AN
Andre, 1992; Chiu & Linn, 2014; Höffler & Leutner, 2007; Honey & Hilton, 2011; Jaakola,
Nurmi, & Veermans, 2011; Windschitl & Andre, 1998). However, research also demonstrates
how students using visualizations can focus on superficial aspects (Lowe, 2004), overestimate
their understanding (Chiu & Linn, 2012), and fail to connect virtual representations to the
M
depicted real-life scientific phenomena (Chiu, 2010). Research that combines physical and
virtual labs either sequentially or side-by-side demonstrates that careful consideration of physical
and virtual affordances can support the development of scientific understanding (Blikstein,
D
Fuhrmann, Greene, & Salehi, 2012; Olympiou & Zacharia, 2012).

Augmented virtual technologies offer an innovative approach to science laboratories by
TE
combining virtual and physical components to provide enhanced educational experiences

(Lindgren & Johnson-Glenberg, 2013). Augmented virtual technologies use virtual tools to
represent scientific phenomena (e.g., simulations, visualizations) that are enhanced by using
physical real-life objects as controls. This paper focuses on the pilot testing of the Frame, a
EP
specific augmented virtual technology that uses probeware (i.e., temperature and pressure
sensors) as inputs to simulations of scientific phenomena (Xie, 2012), enabling students to use
real-world objects to control the simulation. For instance, instead of students moving a gas-filled
piston with a mouse in a simulated environment, students impart a force on a physical spring that
C
inputs the information into the simulation.

The overall goal of this study is to explore if augmented virtual science laboratories such
AC
as the Frame can be used in authentic classroom settings to help students develop scientifically
normative explanations of gas properties. In particular, this study seeks to answer the following
questions:
1. Can augmented virtual Frame labs help middle school students develop explanations
that connect molecular behaviors to macroscopic properties of gas?
2. Can augmented virtual Frame labs help middle school students refine alternative ideas
about macroscopic gas phenomena?
ACCEPTED MANUSCRIPT
EFFECTS OF AUGMENTED VIRTUAL LABS ON SCIENCE UNDERSTANDING 2
2. Background
2.1 Supporting Science Learning through Grounded Knowledge Integration

Students enter the classroom with a wide range of experiences that contribute to the
framework with which they interpret the world (Bransford, Brown & Cocking, 2000). Students’
experiential understandings can serve as fruitful places to build understanding, but can also
counter and even interfere with developing scientifically normative explanations of phenomena
PT
(Duschl & Gitomer, 1991). Research demonstrates that students often struggle to understand
difficult science concepts because they have trouble integrating scientifically normative
explanations into their existing knowledge base (Clough & Driver, 1985; Erickson & Tiberghien,
RI
1985; Jones, Carter, & Rua, 2000), especially for topics that include unobservable levels, such as
gas laws and kinetic molecular theory (e.g., Nakleh, 1992; Novick & Nussbaum, 1981).
Students often confuse molecular and macroscopic levels of a phenomenon (Ardac & Akaygun,
SC
2004; Ben-Zvi, Eylon, & Silberstein, 1986) or misattribute characteristics of one level to another
(e.g., Wilensky & Resnick, 1999). Students’ alternative ideas about molecular behaviors can
stem from everyday macroscopic experiences. For example, many students believe that when
there are more particles in a fixed container, particles have less room to move and thus move
U
slowly (Levy, Novak, & Wilensky, 2006). Conflicts between experiential and scientific
understandings can be challenging to overcome but can also serve as particularly fruitful places
AN
to help students engage in conceptual change or the restructuring of ideas (Chi, 2008; Posner,
Strike, Hewson, & Gertzog, 1982; Vosniadou, 1994). Providing students with opportunities to
make and refine connections between everyday and normative ideas can be particularly
successful for science learning (Clark, 2006; Clark & Jorde, 2004; Levy & Wilensky, 2009; Shen
M
& Linn, 2011).

To help students make connections among molecular and macroscopic levels and help
students leverage everyday ideas with the Frame, this study uses a knowledge integration (KI)
D
learning perspective. The KI framework provides guidance for instructional strategies that
encourage students to use their existing knowledge base and build understandings by making,
TE
refining and sorting connections among ideas (Clark, 2006; Linn & Eylon, 2011). Creating an
environment that fosters knowledge integration elicits students’ existing understandings, adds
new ideas for students to consider, gives students with an opportunity to distinguish between
EP
ideas, and provides a framework for students to sort these ideas (Linn & Eylon, 2006). KI values
experiences where students bring their experiential and scientific knowledge forward, so that
conflicts can be identified and connections made to create more cohesive networks of
understanding (Linn & Eylon, 2011). Many studies demonstrate how KI instructional strategies
C
help students develop connected understanding of science (Chiu & Linn, 2014; Linn, Davis &
Eylon, 2004; McElhaney & Linn, 2011; Zhang & Linn, 2011).
AC
We also leverage embodied and grounded approaches to cognition as a framework for

learning from augmented virtual technologies. Embodied cognition perspectives recognize the
role of bodily actions on cognition (e.g., Lakoff & Johnson, 1980; Wilson, 2002). Grounded
approaches to cognition contend that the mind stores information across perceptual, motor, and
affective states and uses an integrated multimodal representation to build knowledge (Barsalou,
2008). Thus, conceptual understanding involves and connects to ideas that are grounded in the
experiences of the learner, which complements the importance of experiential ideas in the
process of knowledge integration (e.g., Clark, 2006; Lewis & Linn, 1994). A learner’s
understanding of a science concept involves a network of ideas associated with what he or she
ACCEPTED MANUSCRIPT
sees, hears, and feels during instruction, as well as past networks of ideas associated with other,
prior learning. Embodied approaches to learning – that is, having students learn material through
physical interaction as well as through seeing and hearing about the material - can result in
improved conceptual understanding (e.g., Abrahamson, Gutiérrez, Charoenying, Negrete, &
Bumbacher, 2012; Anastopoulou, Sharples, & Baber, 2011; Tolentino, Birchfield, Megowan-
Romanowicz, Johnson-Glenberg, Kelliher, & Martinez, 2009).
PT
2.2 Physical Labs in Science Classrooms
Science instruction has traditionally used hands-on laboratories to help activate
experiential ideas and engage with scientific phenomena. Physical laboratory experiences
RI
benefit students by incorporating concrete objects in science learning; some students can better
engage with science when they are able to touch, move, and examine real objects (Feisel & Rosa,
2005). Physical laboratories also give students opportunities to interact directly with the
SC
scientific phenomena being studied (Lunetta, Hofstein & Clough, 2007).
Despite the widespread use of physical labs in science, hands-on (not computer-based)
labs typically do not provide representations for unseen levels such as atoms and molecules (e.g.,
Hodson, 1996; Hofstein & Lunetta, 2004; Tobin, 1990). As a result, students can have difficulty
U
connecting physical labs with molecular-level ideas and can have difficulties integrating
observable and molecular accounts of phenomena (Gabel, 1999; Johnstone, 1991). For example,
AN
students can engage in a gas laws lab that investigates the relationship between pressure and
volume to help them understand the inverse relationship that as one gets bigger, the other
decreases. However, after completion students may still not understand or connect these results
to the molecular behaviors that explain why this happens on an atomic level (Liu, 2006).
M
2.3 Visualizations in Science Classrooms

Scientific visualizations provide experiences for students to observe and interact with
D
typically unobservable levels of phenomena (Honey & Hilton, 2011). Many studies demonstrate
how visualizations can help students understand the particulate nature of matter, gas laws, and
TE
other molecular-level concepts (e.g., Kozma & Russell, 1997, Levy & Wilensky, 2009, Steiff
and Wilensky, 2003; Wu, Krajcik, & Soloway, 2001). Visualization-based instruction can also
help students make connections from molecular levels to observable levels by providing multiple,
EP
linked representations of phenomena (Kozma, 2003) or explicit guidance through instruction

(Chiu & Linn, 2014; Levy, 2012; Zhang & Linn, 2011). For example, Connected Chemistry
(Levy & Wilensky, 2009) anchored a visualization-based gas laws curriculum in the
macroscopic world and used instruction to help students connect virtual NetLogo models to the
C
particulate level. Students made significant gains from pretest to posttest on items assessing
understanding of submicro-macroscopic connections with a medium effect size. Guiding
AC
students to make connections between visualized molecular-level phenomena and the observable
world, as with the Connected Chemistry curriculum, can enhance conceptual understanding
(Steiff & Wilensky, 2003; Wu, Krajcik, & Soloway, 2001).
Connecting visualizations to everyday, experiential ideas can also benefit student
learning. Clark & Jorde (2004) compared two versions of a simulation: one with a visualization
of heat with an explicit representation of a macroscopic hand and virtual feedback about what it
would feel like touching the object and one with a visualization of heat without the hand or
feedback. Students who used the visualization with the hand outperformed students without the
hand on post and delayed posttests and demonstrated significant improvement on written
ACCEPTED MANUSCRIPT
explanations of thermal concepts. Results demonstrate that connecting visualizations to

experientially acquired ideas can help students refine and articulate understanding of complex
science concepts.
2.4 Combining Physical and Virtual labs

Although many empirical studies demonstrate that visualization-based instruction can
improve student understanding of scientific phenomena (Dori & Belcher, 2005; Finkelstein et al.,
PT
2005; Zacharia & Anderson, 2003), studies also show that students have difficulty learning from
visualizations (Tversky, Morrison, & Betrancourt, 2002). Students may focus on superficial
aspects instead of targeted learning outcomes (Cook, 2006) or overestimate their understanding
RI
of material presented visually (Chiu, King-Chen, & Linn, 2013). For instance, Lowe (2004)
found that students learning from an interactive weather map tended to notice salient, small
aspects of space or time but had difficulty building coherent descriptions of weather that
SC
integrated aspects across the visualization. Similarly, Chiu and Linn (2012) found that students
working with molecular visualizations tended to rate themselves as more knowledgeable
immediately after interacting with a visualization as opposed to after asking them to explain what
they learned from the visualization, suggesting that dynamic visualizations can give students an
U
illusion of understanding the underlying concepts.
Providing sequential physical and virtual experiences has shown promise in helping
AN
students to develop understanding of phenomena (Gire et al., 2010; Zacharia, 2007; Zacharia &
Anderson, 2003; Zacharia & Olympiou, 2011; Zollman, Rebello, & Hogg, 2002). For example,
Olympiou and Zacharia (2012) compared students experimenting with physical, virtual, and both
physical and virtual manipulatives to learn optics. Students using a combination of both physical
M
and virtual manipulatives outperformed students using either physical or virtual manipulatives
alone (Zacharia, 2007). Although the effects of the order of manipulative type on student
outcomes is relatively mixed (Smith & Puntambekar, 2010), the combination of physical and
D
virtual has been shown to be successful and offers greater affordances than either approach in
isolation (de Jong, Linn, & Zacharia, 2013).
TE
In addition to sequential combinations, research demonstrates that students can use both
physical and virtual materials at the same time to develop sophisticated views of science. For
instance, students in after-school settings developed virtual and physical models side-by-side
EP
with biology, chemistry, and physics content. Students used mismatching information from real
and virtual models to develop deeper understanding of phenomena (Blikstein et al., 2012). This
kind of side-by-side or bifocal approach can encourage deep learning in science by engaging
students in authentic scientific practices (Blikstein, 2012).
C
However, combining physical and virtual manipulatives either sequentially or side-by-

side does not guarantee that students will make connections among levels (e.g., McBride,
AC
Murphy, & Zollman, 2010). Building upon these findings, providing students with physical and
virtual manipulatives simultaneously connected has potential to help students make links among
molecular and observable levels in science. Mixed-reality technologies can use physical
interactions to control digital environments, capitalizing on the affordances of both physical and
virtual manipulatives (Borgman et al., 2008). Studies demonstrate the potential of mixed-reality
technologies to enhance science learning (Johnson-Glenberg, Koziupa, Birchfield, & Li, 2011;
Lui & Slotta, 2013; Novellus & Moher, 2011). However, relatively few empirical studies have
investigated learning in authentic classroom settings (Lindgren & Johnson-Glenberg, 2013). The
few studies that have been implemented in classrooms demonstrate the potential of mixed-reality
ACCEPTED MANUSCRIPT
approaches to enhance traditional instruction (e.g., Johnson-Glenberg, Birchfield, Tolentino, &

Koziupa, 2013).
Specifically, this paper explores how sensor-augmented virtual technologies may help
students develop scientifically normative explanations that link molecular-level behaviors to
observable phenomena. Augmented virtual technologies use virtual phenomena augmented with
“real” objects, as opposed to augmented reality, where the real phenomena is augmented with
virtual objects (Milgram & Kishino, 1994). Augmented virtual technologies capitalize on
PT
affordances of simulations and dynamic visualizations by enabling students to not only see but
also interact with typically unobservable processes such as molecular behaviors. The “real”
augmentation leverages advantages of physical labs such as direct manipulation of phenomena
RI
and tactile feedback. By simultaneously connecting real and virtual objects, augmented virtual
technologies have the potential to help students connect molecular behaviors to the observable
world.
SC
2.5 The Frame Technology
This paper presents exploratory work using the Frame, a sensor-augmented virtual
technology that uses physical interactions with sensors to control a molecular simulation (Xie,
U
2012; Figure 1). The Frame engages students through a simulation of a phenomenon (in this
case, gas molecules in a chamber) controlled or augmented by physical inputs. Although the
AN
M
D
TE
C EP
AC
Figure 1. The Frame apparatus (top left) with a molecule “pump” on the left, spring piston controller
on the right, and a jar for holding hot/cold water placed next to the temperature sensor. Students
interact with the visualization (top right) through physical controls (bottom).
ACCEPTED MANUSCRIPT
Frame can be configured to help students understand various topics such as heat transfer and
electricity (Xie, 2012) this study focuses on helping students understand gas properties. The
particulate nature of matter (PNM) is fundamental to complex science understanding (Ozmen,
2013) as it undergirds many chemistry topics (Gabel et al., 1997; Nakleh, 1992; Snir et al., 2003)
as well as topics in physical, life, and earth sciences (Benson et al., 1993; Bouwma-Gearhart et
al., 2009; Lee et al., 1993; Noh & Scharmann, 1997). As a result, the PNM is a central subject in
middle and high school curricula (e.g., NGSS, 2013). However, research demonstrates that
PT
students at all ages, even chemistry graduate students and teachers, have a variety of ideas about
the PNM, gases, and gas laws (Bodner, 1991; Krajcik, 1995; Lin, Cheng, & Lawrenz, 2000;
Nakhleh, 1992). Traditional physical labs typically have students investigate relationships
RI
between volume and pressure or temperature and pressure at the observable level with probeware
or other equipment, but can leave students with alternative ideas about why these macroscopic
relationships occur. Research demonstrates that technology-enhanced approaches can help
SC
students understand chemical phenomena on a molecular level (e.g., Ardac & Akaygun, 2004;
Jones, Jordan & Stillings, 2005; Kelly & Jones, 2007; Madden, Jones & Rahm, 2011; Sanger et
al., 2000; Stieff & Wilensky, 2003; Wu et al., 2001) and in particular, dynamic visualizations can
help students develop molecular understanding of gas behavior (Levy & Wilensky, 2009).
U
Combining physical and virtual approaches has also been beneficial for students studying
gas laws. For example, Liu (2006) had high school students sequentially engage in both
AN
computer-based molecular simulations and a hands-on lab to learn about gas laws. Results
demonstrated that the combination of a physical and virtual lab was better than either approach
alone to develop conceptual understanding of gas laws.
Additionally, evidence suggests that mixed-reality approaches may help students engage
M
in scientific practices that contribute to understanding gas phenomena. Blikstein (2014) piloted
technologies that explicitly linked the volume of a physical syringe to a molecular simulation of
gas in real time. Students could press on a physical syringe and simultaneously see the volume
D
of a virtual syringe change. Students could then compare pressure outputs from both the actual
syringe and the computer model of the syringe through a bifocal pedagogical approach. Pilot
TE
results with 11 high school students over a single 6-hour period suggest that the mixed-reality
technologies encouraged sophisticated scientific practices and fostered inquiry into the targeted
scientific phenomena.
EP
Building from these studies, this paper investigates how the Frame can be used in
classroom settings to help students learn about gas phenomena. Unlike a side-by-side approach,
the Frame uses a simulation as the basis for student experimentation and augments the simulation
with physical controls. Students feel as if they are directly experimenting with gas molecules,
C
leveraging the different affordances of virtual and physical labs into a singular experience. For
example, students place jars filled with hot water near a temperature sensor to increase the
AC
temperature of the simulated gas. Students can physically push on the Frame to increase force
on the force sensor and thus the virtual piston in the simulation. Students can also increase or
decrease the number of molecules in the simulation through a physical pump. The simulation
component of the frame offers a dynamic visualization through which students can also interact
using a touch-screen interface.
We believe that a grounded knowledge integration perspective provides a lens to
understand how the Frame can help students understand gas laws. Specifically, the physical
controls of the Frame can bring forward everyday, intuitive ideas that students may hold about
phenomena through the tactile interaction and feedback. When students interact with physical
ACCEPTED MANUSCRIPT
controls, the feedback from the real world can elicit past experiences and ideas that are
associated with the specific action. For example, when students physically push the spring of the
Frame and feel the force of the spring pushing back, this helps activate related, everyday ideas
about force. The simulation or virtual manipulative can then help students add normative ideas
about the phenomenon – in this case, how a gas exerts pressure and force against a piston wall.
Both the intuitive and scientific ideas are activated at the same time, enabling students to
distinguish productive from less productive ideas and make connections among their ideas with
PT
proper guidance. Activating intuitive and scientific ideas simultaneously provides an
opportunity for students to sort out and refine productive ideas from less productive ideas,
resulting in more coherent networks of ideas (Clark, 2006; Linn & Eylon, 2006). Because
RI
students interact with physical, real world controls and see the resulting behaviors of molecules,
the Frame has the potential to help learners connect macroscopic, experiential ideas to visualized
molecular behaviors and refine alternative ideas.
SC
This paper explores the potential of augmented virtual labs in authentic science
classrooms to help students develop normative explanations of phenomena. As an exploratory
study, this paper first reports on how students used the Frame lab in authentic classroom settings,
then specifically investigates if middle school students can use the Frame to construct and refine
U
explanations that connect molecular behaviors to macroscopic gas properties as well as help
students refine alternative ideas about gas phenomena.
AN
3. Methods
3.1 Curriculum
M
3.1.1 Development
The Gas Laws curriculum used for this implementation of the Frame technology was
D
developed in partnership among educational researchers, scientists, and teachers. The group
chose gas laws as it aligned with both state and national science content standards (e.g., NGSS
TE
DCI MS-PS1.A and MS-PS1-4). The group also chose to emphasize the NGSS practice of
constructing explanations because of the synergies with the goals of the lab, as “scientific
explanations are explicit applications of a theory to a specific situation or phenomenon” (NRC,
EP
2012; p. 52) and the “goal of science is to construct explanations for the causes of phenomena”
(NGSS, 2013; p. 11). Specific learning objectives for scientific practices involved “constructing
an explanation that includes qualitative or quantitative relationships between that predict(s)
and/or describe(s) phenomenon” as well as “apply scientific ideas, principles, and/or evidence to
C
construct, revise, and/or use an explanation for real world phenomenon, examples, or events”
(NGSS, 2013; p. 11).
AC
The collaborative effort identified difficult topics for students when learning about gas
laws based upon both prior research (e.g., Nakhleh, 1992) and the teachers’ experience,
specifically that: gas molecules are in constant motion (Herrman-Abell & DeBoer, 2011),
heating and cooling cause changes in particle motion, not size (Novick & Nussbaum, 1981), gas
particles at the same temperature travel at different speeds (e.g., temperature is a measure of the
average kinetic energy of the molecules), and gas pressure is a result of molecular collisions with
container walls (Lin, Cheng, & Lawrenz, 2000). The Frame lab was specifically designed to
target common alternative student ideas around these concepts through engaging in constructing
explanations.
ACCEPTED MANUSCRIPT
The curriculum design built upon research-based best practices with both physical and
virtual labs to provide the best possible learning experience for the students (e.g., Chang,
Quintana, & Krajcik, 2009; Duschl, Schweingruber, & Shouse, 2007; Hofstein & Lunetta, 2004;
Levy & Wilensky, 2009; Olympiou & Zacharia, 2012; Steiff & Wilensky, 2003; Wu, Krajcik, &
Soloway, 2001). Specifically, curriculum development leveraged research that demonstrates the
benefit of guided inquiry activities for science learning (e.g., Hmelo-Silver, Duncan, & Chinn,
2006; Quintana et al., 2004), as well as the benefit of technology-enhanced inquiry learning
PT
environments to support scientific practices and learning (Donnelly, Linn, & Ludvigsen, 2014).
The Gas Frame lab used a scaffolded knowledge integration instructional approach (Linn, Davis,
& Eylon, 2004) and the Web-based Inquiry Science Environment (WISE) (Figure 1; Slotta &
RI
Linn, 2009), to support students’ investigations with the Frame. The KI instructional approach
promotes design patterns that help students elicit, add, distinguish and refine their ideas about
science concepts (Linn & Eylon, 2006). The Gas Frame Lab primarily leveraged a modified
SC
predict-observe-explain-reflect instructional pattern (e.g., Tien, Teichert, & Rickey, 2007; White
& Gunstone, 1992) to help students engage in knowledge integration as well as provide support
for students developing and refining explanations (e.g., McNeill, Lizotte, Krajcik, & Marx,
2006).
U
The design of the unit also drew upon prior research demonstrating the effectiveness of
WISE to support student learning through virtual investigations with visualizations (Linn et al.,
AN
2006; McElhaney & Linn, 2011), noting the importance of helping students to focus on relevant
aspects of a phenomenon when using simulations (e.g., Chiu & Linn, 2012; Lowe, 2004).
Students were given specific guidance when making observations of the simulation to pay
attention to specific details important to the particular concept. The overall instructional pattern
M
of the lab consisted of helping students (1) make predictions in the form of an initial explanation
of a particular phenomenon, (2) conduct a guided investigation with the Frame of that
phenomenon, (3) make targeted observations during the investigations with the Frame, (4)
D
revise/refine their initial explanations to incorporate observations from their investigation, and
(5) reflect upon how this new information applies to the context of air mattresses.
TE
3.1.2 Content
Learning objectives for the unit consists of students being able to understand pressure,
EP
volume, and temperature with ideal gases on a molecular level as well as articulate relationships
among pressure, volume and temperature. Students worked through a series of five activities in
WISE (Figure 2) to discover scientific phenomena and to interpret data gathered from the Frame.
Specifically, the first activity provided information about the context of the project (air
C
mattresses). The second activity (“What is a gas?”) guided students through an introductory
activity in the Frame that focuses on how a gas differs from liquids and solids at a particulate
AC
level through a simulation of gases, liquids, and solids. The third activity (“What is
temperature?”) concentrated on helping students understand temperature on a molecular level by
having students add and remove energy to the system by placing hot and cold jars next to the
Frame and observing what happens to the molecules in the simulation. Specific learning
objectives included: 1) gas particles at the same temperature move at different speeds and 2)
adding/removing energy results in an average increase/decrease in the speed of the gas
molecules (as opposed to molecules expanding/contracting in size). Activity 4 focused on
helping students understand gas pressure on a molecular level by guiding them to physically
push and pull on the piston and observe what happens to the gas molecules. Specific learning
ACCEPTED MANUSCRIPT
PT
RI
SC
Figure 2. The WISE Frame gas laws curriculum guided students through an investigation of air
mattresses.
objectives involved understanding that decreasing volume leads to increased pressure due to
U
increased collisions with the container walls, and vice versa. The fifth activity guided students
through an investigation of the relationship between temperature and pressure by having students
AN
add energy to the system by putting a hot jar next to the Frame and having students try to hold
the piston to the same location, and similarly with a cold jar. One purpose of this investigation
was to have students feel that it takes more effort to keep the piston in the same place after
adding energy with a hot jar (increased pressure), and that they have to pull the piston to keep it
M
stable for the cold jar (decreased pressure). Learning objectives included understanding that
increasing (or decreasing) the temperature of the molecules increases (or decreases) the average
speed of the molecules, which results in increased (or decreased) pressure on the piston. In case
D
students finished early, additional “Extra for Experts” activities provided students opportunities
to conduct open investigations of the impact of the number of particles on volume and the effect
TE
of mixture of gases on pressure.

The main purpose in using WISE was to facilitate knowledge integration and guide
student interactions with the Frame, as well as to take advantage of research tools such as data
EP
logging capabilities to gain a rough idea of how much time students spent in each activity/step.
Although WISE offers a wide variety of functionality to support knowledge integration, the
Frame lab primarily used text/information steps (6 out of 32 total steps, excluding 10 extra for
expert steps) and multiple-choice/open response steps (26 steps) that roughly corresponded to a
C
paper-and-pencil lab guide. The main ways the Gas Frame lab differed from a paper-and-pencil
lab were: students were typing instead of writing; one step offered students immediate feedback
AC
on a multiple-choice question covering whether actual gas molecules have colors; in the three
reflect/explain steps students’ original predictions were automatically provided for the students
at the top of the page; and if students started completing one portion of a step with multiple
questions (steps often had combinations of multiple-choice and open response questions), WISE
would automatically remind the students to answer the rest of the questions before going to the
next step. The other steps did not offer any kind of feedback or enhanced capabilities.
The Gas Frame Lab and accompanying WISE curriculum was implemented in one class
before this study was conducted. Classroom observations demonstrated that students could
easily interact with the Frame and use another computer to record results in WISE. Data also
informed continued testing and revision of the curriculum and technologies for this study. For
ACCEPTED MANUSCRIPT
example, the WISE curriculum was revised from guiding students through a strict predict-
observe-explain pattern to include more reflection and explanation of a real-life context of an air
mattress product based on student data and feedback from student and teacher interviews.
3.2 Participants
The participants of this study were eighth grade students enrolled in physical science
courses based on teacher involvement with the project; two classroom teachers volunteered to
PT
participate in this study and supplement instruction with Gas Frame activities. One teacher
taught one honors section, and the other teacher taught three standard courses. The teachers had
been part of the development of the Frame and supporting curricula, including a professional
RI
development workshop that familiarized the teachers with the Frame technology and provided
opportunities for teachers to give feedback on curricula, assessments, and implementation in
their classrooms.
SC
Teachers and students came from one middle school with around 600 students enrolled in
grades 6-8. School student demographics include: 50.6% male, 49.4% female; 19.1% Black,
13.8% Hispanic, 53.9% White; 14.2 % Limited English Proficiency, 46.3% on free or reduced
lunch; 14.6% Students with Disabilities and 12.5% of Students categorized as Gifted.
3.3 Treatment
U
AN
Four classes of students worked throughout two 90-minute class sessions on the Gas
Frame lab. The Frame lab followed traditional class instruction in science topics involving
atomic structure and the properties of gases. Teachers introduced the lab and researchers to the
students. Researchers gave a brief introduction to the lab, the Frame technology, and then guided
M
the students through logging into the project in WISE. Students then went through the project at
their own pace. One to two researchers were present during each class. Students worked in
groups of 2 or 3 chosen by the teacher to complete the Gas Frame lab. Each group had a laptop
D
for WISE and a Frame at their lab station.

TE
3.4 Data Sources

Researchers conducted classroom observations to provide insight into how students
interacted with the Frame. Observations did not follow a specific protocol to keep in the spirit of
EP
the exploratory nature of the study. Field notes were taken by at least two researchers in each
class. WISE also recorded student interaction with each step and activity, providing information
about how long students spent in each step as well as if they revised or refined their answers
during the lab. Embedded data within WISE also included predictions, observations, and
C
explanations that students made during the lab.

Pretests and posttests consisted of six open-ended questions that explicitly asked students
AC
to explain observable gas phenomena at the molecular level (Appendix A). Three open-ended
items asked students to explain temperature (T), the relationship between volume and pressure
(V-P), and the relationship between temperature and pressure (T-P) using the same air mattress
context as the lab (proximal). An additional three open-ended items asked students to explain
similar concepts (P, V-P, T-P) using different real-life contexts such as balloons and tires (distal)
For example, one distal question asks students to explain why it is important to check tire
pressure at various temperatures (T-P).
Pretests were administered to all students in the classes a day before the implementation
of the unit, and identical posttests were administered a week after the unit had been completed.
ACCEPTED MANUSCRIPT
Thus, nine days elapsed between the pretest and the posttest. Teachers covered properties of
other states of matter in subsequent lessons between the Gas Frame Lab and the posttest and did
not provide any additional instruction on gas properties.
3.5 Data Analysis

Researchers compared field notes of classroom observations directly following the Frame
implementation. From the field notes, trends in how the students used the Frame lab were
PT
generated. WISE log data were compiled to compute total time in the project as well as the
duration that each student group spent in each step. Each WISE step was classified as text,
predictions, observations, or explanations and then average duration in each kind of step was
RI
calculated.
Analysis of open-ended items focused on the identification of connections among
students’ ideas using a knowledge integration (KI) analysis approach (Liu, Lee, Hofstetter, &
SC
Linn, 2008). Higher KI scores represent more connections among scientifically relevant ideas.
Pretests and posttests for each of six open-ended questions (Appendix A) were coded using a KI
rubric (see Table 1 for more information as well as sample student responses). As part of the KI
analysis, ideas elicited in the open-ended student responses were classified as normative, partial,
U
alternative, and vague/irrelevant. Normative ideas are correct scientific ideas pertinent to the
question. For example, a student response to a question asking why balloons keep their shape of,
AN
“the gas particles are constantly moving and hit the inside of the balloon” would be classified as
normative. Partial ideas represent student ideas that could be considered normative in other
contexts but are tangential to the targeted explanation. A response to the same question that
states, “gas particles take the shape of the container” would be classified as a partial idea as the
M
student recognizes that gases do not have a definite volume, but the idea is not directly
applicable to balloons keeping their shape. Alternative ideas are scientifically non-normative
ideas elicited by students that are relevant to the question. For example, when asked about the
D
relationship of temperature to particle movement, a student response of, “temperature is the thing
that determines the size of particles” would be classified as an alternative idea. Vague/irrelevant
TE
ideas were attributed to off-task or unclear responses (e.g., “This is fun”).

To illustrate the level of normative ideas that were elicited from pretest to posttest,
student responses were also coded for macroscopic, molecular, or linked macroscopic and
EP
molecular ideas present in the responses. Macroscopic ideas represent statements that contained
observable elements with no explicit mention of the particulate level. For example, “balloons
keep their shape because they are filled with helium” was coded as macroscopic. Molecular
ideas represent statements that contained particulate elements. A response that contained, “the
C
particles in a gas go all different directions and fast” was coded as molecular. Macroscopic-
molecular ideas represent statements connecting molecular to macroscopic levels. For example,
AC
“the helium and particles make the balloon stay round” was coded as macroscopic-molecular.
The codes are not mutually exclusive, so one student response could contain multiple types of
ideas. The above examples all came from the same student response, so that response would be
coded as having one macroscopic idea, one molecular idea, and one molecular-macroscopic idea.
ACCEPTED MANUSCRIPT
Table 1. Adapted Knowledge Integration Rubric (from Liu et al., 2008)
KI Level Score Response characteristics Examples Using Student Work
Complex- 5 Students elicit and connect When it was day and it was hot the
PT
link three or more normative and molecules moved faster hitting the
relevant scientific ideas walls inflating the mattress a little.
During the night the slowed down
deflating it a little.
RI
Full-link 4 Students elicit and connect two The particles cooled & slowed down
normative and relevant creating less force which created less
scientific ideas pressure deflating the mattress.
SC
Partial- 3 Students elicit normative and It deflated because it was colder, when
link relevant scientific ideas as well the particles become cooled they
as non-normative ideas become smaller and slower.
No-link 2 Students elicit non-normative There was enough molecules during the
U
ideas or make invalid day so it was inflated then at night they
connections with non- bunched together and it deflated.
AN
normative ideas
Irrelevant 1 Students do not elicit scientific I don't know
ideas
No 0 No response is provided
M
response
Two independent raters coded all data. Cohen’s Kappa ranged from .670 to .756,
indicating substantial inter-rater reliability (Cohen, 1968). Students were assigned overall KI
D
scores based on the number of scientifically normative ideas and the connections among ideas in
their responses. Student responses were classified as Complex-Link, Full-Link, Partial-Link,
TE
No-Link, Irrelevant, and No Response per the KI rubric using a 0-5 point scale (Liu, Lee, & Linn,
2011) in order to address the first research question. Paired-sample t-tests were used to
determine any statistically significant differences in student performance from pretest to posttest
EP
for individual items. Cohen’s d (1988) was calculated to estimate the respective effect size of
each open-ended question analyzed. Chi-squared tests were used to determine any significant
shifts in categories of student ideas from pretest to posttest. Researchers examined the content of
the embedded predictions, observations and explanations in WISE to complement pre/post data.
C
As this study was conducted in an authentic school context, we had missing data.
Although 77 students were enrolled in the four classes, 8 students were absent during the labs, 6
AC
students were absent during either the pretest or the posttest and 18 students did not return IRB
consent forms, resulting in a final sample of 45 (17 male; 28 female) with 14 from the honors
and 31 from the standard physical science classes. We compared pretest scores for students with
missing and complete data and found no statistically significant differences between groups. We
present results for students who completed both a pretest and posttest assessment. The missing
data is also reflected in the log data, which resulted in a final number of 17 student groups across
the three classes (14 groups of 2, 3 groups of 3). We compared log data from students with
complete and missing pretest and posttest data (as students missing either pretests or posttests
ACCEPTED MANUSCRIPT
were removed from final analysis) and found no statistically significant differences between
groups in terms of duration.
4. Results
4.1 Overall Usage

The middle school students had no difficulty operating the Frame or using a separate
PT
laptop concurrently for WISE. Students did not seem surprised that they could control the Frame
using the physical controls of hot jars or spring. Students were generally very engaged and
interacted with the Frame naturally, without questions as to how the physical inputs of the Frame
RI
controlled the simulation. Most groups had one student who would mainly read/record in WISE
and the other student(s) would primarily conduct the investigations with the Frame. Teachers
directed students to switch roles on the second day.
SC
Students spent an average of 110 minutes (SD=17.7) within the WISE project, ranging
from 84 minutes to 143 minutes. Groups spent an average of 3.5 minutes per step (SD=1.2).
Students tended to spend most of the time in the second and fourth activity (average time per
step (min): activity 1: M(SD) = 2.0(0.7); activity 2: M(SD) = 11.8(11.9); activity 3: M(SD) =
U
3.2(0.9); activity 4: M(SD) = 5.1(2.3); activity 5: M(SD) = 2.5(1.3)). All groups completed all
five activities in WISE, with 8 of the student groups getting through part of the extra for experts
AN
sections. All student groups interacted with WISE over 2 days. If students finished early,
teachers directed the students to work on homework during the rest of the block period.
On average, students tended to spend less time on text pages (M(SD) =1.54(0.8)), and
more time in prediction (M(SD )=4.5(3.0)), observation (M(SD) =4.7(1.3)), and explanation
M
steps (M(SD )=4.9(2.5)). There was a significant effect of the type of step on duration (F(3, 64)
= 9.61, p < 0.001). Tukey HSD post hoc comparisons indicate that the duration of text steps
significantly differed from prediction, observation, and explanation steps, but there were no
D
significant differences in the duration among prediction, observation and explanation steps.
Classroom observations aligned with the log data that students spent less time on the
TE
informational text pages and more in prediction, observation, and explanation steps that guided
students’ interactions with the Frame.
Classroom observations also indicated that in general, students conducted investigations
EP
with the Frame as guided in WISE. Many student groups (typically half of the class)
spontaneously used the Frame to test their predictions before being specifically guided in WISE,
and then went back to record their observations and explanations. Some student groups (2-3 per
class) also spontaneously interacted with the Frame, conducting multiple trials of investigations
C
like moving the hot jar back and forth towards the Frame, or conducting more open
investigations like holding in the piston while the hot jar was against the Frame. Observations
AC
revealed that a few students across classes would continuously hold the spring and manipulate
the piston even during discussion of the activities or without looking at the screen of the Frame.
4.2 Pretest and Posttest Performance

Posttest scores were regressed with pretest scores and class (honors or standard) as
explanatory variables to determine any differences in performance across classes (R2 = 0.47, F(2,
42) = 18.75, p < 0.01). Pretest score had a significant effect on posttest scores (β = .96, t(42) =
6.03, p < 0.01). After controlling for pretest score, there were no significant differences between
honors and standard classes (β = .72, t(42) = .76, p = .45). Honors students (Pretest M(SD) =
ACCEPTED MANUSCRIPT
2.25(.51); Posttest M(SD) = 2.82(.71)) performed similarly to standard students (Pretest M(SD) =
2.23(.38); Posttest M(SD) = 2.65(.52)). As there were no differences between honors and
standard, there were also no differences between teachers. We report the rest of the results
without differentiating classes.
Overall, students made significant progress from pretest to posttest with a large effect
size. Table 2 displays pretest and posttest KI scores and effect sizes for each question. On
average, students made progress from no-link to partial links along the KI scale. Students made
PT
significant progress on proximal items that asked students to explain temperature, the
relationship between pressure and volume and the relationship between temperature and pressure
in the same context as the unit. Students also made progress on distal items that asked students
RI
to explain the relationship between pressure and volume as well as temperature and pressure in a
different context than that of the unit. However, scores did not significantly increase on the item
that asked students to explain pressure in a different context.
SC
Table 2. Means and standard deviations for individual pretest and posttest KI items and
overall average score
Pretest
U Posttest Effect 2-tailed
AN
Question Mean(SD) Mean(SD) t df Size p-value
Proximal Temperature 2.49 (.59) 2.87 (.76) 3.39 44 .56 < .01*
(same Pressure-Volume 2.13 (.34) 2.36 (.61) 3.55 44 .47 < .01*
context)
M
Temperature-
Pressure 2.16 (.67) 2.82 (.83) 4.94 44 .88 < .01*
Distal Pressure 2.13 (.81) 2.44 (.87) 1.76 44 .37 .08
(different
D
Pressure-Volume 2.47 (.94) 2.91 (.82) 2.88 44 .50 < .01*

context)
Temperature- 2.80
TE
Pressure 2.07 (.94) (1.32) 4.12 44 .64 < .01*

Overall average KI score* 2.24 (.42) 2.70 (.58) 7.02 44 .91 < .01*
*Note: Overall average KI score is the average across multiple items
EP
4.3 Developing Molecular-level Explanations

In total, posttest responses included more ideas about molecular-level behavior and more
ideas that linked macroscopic to molecular levels, χ2(1, N = 44) = 11.76, p = .003 (Table 3). On
the proximal temperature question, responses tended to contain more macro-molecular links on
C
the posttest. The question asks students to explain what happens to gas particles in an air
mattress at different temperatures. Student responses on the pretest commonly consisted of
AC
descriptions of “air” without explicit discussion of the particulate level (“The mattresses was
[sic] to [sic] full with air and it got so hot that it was abuot [sic] to deform”) whereas on the
posttest students articulated particle-level behavior and connected it to the mattress (“When the
mattress was hot the particles were moving really fast. But when it cooled down the particles
started slowing down”).
Embedded data from the Gas Frame Lab provides evidence that students were observing
the behavior of gas molecules with the Frame and making connections to temperature. For
example, one of the first steps within the temperature activity asked students to make
observations of gas molecules at the same temperature. All but one response contained
ACCEPTED MANUSCRIPT
observations that the gas particles were in constant motion and were colliding with other
molecules and the walls of the container. When students were asked to predict, observe and
explain what happened to the gas molecules when they increased the temperature of the gas with
the hot jar, all students’ observations included that the speed of the gas molecules increased. In
the responses to the embedded reflection prompt that guided students to connect what happens
inside an air mattress when heated and cooled, all but one student group explicitly referred to
particle movement to explain the results.
PT
For the relationship between temperature and pressure, posttest responses included more
molecular ideas and more macro-molecular links. On the proximal temperature-pressure pretest
and posttest question, student responses included more molecular ideas on the posttest. This
RI
question asked students to make the connection from hot or cold temperatures to how that would
translate to pressure differences in an air mattress. For instance, one student responded, “The
cold air caused it to deflate because it became colder” on the pretest and on the posttest
SC
responded, “The air particles in the mattress slowed down and it deflated.” On the distal
temperature-pressure question, posttest responses shifted to include more macro-molecular links.
The question asked students to explain the relationship between temperature and pressure on the
microscopic level in the context of tire pressure on hot and cold days. On the pretest, one student
U
responded that tires need to be checked “so the tire doesn't explode or inplode [sic].” The
posttest response from the same student stated, “In the summer the preassure [sic] builds up more
AN
because the particles move faster, in the winter the particles move slower causing the preassure
to go down.”
Embedded data indicate that the Gas Frame lab helped students make progress in
connecting the effect of temperature on pressure on a molecular level. In activity 5, students
M
were guided to predict, observe and explain what happens to pressure when temperature is
decreased. In the predictions, 8 of the student groups incorrectly predicted that it would become
harder to push in the piston if the temperature was decreased. After conducting the investigation,
D
5 of the 8 groups revised their explanations stating that it was easier to push the piston when the
temperature decreased, with three student groups still stating that it was harder to push.
TE
Although only seven of the student groups included a particle explanation in their explanation of
how temperature relates to pressure, 13 of the 17 student groups used particulate explanations on
the reflection prompt to connect the results to what would happen to the pressure of an air
EP
Table 3. Average number of normative macroscopic, molecular, and macro-molecular

ideas by student response and total number of normative ideas for pretest and posttest
items
C
Macro-
AC
Macroscopic Molecular Molecular Total No. of Ideas

Question Pre Post Pre Post Pre Post Pre Post
Proximal Temperature 0.24 0.27 0.00 0.04 1.04 1.31 58 73
P-V 0.60 0.62 0.00 0.00 0.40 0.38 45 45
T-P 0.33 0.20 0.24 0.82 0.04 0.04 28 46
Distal Pressure 0.64 0.31 0.49 0.53 0.09 0.11 55 43
P-V 0.38 0.44 0.13 0.24 0.27 0.31 35 45
T-P 0.44 0.51 0.02 0.09 0.60 1.07 48 75
Total No. of Ideas 119 106 40 78 110 143 269 327
ACCEPTED MANUSCRIPT
mattress at different temperatures.

For questions relating to pressure and volume, the results were mixed. On the distal
pressure-volume question, posttest responses contained slightly more molecular-level ideas. The
question asked students to explain the relationship between pressure and volume in the context of
a small balloon filled with air. Many students’ pretest responses attributed macroscopic
properties to the molecular level, such as “this happens because it becomes harder to all of the
particles to fit” or “because the air does not release out of the balloon.” Student responses on the
PT
posttest represented a shift towards explicit explanations of particulate behavior, for instance,
“the atoms are moving at faster speeds in a smaller area, causing the balloon to be harder and
harder to squeeze.” This response, while not a completely normative link between pressure and
RI
volume, indicates that the student was adding ideas about molecular behaviors. On the proximal
pressure-volume question, there was little shift in the kinds of ideas from pretest to posttest. On
the distal pressure question, there was a decrease in macroscopic ideas and little change in
SC
molecular and macro-molecular ideas from pretest to posttest and even a decrease in total
normative ideas elicited.
In activity 4 of the lab, students were explicitly guided to decrease the volume of the
container by pushing in the piston to learn about pressure and volume. Although the lab
U
provided specific questions to help students pay attention to how often the gas particles hit the
piston wall, embedded data reveal that 4 out of the 17 student groups did not observe that as they
AN
pushed in the piston, the gas particles hit the piston more often. Of those four groups, two
groups noted that they observed the pressure decreasing instead of increasing.
Interestingly, on the embedded reflection question that asked students to connect the
results to air mattresses by explaining how an air mattress can support weight, 11 out of the 17
M
student groups had responses that used molecular-level ideas (i.e., “The air mattress is able to
support weight because the air mattress is a closed space that the particles move around in. When
someone lays down on it, it compresses the air mattress and the gas particles inside push against
D
the wall of the air mattress, which hold you up. "). However, responses to the reflection prompt
that asked students to connect to what happens when an air mattress bursts (the proximal P-V
TE
question), only one group used molecular-level ideas to explain the result. All other student
groups used macroscopic-level explanations, such as “The air in the mattress couldn't hold that
many people cause they was too heavy” or “Because there is a lot more air and weight it will
EP
compress it more and POP!”
4.4 Refining Alternative Ideas

Table 4 shows the percentages of normative, partial, alternative, and vague/irrelevant
C
ideas in students’ pretest and posttest responses. Overall, there was a significant shift in
categories of responses from pretest to posttest, χ2(1, N = 44) = 42.88, p < .001. Across all
AC
questions except 1, students had fewer alternative ideas in their explanations and more normative
ideas from pretest to posttest. Partial ideas expressed by students decreased from pretest to
posttest, with the exception of question 4. There are no notable trends for the expression of
vague and irrelevant ideas.
For example, the proximal T-P question asks students to consider why a mattress, inflated
when it was hot, would become deflated at night. Students’ pretest responses consisted
predominantly of alternative ideas (e.g. “the particals [sic] got closer together and slowed down”
or “the particles were harder when it was hot but once it cooled down they became softer).
Many of these alternative ideas expressed a mismatch between observed macroscopic properties
ACCEPTED MANUSCRIPT
Table 4. Average number of normative, partial, alternative and vague/irrelevant ideas by student
response and total number of ideas for pretest and posttest items
Normative Partial Alternative Vague/Irrelevant Total No. of Ideas

Question Pre Post Pre Post Pre Post Pre Post Pre Post
Proximal T 1.24 1.56 0.04 0.07 0.36 0.49 0.18 0.18 82 98
P-V 0.04 0.18 0.96 0.82 0.09 0.04 0.04 0.18 51 55
PT
T-P 0.27 0.87 0.36 0.16 0.60 0.29 0.11 0.22 62 69
Distal P 0.31 0.47 0.89 0.44 0.18 0.18 0.02 0.04 66 51
P-V 0.54 0.78 0.36 0.22 0.11 0.13 0.11 0.11 46 56
T-P 0.44 1.29 0.62 0.38 0.36 0.24 0.27 0.18 79 96
RI
Total No. of
Total Ideas 123 231 145 94 76 62 33 36 386 425
SC
and molecular phenomena. In both examples provided, the inflation/deflation of the mattress
and its respective “hardness” or “softness” was attributed to physical proximity or physical
characteristics of particles, neither accurately representing the phenomenon on a microscopic
U
level. By the post-test, students demonstrated a more scientifically normative understanding (e.g.
“the particles stopped moving as fast when it cooled down” and “the temp. made the molecules
AN
speed up which caused the molecules to constantly hit the container, but once it cooled down, the
molecules slowed down not hitting the container as much”). Students’ posttest ideas focus more
on particulate movement and temperature and less upon alternative ideas that were common in
the pretest.
M
Embedded data indicate changes in students’ initial ideas during the lab. For example in
the temperature activity, seven of the student groups incorrectly predicted that molecules would
gain mass when temperature increases. However after their observations with the Frame, all of
D
the student groups explained that molecules do not gain mass but increase in speed when
temperature increases. In the T-P activity, when 8 of the student groups on their predictions
TE
stated that pressure would increase if temperature decreased, many explanations of the
predictions misapplied macroscopic properties to molecular behavior (“It’s going to be harder
because it going to be stuck and go slower”), indicating initial ideas that somehow gas molecules
EP
at decreased speeds relates to the gas being more viscous or harder to move. However, by the
end of the activity 5 of the 8 student groups had revised their explanations to state that when
temperature decreases pressure would decrease.
C
5. Discussion
This study investigated how combining physical and virtual experiences into
AC
simultaneous, augmented virtual science laboratories could help students build upon intuitive
ideas and develop molecular-level explanations of macroscopic phenomena. The findings
suggest that the Gas Frame Lab helped middle school students develop an overall understanding
of gas behavior as evidenced by a significant increase from pretest to posttest KI scores with a
large effect size (0.91). As the KI scores represent the development of explanations that apply
scientific ideas to explain real-world phenomena, results also suggest that the Gas Frame Lab
helped students make progress in the NGSS practice of constructing explanations.
Results align with other studies that demonstrate benefit from visualizations of molecular
phenomena (e.g., Steiff & Wilensky, 2003; Wu, Krajcik, & Soloway, 2001) as well as other
research that demonstrates the combination of physical and virtual materials benefits students
ACCEPTED MANUSCRIPT
(e.g., Blikstein, 2014; Liu, 2006; Olympiou & Zacharia, 2012). Results are comparable to other
studies that use WISE to support inquiry around molecular concepts. For example, Zhang &
Linn (2011) investigated middle school students’ (n=133) understanding of chemical reactions
through a unit on hydrogen fuel cells lasting 5 instructional days (total of around 300 minutes).
They found an effect size of 1.58 from students’ average pretest to posttest KI scores. Similarly,
Chang and Linn (2013) found effect sizes of 0.57 (n=60), 0.63 (n=62), and 1.21 (n=83) from
pretest to posttest KI scores for three studies of middle school students using a weeklong WISE
PT
unit on thermodynamics. Chiu and Linn (2014) found effect sizes of .99 (n=21) and 0.53 (n=24)
for two studies of high school students using a weeklong WISE unit to learn about chemical
reactions. The effect size of 0.91 found for the Gas Frame Lab is similar to if not more favorable
RI
than these studies, considering that the Gas Frame lab was roughly half of the length of
instruction and assessments were administered at similar times across all studies.
Pretest to posttest improvements in KI score demonstrate that students progressed in
SC
constructing explanations that used molecular-level ideas of gas particles and refined partial and
alternative ideas for concepts of temperature and relating temperature to pressure. During the lab,
students were guided to use hot and cold jars to control the temperature of the Frame, and feel
the effect of the different temperatures on pressure by controlling the piston through the spring.
U
Embedded data suggest that students were able to use these Frame activities to add normative
molecular ideas as well as refine alternative ideas about temperature and the relationship
AN
between temperature and pressure.
However, students did not make as much progress in articulating explanations that added
molecular-level ideas or macro-molecular connections for pressure and the relationship between
pressure and volume. Improvements in KI score were generally due to the increase in normative
M
ideas and refinement of partial/alternative ideas. Students tended to have explanations at

macroscopic levels that demonstrated a general relationship that as volume decreases, pressure
increases, and vice versa. For the pressure-volume activity, the Frame lab guided students to
D
decrease the volume of the container by pushing on the spring. Interestingly, students tended to
spend more time in the pressure-volume activity as compared to the temperature and
TE
temperature-pressure activities. Although most students stated they observed increased

collisions on the container wall when decreasing the volume of the container, when explaining
related situations of gas behavior in air mattresses or balloons, students tended not to use
EP
molecular-level explanations, remaining at the macroscopic level. One possible explanation

could be that students have robust experiential ideas for explaining everyday phenomena
(diSessa, 1998; 1993) and the particular assessment item contexts of an air mattress bursting or
squeezing a balloon failed to elicit students’ understanding of pressure on a molecular level,
C
instead eliciting revision of their explanations on a macroscopic level.

One major goal of this study was to explore if augmented virtual technologies could help
AC
students learn in authentic classrooms settings. Across classes, students had very little difficulty
using the Frame and understanding how the physical inputs controlled the simulation. As the
Frame is a novel technology and very few studies address mixed-reality approaches in
classrooms (Lindgren & Johnson-Glenberg, 2013), this study indicates that augmented virtual
approaches can be successful in real classroom contexts.
Results from this exploratory study also point to the need to explore how particular
combinations of specific physical controls and visualizations for particular content can affect
student learning in authentic settings. We predicted that embodied experiences with physical
controls could help students elicit relevant intuitive knowledge at the same time as adding
ACCEPTED MANUSCRIPT
normative molecular-level ideas through visualizations. Having both sets of ideas activated at
the same time could encourage students to connect molecular and macroscopic ideas and refine
everyday ideas. However, the results suggest that certain combinations of augmented virtual
interactions were more successful for particular concepts. The combination of using a hot and
cold jar to control the temperature of the simulation as well as the spring to control the piston
seemed to help students make molecular-level explanations and refine alternative ideas about
temperature and the effect of temperature on pressure. The combination of students using the
PT
spring to understand pressure and the effect of volume on pressure was not as successful at
helping students develop molecular-level explanations, but instead somewhat helped students
add normative and refine ideas at the macroscopic level. Future work that articulates exactly
RI
how students interpret and interact with specific combinations of physical and virtual
manipulatives for specific concepts can contribute to an understanding of when and why
augmented virtual approaches can be successful.
SC
Similarly, the Frame was developed to help students connect real-world gas phenomena
to unobservable molecular behaviors. The design and revisions of the technology were guided
by the overall goal for use in authentic classrooms with teachers playing a large role in the
design and refinement of the technologies. Although our augmented virtual lab proved to be
U
generally beneficial for learning with these specific learning goals, designers may need to tailor
or choose mixed-reality approaches based on targeted learning outcomes. For example, our
AN
activities did not emphasize scientific practices of modeling, where having virtual and physical
representations side-by-side may have the most benefit (e.g., Blikstein, 2014). However, our
study does provide guidance for other mixed-reality approaches that aim to help students connect
molecular and macroscopic levels in science.
M
Results also suggest implications for practitioners teaching or developing curricula for
gas laws. Similar to other studies (e.g., de Jong, Linn, & Zacharia, 2013; Liu, 2006), results
indicate that providing both hands-on and virtual experiences can support science learning.
D
Since many teachers already incorporate probeware into labs on gas laws, teachers can consider
providing complementary visualizations and simulations to maximize benefit to students.
TE
6. Limitations
This study is based on a pre/post design without a control group; therefore these results
EP
serve as indications that the Frame lab can be successfully implemented in authentic classrooms.
As the pretest and posttest were identical, test or practice effects may contribute to the findings
of this study. Additionally, the treatment in this study involved two 90-minute class periods.
More research is necessary to determine the extent of the impact that sensor-augmented virtual
C
experiences have on student understanding after a longer duration, as well as investigating the
role of the teacher in these labs.
AC
As this study focused on the development of explanations, assessment items focused on

text-based open responses. However, research demonstrates that drawing tasks can provide
valuable insight into how students understand molecular concepts (Marbach-Ad, Rotbain, &
Stavy, 2008; Zhang & Linn, 2011). Future work should include both drawing and explanation
tasks to better understand how augmented approaches might help students further develop
understanding of complex molecular concepts.
Another major limitation to this study involves the relative impact of the physical and
virtual manipulatives, and similarly, the influence of embodied experiences on student learning.
As this was an exploratory study to determine if these kinds of approaches can be successful in
ACCEPTED MANUSCRIPT
real classrooms, this study highlights the need for more detail as to if or how embodied
experiences can contribute to learning in authentic settings. Although we used a grounded
knowledge integration framework to shape this exploratory work, this study only provides
evidence that embodied approaches can be as successful as other technology-enhanced
visualization-based inquiry labs. It is possible that the gains from pretest to posttest could be
attributed solely to students interacting with molecular-level visualizations and the physical
controls could have a small or negligible effect. As other meta-analyses demonstrate moderate
PT
effects of visualizations on science understanding (e.g. Höffler & Leutner, 2007), future research
needs to compare student learning outcomes from only visualizations to augmented approaches
to isolate what, if any, benefits there are grounded or embodied experiences. Likewise, more
RI
fine-grained, qualitative studies of students working with these technologies can provide needed
insight into exactly if and how students interpret and use the physical augmentations for learning.
SC
7. Conclusion
Augmented virtual science laboratories have the potential to help students develop
sophisticated scientific explanations as called for by the NGSS by connecting molecular-level
U
visualizations to observable, macroscopic phenomena. This paper investigated if students in
authentic classrooms could use the Frame, an augmented virtual technology that uses physical
AN
inputs to control molecular simulations simultaneously, to develop molecular-level explanations
and refine alternative ideas about gas properties. Overall results indicate students were able to
make more connections among ideas as well as refine alternative to normative ideas about gas
phenomena from pretest to posttest, but these results differed for different topics. This study
M
contributes to the field of mixed-reality technologies by providing evidence that providing

simultaneous hands-on and virtual experiences can promote science learning.
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
References
AAAS Project 2061 (n.d.) [Pilot and field test data collected between 2006 and 2010].
Unpublished raw data.
Abrahamson, D., Gutiérrez, J. F., Charoenying, T., Negrete, A. G., & Bumbacher, E.
(2012). Fostering hooks and shifts: tutorial tactics for guided mathematical
PT
discovery. Technology, Knowledge, and learning, 17(1-2), 61-86. DOI: 10.1007/s10758-
012-9192-7.
Anastopoulou, S., Sharples, M., & Baber, C. (2011). An evaluation of multimodal interactions
RI
with technology while learning science concepts. British Journal of Educational
Technology, 42(2), 266-290.
Ardac, D., & Akaygun, S. (2004). Effectiveness of multimedia‐based instruction that emphasizes
SC
molecular representations on students' understanding of chemical change. Journal of
Research in Science Teaching, 41(4), 317-337.
Barsalou, L. W. (2008). Grounded cognition. Annual Review of Psychology, 59, 617–45.
doi:10.1146/annurev.psych.59.103006.093639
U
Bell, R. L., & Trundle, K. C. (2008). The use of a computer simulation to promote scientific
conceptions of moon phases. Journal of Research in Science Teaching, 45, 3, 346-372.
AN
Benson, D. L., Wittrock, M. C., & Baur, M. E. (1993). Students’ preconceptions of the nature of
gases. Journal of Research in Science Teaching, 30(6), 587-597.
Ben-Zvi, R., Eylon, B-S, & Silberstein, J. (1986). Is an atom of copper malleable? Journal of
M
Chemical Education, 63, 64-66.

Blikstein, P. (2012, October). Bifocal modeling: a study on the learning outcomes of comparing
physical and computational models linked in real time. In Proceedings of the 14th ACM
international conference on Multimodal interaction (pp. 257-264). ACM.
D
Blikstein, P. (2014). Bifocal modeling: Promoting authentic scientific inquiry through exploring
and comparing real and ideal systems linked in real-time. In A. Nijholt (Ed.) Gaming
TE
Media and Social Effects. Singapore: Springer.

Blikstein, P., Fuhrmann, T., Greene, D., & Salehi, S. (2012). Bifocal modeling: mixing real and
virtual labs for advanced science learning. In Proceedings of the 11th International
EP
Conference on Interaction Design and Children (pp. 296-299). ACM.

Blikstein, P., & Wilensky, U. (2007). Bifocal modeling: a framework for combining computer
modeling, robotics and real-world sensing. Paper presented at the Annual meeting of the
American Educational Research Association. Chicago, IL.
C
Bodner, G. (1991). I have found you an argument: The conceptual knowledge of beginning
chemistry graduate students. Journal of Chemistry Education, 68, 385-388.
AC
Borgman, C. L., Abelson, H., Dirks, L, Johnson, R., Koedinger, K.R., Linn, M.C., et al. (2008).
Fostering learning in the networked word: The cyberlearning opportunity and challenge.
Report of the NSF Task Force on Cyberlearning. Washington, D.C.: National Science
Foundation.
Bouwma-Gearhart, J., Stewart, J. & Brown, K. (2009). Student misapplication of a gas-like
model to explain particle movement in heated solids: Implications for curriculum and
instruction towards students’ creation and revision of accurate explanatory models.
International Journal of Science Education, 31(9), 1157-1174.
ACCEPTED MANUSCRIPT
Bransford, J., Brown, A. L., & Cocking, R. R. (2000). How people learn. Washington, DC:
National Academy Press.
Carlsen, D. D., & Andre, T. (1992). Use of a microcomputer simulation and conceptual change
text to overcome student preconceptions about electric circuits. Journal of Computer-
Based Instruction, 19, 4, 105-9.
Chang, H. Y., Quintana, C., & Krajcik, J. S. (2010). The impact of designing and evaluating
molecular animations on how well middle school students understand the particulate
PT
nature of matter. Science Education, 94(1), 73-94.
Chao, J., Chiu, J. L., Pan, E., DeJaegher, C., Hazzard, E. & Xie, C. (2014). The Effects of
Mixed-Reality Laboratories on High School Students’ Conceptual Understanding of Gas
RI
Laws. Paper presented at the Annual Meeting of the American Educational Research
Association, Philadelphia, PA.
Chiu, J. L. (2010). Developing students’ criteria for visualizations by prompting judgments of
SC
fidelity. The International Conference of the Learning Sciences. Chicago, IL.
Chiu, J. L., King Chen, J. & Linn, M. C. (2013). Overcoming deceptive clarity by encouraging
metacognition in the web-based inquiry science environment. In R. Azevedo & V.
Aleven (Eds.) International Handbook of Metacognition and Learning Technologies (pp.
U
517-531). New York: Springer.
Chiu, J. L. & Linn, M. C. (2012). Supporting self-monitoring with dynamic visualizations. In J.
AN
Dori & A. Zohar (Eds.) Metacognition and Science Education (pp. 133-164). Mahwah,
NJ: Lawrence Erlbaum.
Chiu, J. L. & Linn, M. C. (2014). Supporting knowledge integration in chemistry with a
visualization-enhanced inquiry unit. Journal of Science Education and Technology, 23(1),
M
37-58.
Clark, D. B. (2006). Longitudinal conceptual change in students' understanding of thermal
equilibrium: An examination of the process of conceptual restructuring. Cognition and
D
Instruction, 24(4), 467–563.

Clark, D. B., & Jorde, D. (2004). Helping students revise disruptive experientially-supported
TE
ideas about thermodynamics: Computer visualizations and tactile models. Journal of

Research in Science Teaching, 41(1), 1–23.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:
EP
Lawrence Earlbaum Associates.

de Jong, T., Linn, M. C., & Zacharia, Z. C. (2013). Physical and Virtual Laboratories in Science
and Engineering Education. Science, 340(6130), 305-308.
Dori, J.Y., & Belcher, J. (2005). How does technology-enabled active learning affect
C
undergraduate students’ understanding of electromagnetism concepts? Journal of the

Learning Sciences, 14, 243-279.
AC
Duschl, R. & Gitomer, D. (1991). Epistemological perspectives on conceptual change:

Implications for educational practice. Journal of Research in Science Teaching, 28, 839-
858.
Duschl, R. A., Schweingruber, H. A., & Shouse, A. E. (Eds.). (2007). Taking science to school:
Learning and teaching science in grades K-8. Washington, DC: National Academies
Press.
Erickson, G., & Tiberghien, A. (1985). Heat and temperature. In R. Driver, E. Guesne, &
Tiberghien (Eds.), Children’s ideas in science (pp. 52–83). Philadelphia: Open
University Press.
ACCEPTED MANUSCRIPT
Feisel, L.D., & Rosa, A.J. (2005). The role of the laboratory in undergraduate engineering
education. Journal of Engineering Education, 94, 121-130.
Finkelstein, N., Adams, W., Keller, C., Kohl, P., Perkins, K., Podolefsky, N., Reid, S., et al.
(2005). When learning about the real world is better done virtually: A study of
substituting computer simulations for laboratory equipment. Physical Review Special
Topics - Physics Education Research, 1(1), 1–8.
Gabel, D. (1999). Improving teaching and learning through chemistry education
PT
research: A look to the future. Journal of Chemical Education, 76 (4), 548.
Gabel, D., Samuel, K. & Hunn, D. (1987). Understanding the particulate nature of matter.
Journal of Chemical Education, 64(8), 695-697.
RI
Gire, E., Carmichael, A., Chini, J.J., Rouinfar, A., Rebello, S., Smith, G., et al. (2010). The
effects of physical and virtual manipulatives on students’ conceptual learning about
pulleys. In K. Geomez, L. Lyons, & J. Radinsky (Eds.) Learning in the disciplines:
SC
Proceedings of the 9th international conference of the learning sciences (ICLS 2010) (Vol.
1, pp. 937-944).
Herrmann-Abell, C. F., & DeBoer, G. (2011). Using distractor-driven standards-based multiple-
choice assessments and Rasch modeling to investigate hierarchies of chemistry
U
misconceptions and detect structural problems with individual items. Chemistry
Education Research and Practice, 12, 184-192.
AN
Hodson, D. (1996). Laboratory work as scientific method: Three decades of confusion and
distortion. Journal of Curriculum Studies, 28(2), 115-135.
Höffler, T. N., & Leutner, D. (2007). Instructional animation versus static pictures: A met-
analysis. Learning and Instruction, 17, 722-738.
M
Hofstein, A., & Lunetta, V. (2004). The laboratory in science education: Foundations for the
twenty-first century. Science Education, 88(1), 28-54.
Honey, M. A., & Hilton, M. L. (Eds.). (2011). Learning science through computer games and
D
simulations. Washington, DC: National Academy Press.

Jaakkola, T., Nurmi, S., & Veermans, K. (2011). A comparison of students' conceptual
TE
understanding of electric circuits in simulation only and simulation-laboratory contexts.

Journal of Research in Science Teaching, 48, 1, 71-93.
Johnson-Glenberg, M. C., Koziupa, T., Birchfield, D., & Li, K. (2011, June). Games for learning
EP
in embodied mixed-reality environments: principles and results. In Proceedings of the 7th

International Conference on Games, Learning, & Society (pp. 129-137). ETC Press.
Johnstone, A. H. (1991). Why is science difficult to learn? Things are seldom what they seem.
Journal of Computer Assisted Learning, 7, 75-83.
C
Jones, L. L., Jordan, K. D. & Stillings, N. A. (2005). Molecular visualization in chemistry

education: the role of multidisciplinary collaboration. Chemistry Education Research and
AC
Practice, 6(3), 136-149.

Jones, M. G., Carter, G., & Rua, M. J. (2000). Exploring the development of conceptual
ecologies: Communities of concepts related to convection and heat. Journal of Research
in Science Teaching, 37(2), 139–159.
Kelly, R. M. & Jones, L. L. (2007). Exploring how different features of animations of sodium
chloride dissolution affect students’ explanations. Journal of Science Education and
Technology, 16(5), 413-429.
Kozma, R. (2003). The material features of multiple representations and their cognitive and
social affordances for science understanding. Learning and Instruction, 13(2), 205-226.
ACCEPTED MANUSCRIPT
Kozma, R. B., & Russell, J. (1997). Multimedia and understanding: Expert and novice responses
to different representations of chemical phenomena. Journal of Research in Science
Teaching, 34(9), 949-968.
Krajcik, J. (1991). Developing students' understandings of chemical concepts. In S. Glynn, R.
Yeany, & B. Britton (Eds.), The psychology of learning science (pp. 117-147). Hillsdale,
NJ: Erlbaum.
Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago: University of Chicago Press.
PT
Lee, O., Eichinger, D. C., Anderson, C. W., Berkheimer, G. D. & Blakeslee, T. D. (1993).
Changing middle school students’ conceptions of matter and molecules. Journal of
RI
Levy, D. (2012). How dynamic visualization technology can support molecular reasoning.
Journal of Science Education and Technology, 22(5), 702-717.
Levy, S. T., & Wilensky, U. (2009). Students’ learning with the Connected Chemistry (CC1)
SC
curriculum: navigating the complexities of the particulate world. Journal of Science
Education and Technology, 18(3), 243-254.
Levy, S. T., Novak, M., & Wilensky, U. (2006). Students’ foraging through the complexities of
the particulate world: Scaffolding for independent inquiry in the connected chemistry
U
(MAC) curriculum. In Annual Meeting of the American Educational Research
Association, San Francisco, CA.
AN
Lewis, E., & Linn, M. C. (1994). Heat energy and temperature concepts of adolescents, adults,
and experts: Implications for curricular improvements. Journal of Research in Science
Teaching, 31(6), 657-677.
Lin, H., Cheng, H., & Lawrenz, F. (2000). The assessment of students and teachers’
M
understanding of gas laws. Journal of Chemical Education, 77(2), 235-238.

Lindgren, R. & Johnson-Glenberg, M. (2013). Emboldened by embodiment: Six precepts for
research on embodied learning and mixed reality. Educational Researcher, 42(8), 445-
D
452.
Linn, M. C. (2005). WISE design for lifelong learning—Pivotal Cases. In Peter Gärdenfors and
TE
Petter Johansson (Eds.) Cognition, Education and Communication Technology. Mahwah,

NJ: Lawrence Erlbaum Associates.
Linn, M. C., Davis, E. A., & Eylon, B. –S. (2004). The scaffolded knowledge integration
EP
framework for instruction. In M. C. Linn, E. A. Davis, & P. Bell (Eds.), Internet

environments for science education (pp. 73-83). Mahwah, NJ: Erlbaum.
Linn, M. C., & Eylon, B. -S. (2006). Science education: Integrating views of learning and
instruction. Handbook of Educational Psychology, 2, 511-544.
C
Linn, M. C. & Eylon, B.-S. (2011). Science Learning and Instruction: Taking Advantage of
Technology to Promote Knowledge Integration. New York: Routledge.
AC
Liu, O. L., Lee, H. S., Hofstetter, C., & Linn, M. C. (2008). Assessing knowledge integration in
science: Construct, measures, and evidence. Educational Assessment, 13(1), 33-55.
Liu, X. (2006). Effects of combined hands-on laboratory and computer modeling on student
learning of gas laws: A quasi-experimental study. Journal of Science Education and
Technology, 15(1), 89–100. doi:10.1007/s10956-006-0359-7
Liu, L., Lee, H.-S., & Linn, M. C. (2011). Measuring knowledge integration: Validation of four-
year assessments. Journal of Research in Science Teaching, 48(9), 1079-1107.
Lowe, R. (2004). Interrogation of a dynamic visualization during learning. Learning and
Instruction, 14(3), 257-274.
ACCEPTED MANUSCRIPT
Lui, M., & Slotta, J. D. (2013). Exploring Evolutionary Concepts with Immersive Simulations.
In Proceedings of the Tenth Computer-Supported Collaborative Learning Conference,
Madison (Vol. 1, pp. 304-311).
Lunetta, V. N., Hofstein, A., & Clough, M. P. (2007). Learning and teaching in the school
science laboratory: An analysis of research, theory, and practice. In S. K. Abell & N. H.
Lederman (Eds.), Handbook of research on science education (pp. 393-441). Mahwah,
NJ: Lawrence Erlbaum Associates.
PT
Madden, S. P., Jones, L. L. & Rahm, J. (2011). The role of multiple representations in the
understanding of ideal gas problems. Chemistry Education Research and Practice, 12(3),
283-293.
RI
Marbach‐Ad, G., Rotbain, Y., & Stavy, R. (2008). Using computer animation and illustration
activities to improve high school students' achievement in molecular genetics. Journal of
SC
McBride, D. L., Murphy, S., & Zollman, D. A. (2010). Student understanding of the correlation
between hands-on activities and computer visualizations of NMR⁄MRI. AIP Conference
Proceedings (Vol. 225, pp. 225–228).
McElhaney, K. W., & Linn, M. C. (2011). Investigations of a complex, realistic task: Intentional,
U
unsystematic, and exhaustive experimenters. Journal of Research in Science Teaching,
48(7), 745-770.
AN
McNeill, K. L., Lizotte, D. J., Krajcik, J., & Marx, R. W. (2006). Supporting students’
construction of scientific explanations by fading scaffolds in instructional materials.
Journal of the Learning Sciences, 15(2), 153–191.
M
Milgram, P., & Kishino, F. (1994). A taxonomy of mixed reality visual displays. IEICE
TRANSACTIONS on Information and Systems, 77(12), 1321-1329.
Nakhleh, M.B. (1992). Why some students don't learn chemistry: Chemical misconceptions.
Journal of Chemical Education, 69, 191-196.
D
National Research Council [NRC]. (2006). America’s lab report: Investigations in high school
science. Washington, DC: National Academy Press.
TE
NGSS Lead States. (2013). Next generation science standards: For states, by states. Washington,
DC: The National Academies Press.
Noh, T. & Scharmann, L. (1997). Instructional influence of a molecular-level pictorial
EP
presentation of matter on students’ conceptions and problem-solving ability. Journal of

Novellis, F., & Moher, T. (2011). How real is “real enough”? Designing artifacts and
procedures for embodied simulations of science practices. In Proceedings of the 10th
C
International Conference on Interaction Design and Children - IDC ’11 (pp. 90–98).
Novick, S., & Nussbaum, J. (1981). Pupils' understanding of the particulate nature of matter: A
AC
cross-age study. Science Education, 65(2), 187-196.

Olympiou, G., & Zacharia, Z. (2012) Blending physical and virtual manipulatives: An effort to
improve students’ conceptual understanding through science laboratory experimentation.
Science Education, 96(1), 21-47. doi: 10.1002/sce.20463
Osborne, R. J. & Cosgrove, M. M. (1983). Children’s conceptions of the changes of state of
water. Journal of Research in Science Teaching, 20(9), 825-838.
Ozmen, H. (2013). A cross-national review of the studies on the particulate nature of matter and
related concepts. Eurasian Journal of Physics and Chemistry Education, 5(2), 81-110.
ACCEPTED MANUSCRIPT
Posner, G. J., Strike, K. A., Hewson, P. W., & Gertzog, W. A. (1982). Accommodation of a
scientific conception: Toward a theory of conceptual change. Science Education, 66(2),
211-227.
Sanger, M., Phelps, A. & Fienhold, J. (2000). Using a computer animation to improve students’
conceptual understanding of a can-crushing demonstration. Journal of Chemical
Education, 77(11), 1517-1520.
Shen, J., & Linn, M. C. (2011). A technology‐enhanced unit of modeling static electricity:
PT
Integrating scientific explanations and everyday observations. International Journal of
Science Education, 33(12), 1597-1623.
Slotta, J. D., & Linn, M. C. (2009). WISE science: Web-based inquiry in the classroom. New
RI
York, NY: Teachers College Press.
Smith, G. W., & Puntambekar, S. (2010). Examining the combination of physical and virtual
experiments in an inquiry science classroom. In Proceedings of the Conference on
SC
Computer Based Learning in Science, Warsaw, Poland.
Snir, J., Smith, C. L. & Raz, G. (2003). Linking phenomena with competing underlying models:
A software tool for introducing students to the particulate model of matter. Science
Education, 87(6), 794␣830.
U
Steiff, M., & Wilensky, U. (2003). Connected chemistry – incorporating interactive simulations
into the chemistry classroom. Journal of Science Education and Technology, 12 (3), 285-
AN
302.
Tien, L., Teichart, M., & Rickey, D. (2007). Effectiveness of a MORE laboratory module in
prompting students to revise their molecular-level ideas about solutions. Journal of
M
Chemical Education, 84(1), 175-181.

Tobin, K. (1990). Research on science laboratory activities: In pursuit of better questions and
answers to improve learning. School Science and Mathematics, 90(5), 403-418.
Tolentino, L., Birchfield, D., Megowan-Romanowicz, M. C., Johnson-Glenberg, M. C., Kelliher,
D
A., & Martinez, C. (2009). Teaching and learning in the mixed reality science classroom.
Journal of Science Education and Technology, 18(6), 501-517. doi:10.1007/s10956-009-
TE
9166-2
Vosniadou, S. (1994). Capturing and modeling the process of conceptual change. Learning and
Instruction, 4(1), 45-69.
EP
Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin & Review, 9(4),
625-636.
Windschitl, M., & Andre, T. (1998). Using computer simulations to enhance conceptual change:
The roles of constructivist instruction and student epistemological beliefs. Journal of
C
Research in Science Teaching, 35, 2, 145-60.

White, R. T., & Gunstone, R. F. (1992). Probing understanding. London: Falmer Press.
AC
Wu, H. K., Krajcik, J. S., & Soloway, E. (2001). Promoting understanding of chemical
representations: Students' use of a visualization tool in the classroom. Journal of
Xie, C. (2012). Framing Mixed-Reality Labs. @Concord Newsletter, Spring 2012. Retrieved
from http://concord.org/publications/newsletter/2012-spring/framing-mixed-reality-labs.
Zacharia, Z. C. (2007). Comparing and combining real and virtual experimentation: an effort to
enhance students’ conceptual understanding of electric circuits. Journal of Computer
Assisted Learning, 23(2), 120–132.
ACCEPTED MANUSCRIPT
Zacharia, Z. C., & Anderson, O.R. (2003). The effects of an interactive computer-based
simulation prior to performing a laboratory inquiry-based experiment on students’
conceptual understanding of electric circuits. American Journal of Physics, 71, 618-629.
Zacharia, Z. C., & Olympiou, G. (2011). Physical versus virtual manipulative experimentation
in physics learning. Learning and Instruction, 21(3), 317-331.
Zhang, Z. H., & Linn, M. C. (2011). Can generating representations enhance learning with
dynamic visualizations? Journal of Research in Science Teaching, 48(10), 1177-1198.
PT
Zollman, D. A., Rebello, N. S., & Hogg, K. (2002). Quantum mechanics for everyone: Hands-on
activities integrated with technology. American Journal of Physics, 70, 252.
RI
U SC
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
Appendix A – Pretest/Posttest Questions and Sample Normative Responses
Questions 1-3: Air mattresses

Air mattresses are inflatable pads used for sleeping. A company tested air mattresses in a variety
of usages and environmental conditions. Please provide explanations for each of these results.
1. (Proximal, Temperature) Result 1: A manager put the mattress next to the hot fireplace,
PT
warming up the air inside the mattress. The manager noticed it was getting hot, so then she
placed it by the window where it was cooler. Explain what happened to the gas particles inside
the mattress when heated and cooled:
RI
Example normative student answer: When the mattress was next to the hot fireplace, the gas
particles inside the mattress increased in speed because molecules with more energy move faster.
SC
When the mattress was placed by the window, the particles inside the mattress slowed down
because less energy means the molecules will move slower.
2. (Proximal, Pressure-Volume) Result 2: One of the company's employees sent an air mattress
U
to his son in college. Unfortunately, the mattress burst when six students sat on the mattress to
play cards. Explain why this happened:
AN
Example normative student answer: The air mattress normally is inflated because there are air
molecules inside the mattress that are constantly moving and bouncing against the walls, putting
M
a force on the inside of the walls that pushes the mattress out. Normally there is also an external
pressure on the mattress from the outside air molecules. But if too many students sit on the
mattress, the outside pressure will be too much for the mattress to handle and the material will
break.
D
3. (Proximal, Temperature-Pressure) Result 3: A few employees took the air mattress out for
TE
camping. They inflated it on a hot summer day but then it became substantially deflated during
the night when it was a much cooler temperature. Explain why this happened:
EP
Example normative student answer: When the temperature cooled down, the speed of the air
molecules on average decreased. This leads to them bumping into the walls of the mattress less
and that decreases the pressure, which means that it will be deflated.
C
4. (Distal, Pressure) Why do inflated balloons keep their shape? Please explain at the particle
level.
AC
Example normative student answer: The gas molecules inside the balloons are constantly moving
and colliding with each other and the balloon walls. The collisions with the balloon wall exert a
force and a pressure on the balloon from the inside pushing out.
5. (Distal, Pressure-Volume) A small balloon is filled with air. When you squeeze the balloon,
you feel it’s harder and harder to squeeze. Please explain why this happens at the particle level.
Example normative student answer: Air molecules inside the balloon are moving around and
ACCEPTED MANUSCRIPT
colliding with the walls. When you squeeze the balloon, you are decreasing the volume of the
space that the molecules are in, which means that they will hit the walls on average more often
than before. If you increase the collisions, that increases the force and the pressure from the gas,
which makes it feel harder and harder to squeeze.
6. (Distal, Temperature-Pressure) Car tire pressure needs to be checked and kept within a
recommended range both in the summer and in the winter. Please explain why at the particle
PT
level.
Example student answer: In the summer, the temperature is higher than in winter and the gas
RI
molecules in the tire will be moving faster, which means they collide with the inside of the tire
more often that can result in increased pressure. In the winter, the temperature is lower and the
gas molecules will be moving more slowly, which means that they collide with the inside of the
SC
tire less often which can result in decreased pressure.
U
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
Acknowledgements
This material is based upon work supported by the National Science Foundation under
grants IIS-1123868 and IIS-1124281. Any opinions, findings, and conclusions or
recommendations expressed in this material are those of the authors and do not
necessarily reflect the views of the National Science Foundation. The authors would
especially like to thank Charles Xie and Edmund Hazzard from the Concord Consortium
for their collaboration on this project, as well as the teachers and students who
PT
participated in this project.
RI
U SC
AN
M
D
TE
C EP
AC

Chiu2015 PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Chiu2015 PDF

Hochgeladen von

Copyright:

Verfügbare Formate

Accepted Manuscript

The Effects of Augmented Virtual Science Laboratories on Middle School Students’

Jennifer L. Chiu, Crystal J. DeJaegher, Jie Chao

To appear in: Computers & Education

Received Date: 7 August 2014

The Effects of Augmented Virtual Science Laboratories on Middle School Students’

Jennifer L. Chiu, Crystal J. DeJaegher, Jie Chao

Corresponding author: Jennifer L. Chiu

The Effects of Augmented Virtual Science Laboratories on Middle School Students’

Fuhrmann, Greene, & Salehi, 2012; Olympiou & Zacharia, 2012).

combining virtual and physical components to provide enhanced educational experiences

inputs the information into the simulation.

2.1 Supporting Science Learning through Grounded Knowledge Integration

& Linn, 2011).

We also leverage embodied and grounded approaches to cognition as a framework for

2.3 Visualizations in Science Classrooms

linked representations of phenomena (Kozma, 2003) or explicit guidance through instruction

explanations of thermal concepts. Results demonstrate that connecting visualizations to

2.4 Combining Physical and Virtual labs

However, combining physical and virtual manipulatives either sequentially or side-by-

approaches to enhance traditional instruction (e.g., Johnson-Glenberg, Birchfield, Tolentino, &

of mixture of gases on pressure.

for WISE and a Frame at their lab station.

3.4 Data Sources

explanations that students made during the lab.

3.5 Data Analysis

ideas were attributed to off-task or unclear responses (e.g., “This is fun”).

Table 1. Adapted Knowledge Integration Rubric (from Liu et al., 2008)

KI Level Score Response characteristics Examples Using Student Work

4.1 Overall Usage

4.2 Pretest and Posttest Performance

Pressure-Volume 2.47 (.94) 2.91 (.82) 2.88 44 .50 < .01*

Pressure 2.07 (.94) (1.32) 4.12 44 .64 < .01*

4.3 Developing Molecular-level Explanations

Table 3. Average number of normative macroscopic, molecular, and macro-molecular

Macroscopic Molecular Molecular Total No. of Ideas

mattress at different temperatures.

compress it more and POP!”

4.4 Refining Alternative Ideas

Normative Partial Alternative Vague/Irrelevant Total No. of Ideas

ideas and refinement of partial/alternative ideas. Students tended to have explanations at

temperature-pressure activities. Although most students stated they observed increased

molecular-level explanations, remaining at the macroscopic level. One possible explanation

instead eliciting revision of their explanations on a macroscopic level.

As this study focused on the development of explanations, assessment items focused on

contributes to the field of mixed-reality technologies by providing evidence that providing

Chemical Education, 63, 64-66.

Media and Social Effects. Singapore: Springer.

Conference on Interaction Design and Children (pp. 296-299). ACM.

Instruction, 24(4), 467–563.

ideas about thermodynamics: Computer visualizations and tactile models. Journal of

Lawrence Earlbaum Associates.

undergraduate students’ understanding of electromagnetism concepts? Journal of the

Duschl, R. & Gitomer, D. (1991). Epistemological perspectives on conceptual change:

simulations. Washington, DC: National Academy Press.

understanding of electric circuits in simulation only and simulation-laboratory contexts.

in embodied mixed-reality environments: principles and results. In Proceedings of the 7th

Jones, L. L., Jordan, K. D. & Stillings, N. A. (2005). Molecular visualization in chemistry

Practice, 6(3), 136-149.

understanding of gas laws. Journal of Chemical Education, 77(2), 235-238.

Petter Johansson (Eds.) Cognition, Education and Communication Technology. Mahwah,

framework for instruction. In M. C. Linn, E. A. Davis, & P. Bell (Eds.), Internet

presentation of matter on students’ conceptions and problem-solving ability. Journal of