Conceptual change in an organic chemistry laboratory

Material Information

Conceptual change in an organic chemistry laboratory a comparison of computer simulations and traditional laboratory experiments
Gaddis, Barbara A
Publication Date:
Physical Description:
xix, 417 leaves : ; 28 cm

Thesis/Dissertation Information

Doctorate ( Doctor of philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Education and Human Development, CU Denver
Degree Disciplines:
Educational leadership and innovation


Subjects / Keywords:
Chemistry, Organic -- Experiments -- Computer simulation ( lcsh )
Laboratories ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 389-417).
General Note:
School of Education and Human Development
Statement of Responsibility:
by Barbara A. Gaddis.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
48805127 ( OCLC )
LD1190.E3 2001d .G32 ( lcc )

Full Text
Barbara A. Gaddis
B.S., University of Colorado at Colorado Springs, 1982
M.S., University of Colorado at Boulder, 1990
M.A.T, Colorado College, 1990
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Educational Leadership and Innovation

2001 by Barbara A. Gaddis
All rights reserved.

This thesis for the Doctor of Philosophy
degree by
Barbara Ann Gaddis
has been approved
David R. Anderson
Rodney Muth
R. Paul Sale
A 8 01

Gaddis, Barbara (Ph.D., Educational Leadership and Innovation)
Conceptual Change in An Organic Chemistry Laboratory: A Comparison of
Computer Simulations and Traditional Laboratory Experiments
Thesis directed by Associate Professor David R. Anderson
This quasi-experimental research study examined the effect of computer
simulations and hands-on laboratory experiments in enhancing conceptual
understanding and alleviating misconceptions of organic chemistry reaction
mechanisms. Subjects were sixty-nine sophomore-level organic chemistry students
enrolled in four laboratory sections. Laboratory sections were stratified across
instructor and randomly assigned to serve as a control or treatment laboratory.
Students in the control group performed all hands-on experiments. Students in the
treatment group performed hands-on experiments for the first and last part of the
semester but performed computer simulations for a five-week period in the middle of
the semester. Prior to treatment, groups were equivalent with respect to academic
orientation, motivation, formal reasoning ability, and spatial visualization ability.
Fifteen common misconceptions held by beginning organic chemistry students
were identified from the Covalent Bonding and Structures Test. At the end of the
semester, thirteen of these misconceptions persisted. Molecular geometry was the
only category of misconceptions that significantly improved as a result of computer
simulations, F(l,58) = 6.309, p = .015. No significant differential change was
observed in misconceptions about bond polarity, molecular polarity, intermolecular
forces, lattice structures, or the octet rule.

Computer simulations were found to result in significantly greater conceptual
understanding of organic chemistry reactions on two of the experiments,
Stereochemistry, F(l,55) = 6.174, p = .016, and Nucleophilic Substitution, F(l,57) =
6.093, p = .017. The other three experiments, Infrared Spectroscopy, Elimination, and
Oxymercuration, did not show a significant differential effect between types of
laboratory experiences. No significant differences were observed on long-term
retention of concepts.
Overall conclusions from the study are that neither computer simulations nor
hands-on laboratory experiments are effective in alleviating misconceptions, but that
computer simulations can significantly improve conceptual understanding of organic
reaction mechanisms.
This abstract accurately represents the content of the c;
its publication.
I recommend
David R. Anderson

I dedicate this thesis first to my husband Larry, whose love and encouragement were
of tremendous help throughout this exciting process. Larry, you are my role model of
an educated person. To Jeff, Aaron, Lisa, and Mackenzie, I thank you for giving me
love and support. According to Darwin, it is not the strongest of the species that
survive, not the most intelligent, but the ones most responsive to change. Thank you
for helping me to grow. To my father, Paul, I thank you for giving me enthusiasm for
learning that has motivated, guided, and sustained me through the years.

My journey was possible only because of outstanding people who generously gave
their time, expertise, and encouragement throughout these past few years. My thanks
go to my advisor, David Anderson, for his technical expertise, creativity, and his love
of teaching. I sincerely thank my dissertation committee, Rod Muth, Nadyne
Guzman, and R. Paul Sale, who gave me unlimited encouragement, guidance, and
support and Pam Shockley, who instilled the desire to begin the journey and the drive
to continue it. This project would not have been possible were it not for the
cooperation of Allen Schoffstall, Chair of the Chemistry Department, and two very
fine laboratory instructors, Sandhya Buchanan and Shannon Coleman. This project
was made all the more exciting by their genuine interest and enthusiasm.

Significance of the Problem.................
Rationale for Study.........................
Theoretical Framework.......................
Misconceptions and Alternative Conceptions....
Conceptual Change Theory................
Undergraduate Science Laboratory Curriculum
Computer Laboratory Simulations.........
Other Factors Affecting Learning........
Statement of the Problem....................
Research Questions..........................
Significance of the Study...................
2. REVIEW OF LITERATURE.........................
. xvii
.. 6
.. 7
.. 8
Misconceptions and Alternative Conceptions
Misconceptions in Chemistry.............

Misconceptions about Chemical Reactions...............27
Misconceptions about Dynamic Reactivity...............32
Misconceptions about Atomic and Molecular Structure...34
Misconceptions about Covalent Bonding and Structure...38
Misconceptions about Geometry, Bonding, and Polarity..42
Misconceptions about Organic Structure and Reactivity.45
Summary of Misconception Literature...................46
Conceptual Change.........................................47
Origins of Misconceptions.............................47
Prior Knowledge as a Cause of Misconceptions..........49
Instruction as a Cause of Misconceptions..............52
Instructor as a Cause of Misconceptions...............53
Persistence of Misconceptions.........................55
Conceptual Change Theories................................56
A Piagetian Model of Conceptual Change................56
Conceptual Change Teaching Strategies.................57
Discrepant Event Model of Conceptual Change...........58
Constructivist Computer Learning Model................60
Dynamic Model of Conceptual Change....................61

Learner Control Model
Commonalities of Conceptual Change Models..............63
Summary of Conceptual Change Teaching Strategies......67
Conceptual Change in the Chemistry Laboratory..............67
Goals of Laboratory Work...............................67
Actualized Learning Outcomes from Laboratory...........69
The Laboratory Curriculum..............................71
Analysis of Laboratory Instruction.....................75
Pedagogical Benefits of the Laboratory.................76
Computer Simulations and Conceptual Change..................77
Benefits of Computer Simulations.......................80
Limitations of Computer Simulations....................82
Learning from Computer Simulations.....................83
Studies of Computer-Based Learning and Simulations....84
Studies of Affective Factors...........................86
Summary of Meta-Analyses of Computer Simulations......88
Computer Simulations in the Laboratory......................89
Simulations Involving General Chemistry Concepts.......89
Simulations and Particulate Properties.................96
Simulations and Alternate Representations of Matter...98

Simulations and Cognitive Skills.....................99
Simulations and Laboratory Technical Skills........102
Simulations, Transfer, and Retention...............104
Summary of Computer Simulation Studies.............105
Factors Influencing Effectiveness of Computer Simulations.105
Learner Control....................................114
3. METHODOLOGY................................................118
Research Design.........................................118
Factors Affecting Performance in Organic Chemistry......121
Covalent Bonding and Structure Test................123
Group Assessment of Logical Thinking Test (GALT).....124
Purdue Visualization of Rotations Test (ROT).......125
Motivated Strategies for Learning Questionnaire
Performance Measurements...........................127

Materials: Computer Simulations and Experiments............129
Establishing Pre-Treatment Equivalence....................133
Establishing Instructor Equivalence.......................134
Experimental Conditions.................................. 137
Timetable of Events...................................137
Research Design and Data Analysis.........................143
Control for Effects and Biases........................144
Control for Alternative Hypotheses....................144
Summary of Research Methodology...........................146
4. RESULTS.......................................................147
General Approach to Statistical Analysis..................148
Establishing Pre-Treatment Equivalence....................148
Academic Orientation..................................149
Reasoning Ability and Logical Thinking................150
Spatial Visualization Ability.........................151
Assignment to Treatment Groups........................153
Research Question 1: Identification of Misconceptions.....155
Research Question 2: Treatment Effects on Misconceptions.160

Analysis of Pre-CBST Results..........................161
Identifying Persistent Misconceptions.................163
Instructor Effects....................................173
Research Question 3: Computer Simulations...................174
Strategy for Analyzing Laboratory Experiments.........175
Results of Laboratory Treatment.......................177
Analysis of Pre-Treatment Laboratory Experiments......179
Analysis of Treatment Laboratory Experiments..........181
Analysis of Post-Treatment Laboratory Experiments....195
Summary of Treatment Effects and Conceptual Change.... 197
Research Question 4: Retention..............................202
Summary of Research Study Results...........................204
5. DISCUSSION AND CONCLUSIONS...................................206
Establishing Pre-Treatment Equivalence......................206
Cognitive and Non-Cognitive Factors...................207
Adjusting for Instructor Effects......................211
Synopsis of Research Study..................................212
Research Question T. Identification of Misconceptions.......213
Misconceptions about Bond Polarity....................215
Misconceptions about Molecular Geometry

Misconceptions about Molecular Polarity............. 221
Misconceptions about Intermolecular Forces...........223
Misconceptions about the Octet Rule..................224
Misconceptions about Lattice Structure...............225
Summary of Misconceptions Identified in Study.......226
Comparison of Misconceptions with Other Studies.....228
Analysis of Persistent Misconceptions................231
Research Question 2: Effect of Treatment on Misconceptions.235
Numbers of Misconceptions by Treatment Group.........235
Changes in Misconceptions by Treatment...............236
Attributes of Computer Simulations...................237
Summary of Treatment and Conceptual Change...........239
Research Question 3 : Simulations and Conceptual
Comparison of Experiments............................240
Treatment Experiment 1: Infrared Spectroscopy........243
Treatment Experiment 2: Elimination..................247
Treatment Experiment 3: Oxymercuration...............250
Treatment Experiment 4: Stereochemistry..............253
Treatment Experiment 5: Nucleophilic Substitution...256

Computer Simulations and Instructional Methodology.258
Effect of Instructional Strategy on Learning........263
Factors Influencing Effectiveness of Simulations....264
Research Question 4: Long-Term Effects of Simulations.....270
Conceptual Change Strategy and Simulations..........271
Alternate Learning Strategies.......................272
Animation Effects...................................273
Summary of Research Findings..............................273
Limitations of the Research Study.........................276
Suggestions for Future Research...........................277
Implications of Learning in a Virtual Organic Laboratory..279
Pedagogical Benefits................................281
Practical Benefits..................................283
Non-Cognitive Benefits..............................285
Misconceptions about Covalent Bonding and Structure.287
Implementing Conceptual Change in a Virtual Laboratory....290
QUESTIONNAIRE............................................ 304

G. PRE-LAB AND POST-LAB TESTS.............336
H. RETENTION EXAM.........................369
I. DATA TABLES............................377

2.1 Dual coding model of multimedia learning.....................108
4.1 Percent correct answers on tier 1 questions (content knowledge)
and tier 2 questions (conceptual understanding)...............159
4.2 Percentage of class holding each of the fifteen identified
misconceptions at the beginning (initial) and end (persistent) of
the semester..................................................164
4.3 Initial (Pre-CBST) and persistent (Post-CBST) misconceptions by
treatment group in six conceptual categories: bond polarity (1),
geometiy (2), molecular polarity (3), intermolecular forces (4),
lattice energy (5), and octet rule (6)........................169
4.4 Post-lab test scores on Elimination by treatment and teacher.184
4.5 Post-lab test scores of Stereochemistry by treatment and teacher... 190
4.6 Post-lab test scores of Nucleophilic Substitution by treatment and
4.7 Mean post-lab test scores for Pre-Treatment experiments
(Experiments 1-4), Treatment experiments (Experiments 5-9),
and Post-Treatment experiments (Experiments 10-12)............199
4.8 Normalized Pre-Treatment, Treatment, and Post-Treatment post-lab
test scores...................................................201

3.1 Timetable for Treatment Experiments........................140
4.1 Factors Affecting Academic Performance.....................154
4.2 Mean of Conceptions and Misconceptions on Pre-CBST Test...158
4.3 Pre-CBST Scores and Misconceptions by Treatment Group......161
4.4 Initial Misconceptions by Treatment Group..................162
4.5 Post-CBST Scores by Treatment Group........................166
4.6 Persistent Misconceptions in Conceptual Categories.........168
4.7 Persistent Misconceptions with Initial Misconceptions as
4.8 Persistent Misconceptions in Conceptual Categories by
4.9 Mean Adjusted Post-Lab Test Scores.........................178
4.10 Adjusted Means of Elimination Experiment by Instructor...185
4.11 Means of Pooled Laboratory Experiments....................198
4.12 Means on Retention Exam by Treatment......................203
1.1 Academic Factors by Lab Section............................377
1.2 GALT Results by Lab Section................................378
1.3 Academic Factors by Treatment..............................379
1.4 GALT Results by Treatment..................................380

1.5 Responses on Pre-CBST Test..............................381
1.6 Responses on Post-CBST Test.............................384

Illustrating organic reactions discussed in lecture is an important goal of the
undergraduate chemistry laboratory (Hegarty-Hazel, 1990). Organic chemistry
laboratory experiments illustrate important name reactions, such as the Grignard
reaction, the SnI reaction, and the Aldol condensation. The underlying assumption of
these types of experiments is that students who synthesize a product in the laboratory
will be better able to understand this reaction in lecture.
Little evidence suggests, however, that these experiments help students to
understand better the conceptual principles behind the experiment (Gallet, 1998;
Pickering, 1988). Lazarowitz and Tamir (1993, p.120) bemoan that laboratory
activities, as they are currently being implemented, do not enhance students learning
or understanding of science.
Neither do the experiments help students to visualize three-dimensional
mechanisms and reactions occurring at the molecular level. Being able to visualize at
the molecular levelhaving molecular mental modelsis even more critical in
organic chemistry than in general chemistry. Students who are unable to visualize at
the molecular level, to relate structural representations to functional group reactivity,
and to translate mentally two-dimensional drawings into three-dimensional molecular

structures may hold misconceptions that hinder further learning of organic reaction
mechanisms (Russell, Kozma, Jones, & Wykoff, 1997).
By the time a science student graduates from college, she will have spent
more than 450 hours in lecture and 500 hours in the laboratory (Bodner, 1991). From
a pedagogical standpoint, then, it would appear that much of that time is wasted:
students leave the chemistry classroom with many of the same misconceptions with
which they entered (Birk & Kurtz, 1999; Griffiths & Preston, 1992; Williamson,
1992). The problem may lie in the laboratory curriculum. Students perform
cookbook experiments, follow step-by-step procedures, and engage in
predominantly lower-order cognitive activities. Small wonder that some laboratory
critics conclude that little thinking occurs in the laboratory (Domin, 1999a, 1999b;
Kirschner & Meester, 1988; Meester & Kirschner, 1995).
The undergraduate science laboratory also is coming increasingly under fire
for being too costly, too time consuming, too labor intensive, and too hazardous
(Bennett, 1995). These practical problems, coupled with the apparent lack of learning
outcomes, have raised questions about the wisdom of continuing to teach
undergraduate science labs (Hilosky, Sutman, & Schmuckler, 1998). In spite of these
problems, few critics advocate doing away with the laboratory, preferring instead to
change the nature of the lab, making it more conducive to student learning.

Interactive computer laboratory simulations may be one solution to the
practical and pedagogical problems associated with the traditional undergraduate
science lab. Typically, computer simulations have been acclaimed for their pragmatic
advantages, replacing laboratory experiments that are too hazardous, too costly, too
complex, or that occur too quickly or too slowly to be observed (Fletcher, Hawley, &
Piele, 1990). However, computer laboratory simulations may also have strong
instructional benefits that can help organic chemistry students improve conceptual
understanding and develop better molecular mental models.
Significance of the Problem
Organic chemistry is a highly abstract and visual discipline that requires
students to envision organic compounds simultaneously as molecules, as symbolic
two-dimensional drawings, and as three-dimensional structures. Students with
misconceptions about bonding, structure, and polarity at the molecular level may be
unable to understand mechanisms and reactions (Russell et al., 1997). The high
failure rate in chemistry classes may be at least partly attributable to students
inability to visualize at the molecular level and to form correct concepts about
microscopic chemical systems (Gabel, 1999).
Theoretically, students completing organic chemistry should have few
persistent misconceptions about structure or bonding, since these topics are discussed

in detail throughout the course. However, research shows that misconceptions do not
significantly decrease with formal instruction: Third-year chemistry students have
many of the same misconceptions as college freshmen (Birk & Kurtz, 1999), and
even chemistry graduate students hold misconceptions about pressure, structure,
density, and other concepts (Bodner, 1991).
Neither lecture nor laboratory has proved effective in replacing
misconceptions or in promoting conceptual understanding (Duchovic, 1998).
Interactive computer laboratory simulations, on the other hand, have had
demonstrable success in general chemistry, biology, anatomy and physiology, and
physics labs (Akpan & Andre, 2000; Bertrand, 2000; Escalada & Zollman, 1997;
Kinzie, Larsen, Burch, & Baker, 1996; Sanger, Phelps, & Fienhold, 2000; Vining,
2000; White, 1993; Windschitl & Andre, 1998). Although computer laboratory
simulations have not been widely implemented in the organic chemistry laboratory
curriculum, factors that make interactive computer simulations effective in other
disciplines such as animation (Rieber, 1990; Williamson, 1992), feedback (Azevedo
& Bernard, 1995), and learner control (Friend & Cole, 1990; Merrill, 1994; Milheim
& Martin, 1991) are also expected to enhance learning in organic chemistry.
This study will identify misconceptions about bonding, structure, and polarity
held by organic chemistry students at the beginning and at the conclusion of a first-
semester organic chemistry class. Interactive computer laboratory simulations will be

used as a conceptual change strategy. The simulations will provide animations of
reaction mechanisms and experiments, feedback, practice, and learner control.
Conducting parallel experiments in the traditional lab and in a virtual, computer
simulated lab, this study will compare students understanding of organic chemistry
Rationale for Study
The importance of misconceptions, visualization, and molecular mental
models in organic chemistry is arguably even greater than in general chemistry.
However, no research has yet examined misconceptions held by sophomore-level
organic chemistry students. This study, by identifying common misconceptions about
covalent bonding, molecular geometry, polarity, reactivity, and other important
chemical concepts of organic chemistry students, will fill an important void in the
The results of this research study will prove pivotal in assessing the efficacy
of computer simulations as replacements for traditional experiments. Whatever the
outcome, the results will provide direction to college science departments across the
country, as they struggle with increasing costs and decreasing resources. Computer
laboratory simulations may provide an alternative for experiments that can no longer
be conducted safely or effectively in an undergraduate chemistry lab.

The outcome of this research will have implications for distance education as
well. The population of higher education is changing, becoming older, less mobile,
less able or willing to be traditional, on-campus full-time students (Dolence & Norris,
1995). Changing demographics, funding deficiencies, and budgetary constraints are
expected to increase exponentially the numbers of distance learners (Harasim, 1990),
many of whom will need to enroll in the chemistry laboratory. The results of this
study will indicate whether computer laboratory simulations will adequately function
as virtual laboratory experiences for students who cannot be in the actual chemistry
Theoretical Framework
The theoretical framework for this study draws from the literature of
misconceptions and alternative conceptions, conceptual change theory, undergraduate
chemistry laboratory pedagogy, and computer simulations.
Misconceptions and Alternative Conceptions
Many theories of misconceptions consider the origin of misconceptions to be
the learners own interactions with the physical, social, and natural environment
(Wittrock, 1994). This constructivist rationale reflects the observations that
misconceptions are widely held across a range of ages, experiences, and social levels

(Birk & Kurtz, 1999; Driver, 1989; Osborne & Cosgrove, 1983; Wandersee, Mintzes,
& Novak, 1993). Numerous descriptive research studies have identified
misconceptions in preschool, elementary, secondary school, and undergraduate
chemistry students. Many of the misconceptions have important implications for
learning and for conceptual change teaching. Specific findings from these research
studies will be presented in Chapter 2, while the instruments used to identify
misconceptions will be discussed more thoroughly in Chapter 3.
Conceptual Change Theory
If conceptual frameworks are constructed prior to starting school, as
constructivists claim (Hawkins & Pea, 1987; Jaramillo, 1996), then formal instruction
may not be effective in changing prior concepts. In fact, students with more chemistry
experience may actually have more deeply held misconceptions than do novice
chemistry students (Griffiths & Preston, 1992). This paradox is due to several factors,
which will be explored more fully in Chapter 2.
Strike and Posner (1992) use the concept of conceptual ecology to describe
how new learning is continually being integrated into, modified by, or rejected from a
learners conceptual framework (Strike & Posner, 1992). Models of conceptual
change generally involve recognition of the misconception, dissatisfaction with the
current misconception, and practice using a replacement concept in solving problems.

The Strike and Posner model of conceptual change suggests criteria that the
replacement concept must meet in order to be effectively incorporated into the
learners conceptual ecology. This model and four other conceptual change models
are discussed in Chapter 2. Examples of conceptual change teaching strategies will
also be presented.
Undergraduate Science Laboratory Curriculum
Unfortunately, lecturethe predominant mode of instruction in college
chemistry coursesis not effective in changing students misconceptions (Dinan &
Frydrychowski, 1995). Nor has the laboratory been effective in reducing
misconceptions or in enhancing comprehension of chemical concepts (Kirschner &
Meester, 1988). In fact, some critics suggest that little learning is occurring in the
expository laboratory (Gallet, 1998; Gunstone & Champagne, 1990; Pickering, 1988).
Yet, this style of instruction predominates in the college chemistry laboratory
(Abraham et al., 1997).
Discovery-oriented labs, inquiry labs, and project-based labs have had some
success in improving student performance when implemented in organic chemistry
and general chemistry. Yet the perception of faculty that these types of lab
experiences are too time intensive for the cognitive gains obtained has limited their
adoption into the laboratory curriculum (Domin, 1999b). Goals of the undergraduate

laboratory, instructional methodologies, and relevant examples from the literature are
discussed in more detail in Chapter 2.
Computer Laboratory Simulations
On the other hand, computer laboratory simulations have been quite effective
when used as a supplement to or replacement for traditional experiments. Interactive
computer laboratory simulations can improve conceptualization, make abstract
concepts more concrete, aid retrieval and encoding tasks (Rieber & Parmley, 1995),
enhance visualization and spatial relational ability (Aldahmash, 1995), improve
student learning (Bonk & Reynolds, 1997; Bostock, 1997; Dede, 1996; Nickerson,
1995; Windschitl & Andre, 1998), and thereby contribute to conceptual change
(Driver & Scanlon, 1988). If the predominant goal of the laboratory is to illustrate
principles or reactions discussed in lecture, then computer simulations can illustrate
those principles often better than a traditional laboratory experiment.
Conversely, computer simulations may present too simplified a system, giving
a false idea of the phenomenon. Lack of interaction with lab instructor and other
students may reduce motivation to learn. These are only two of numerous objections
that are raised against using computer simulations in the lab (Jacobson, 1999; Prosser
& Tamir, 1990). On balance, however, research shows computer simulations to be an
effective way of learning. Current research on computer simulations will be

discussed in Chapter 2, with relevant examples of computer laboratory simulations
applicable to organic chemistry.
Several theories have attempted to rationalize the apparent effectiveness of
computer simulations in learning. Paivios Dual Coding Theory (1991) and the Mayer
and Sims (1994) Multimedia Theory assume interconnections between two cognitive
subsystems, one of which operates in the verbal/text-based system, the other which
pertains to images. Animations in computer simulations cause linkages to be formed
between the two subsystems, allowing more efficient retrieval of ideas from working
memory. These theories are explained in greater detail in Chapter 2, as are theories
explaining how feedback and learner control impact learning (Mayer & Sims, 1994;
Paivio, 1986; Paivio & Clark, 1991).
Other Factors Affecting Learning
The essence of the literature is that misconceptions hinder students ability to
understand organic chemistry and diminish academic performance. However, factors
other than misconceptions may also affect performance in college chemistry. From
the literature, four main factors were isolated that influence academic performance in
college-level general chemistry and organic chemistry classes. These four factors are
academic orientation (grade in general chemistry and SAT or ACT scores), formal
reasoning ability, spatial visualization ability, and motivational level.

Academic Factors. Prior chemistry knowledge is an important predictor of
academic achievement in college science classes (Yu, 1999). Prior knowledge helps
students to incorporate new concepts into existing schema, thereby improving
conceptual understanding (Alao & Guthrie, 1999). Successful performance in second
semester chemistry (General Chemistry II) is a significant predictor of academic
performance in organic chemistry, as is grade received in General Chemistry I
(Sevenair, 1987). For both community college and university students, prior
knowledge is the most important variable associated with conceptual understanding
of organic reaction mechanisms and stereochemistry (Krylova, 1997).
SAT and ACT scores are also correlated with academic performance,
although the correlation varies with student population (Burchfield, 1995; Nordstrom,
1990; Spencer, 1996; Stehn, 1994). Composite ACT scores and high-school GPA
have also been identified as strong predictors of academic achievement in general
chemistry (Hendrickson, 1997). For organic chemistry students, however, Rixse and
Pickering (1987) found that grades in general chemistry were a better predictor of
academic success than math ability (Rixse & Pickering, 1985).
For science majors, in particular, mathematical ability is strongly correlated to
performance (Bunce & Hutchinson, 1993). Students with high SAT-Math or ACT-
Math scores tend to earn higher grades in freshmen chemistry while students with
lower math scores tend to earn lower grades (Spencer, 1996). However, studies of

upper division chemistry classes have found little correlation between successful
performance and math diagnostic scores and no correlation to the number of college
math courses taken (Nicoll & Francisco, 2001). This study did find a strong
correlation to verbal ability, formal reasoning ability, and academic performance in
physical chemistry'.
Formal Reasoning Ability. Formal reasoning is one of the strongest predictors
of successful achievement in physical sciences classes, such as chemistry and physics
(Bitner, 1991; Chandran et al., 1987), accounting for approximately twenty-five
percent of the variance in conceptual understanding (Williamson, 1992). For
community college organic chemistry students, reasoning ability was the most
important predictor of conceptual understanding of organic reaction mechanisms and
stereochemistry (Krylova, 1997).
Reasoning ability can be classified into three levels, based upon the Piagetian
scheme of concrete, transitional, and formal reasoning. Concrete reasoners require
physical models and experimental evidence in order to draw conclusions, while
formal reasoners are able to generalize abstractly. Understanding abstract scientific
principles requires formal reasoning ability; however, many college students are not
at this level. Graves (1998) finds that sixty-four percent of community college
students and forty percent of university students have not acquired formal reasoning

Reasoning ability can be categorized into six types of reasoning modes:
conservational reasoning (recognizing that quantity remains constant if nothing is
added or removed); proportional reasoning (interpreting functional relationships);
controlling variables reasoning (recognizing that all variables but the one of interest
must be held constant); probabilistic reasoning (determining likelihood of events);
correlational reasoning (recognizing relationships between variables); and
combinatorial reasoning (seeing all conceivable combinations by recognizing
patterns). Conservational reasoning is the only reasoning mode associated with
concrete reasoning ability; the other five formal modes have been found to account
for much of the variance in chemistry performance (Bitner, 1991). Logical thinking
skills are important for all levels of chemistry, being the most significant predictor of
academic success in non-majors introductory level classes (Bunce & Hutchinson,
1993) and in upper-division physical chemistry (Nicoll & Francisco, 2001).
Spatial Visualization Ability. Spatial visualization is the ability to
interconvert two-dimensional and three-dimensional structures, to mentally re-orient
objects, and to see relationships between the rotated objects. This skill is especially
important in organic chemistry, where stereochemistry, reaction mechanisms, and
reactivity depend upon three-dimensional orientations. Carter, LaRussa, and Bodner
(1987) found that spatial ability had only a small but positive effect on overall
performance in general chemistry, but was highly correlated with ability to solve

complex problems, such as complex word problems or problems requiring more than
algorithmic manipulation. Solving complex problems requires the ability to process
and transform data, to generate and critically analyze multiple solutions, and to be
able to separate extraneous and irrelevant facts from important ones. Spatial
visualization ability requires these same analytic, disembedding skills.
Spatial ability plays an even more important role in organic chemistry. Pribyl
and Bodner (1985,1987) compared performance on organic exams and found that
students with low spatial ability scored significantly lower on organic chemistry
exams than students with higher spatial ability. Overall, spatial ability accounts for up
to 15% in exam scores in organic chemistry courses. Kuo (1995) found that
visualization was moderately and very significantly correlated to learning
stereochemistry, particularly when students used computer-based learning. Aldamash
(1995) came to these same conclusions, finding a high correlation between
performance and spatial ability when viewing computer animations or static visuals
about organic reaction mechanisms. Conversely, Keig and Rubba (1993) found
visualization ability to be unrelated to students ability to translate between
particulate, macroscopic, and symbolic representations of matter.
Motivation. Motivation is a complex, important, and frequently overlooked
attribute of learning in chemistry classes (Ward & Bodner, 1993). Motivation is not
just one factor, but a compilation of factors that affect learning. Extrinsic and

extrinsic rewards, importance of the class to immediate and future goals, extent to
which students can control learning outcomes, self-efficacy, all these factors affect
academic success. Motivated students use a variety of learning strategies, such as
rehearsal, elaboration, organization, critical thinking, and metacognitive self-
regulation, and make efficient use of learning resources, such as peer learning, time
management, and other help-seeking strategies (Pintrich & Johnson, 1990). Together,
values, resource use, and learning strategies are associated with conceptual change
(Pintrich & Schrauben, 1992).
Motivation helps students become cognitively engaged in academics. Steiner
and Sullivan (1984) found that the primary difference between successful and
unsuccessful organic chemistry students was a positive attitude, with all other
variables, including math SAT scores, being less important. Along with prior
knowledge and use of a variety of cognitive strategies, motivation plays a critical role
in the learning process (Pintrich & Schrauben, 1992). Motivation, described as a
process whereby goal-directed activity is instigated and sustained (Pintrich &
Schunk, 1996, p. 4), correlates to greater cognitive engagement and improved
performance. Bligh (2000) suggests that motivated students tend to work longer at
their learning, rather than working more intensely.
In addition to having sustained activity, motivated students may use different
learning strategies. Alao and Guthrie (1999) suggest that students who are motivated

use higher level strategies and so are better able to incorporate new knowledge into
their conceptual ecologies. Investigating the role that motivation plays in academic
achievement by women in science classes, Yu (1999) found that motivation and use
of learning strategies were significant predictors of course achievement, but that prior
achievement was the most important. Shih (1998) concurred, finding that motivation
and learning strategy explained 35% of academic performance. This same study
found student motivation to be heavily influenced by task orientation (what students
perceive as important) and self-efficacy (what students believe they can accomplish).
Not surprisingly, students with higher motivation showed higher academic
performance in general chemistry (House, 1994,1995; Yu, 1999). Motivation and
learning strategies are especially important in organic chemistry. Garcia and Pintrich
(1993) studied the relationships between academic performance, gender, ethnicity,
and motivation levels. For male students, the highest correlates with successful
performance in organic chemistry were prior achievement and motivation, while for
females, learning strategies were even more important than motivation.
The complex role that motivation plays in learning has not been fully
investigated with respect to performance in college chemistry courses, nor in organic
chemistry courses in particular. However, research suggests that motivation, along
with prior chemistry knowledge, formal reasoning ability, and spatial visualization
ability, all play an important role in academic achievement.

In addition to these cognitive and affective factors identified from the
literature, several situational variables, such as employment and study time, also
impact academic performance. Working full-time while attending school is strongly
negatively correlated with retention and off-campus employment, even part-time,
negatively affects academic performance. Conversely, time spent studying has a
positive correlation with academic performance (Astin, 1993).
The body of literature on misconceptions, conceptual change theory,
undergraduate laboratory pedagogy, and computer simulations suggests that organic
chemistry students might benefit from interactive computer simulations that illustrate
important reaction mechanisms through dynamic animations. The literature
demonstrates that both cognitive and non-cognitive factors affect academic
performance in organic chemistry.
Statement of the Problem
Holding misconceptions about chemistry concepts hinders further learning
(Griffiths & Preston, 1992). Because general chemistry forms the foundation for
organic chemistry, many of the misconceptions that plague students in general
chemistry will have even greater impact on them when they are organic chemistry

students. While several studies have examined misconceptions of general chemistry
students, no studies have examined the misconceptions held by organic students. Nor
have any studies examined how effective organic chemistry lecture or laboratory
instruction is in changing misconceptions. This area of research has been sadly
When students have sound understanding of general chemistry principles, they
can more easily make connections between the macroscopic reactions they observe in
the laboratory, the mechanism of the molecular reaction presented in lecture, and the
symbolic representations used to explain their observations (Hinton & Nakhleh, 1999;
Russell et al., 1997). One of the goals of the laboratory, then, might be in helping
students to make this connection.
However, little evidence suggests that these connections are being made in the
traditional chemistry laboratory. In spite of the intentions to offer a cognitively sound
laboratory experience that fosters student learning, it would appear that little learning
occurs. Rubin (1993, p. 2) states that involving students in lab experiences to
develop process and higher order cognitive skills and to provide college level students
with opportunities for investigative type experiments, though useful objectives,
appear not often to be met. If traditional lab experiences do not foster conceptual
development, perhaps computer simulations can.

One solution for improving comprehension of chemical principles while
performing safe, timely, and inexpensive reactions is through computer laboratory
simulations. Computer simulations may prove educational to the many organic
students who have problems visualizing reactivity and properties at the molecular
level and therefore have numerous misconceptions that may interfere with learning
organic chemistry.
This study will provide valuable information about misconceptions held by
sophomore-level organic chemistry students and about the impact of computer
simulations on conceptual understanding. A complete description of the research
design for the study appears in Chapter 3.
Research Questions
This study will identify misconceptions in sophomore-level organic chemistry
students and compare the effects of interactive computer laboratory and hands-on
laboratory experiments in reducing the number of misconceptions and in promoting
conceptual understanding of organic chemistry principles. This study will investigate
the following four research questions:
1. What misconceptions do sophomore-level organic chemistry students

2. What is the impact of computer simulations on students misconceptions as
compared to traditional laboratory experiments?
3. What is the impact of computer simulations on students conceptual
understanding of nucleophilic substitution, elimination, addition, stereochemistry, and
spectroscopy as compared to traditional laboratory experiments?
4. What is the impact of computer simulations on retention of conceptual
understanding of mechanisms, reactions, and laboratory techniques as compared to
traditional laboratory experiments?
These research questions will be answered following the completion of the
Significance of the Study
Learning organic chemistry involves understanding abstract concepts and
visualization of reactions and mechanisms occurring at the molecular level. Holding
misconceptions hinders learning chemistry. To date, no studies have investigated
misconceptions held by organic chemistry students. Yet, these misconceptions may
interfere with developing conceptual understanding of this complex subject.
Conceptual change involves modifying or replacing misconceptions to
accommodate new ideas. Computer simulations have been used effectively to teach
chemistry concepts, improve problem-solving skills and critical thinking, improve

visualization skills, and effect conceptual change when used in chemistry, physics,
and biology labs. Yet, little is known about how computer simulations affect learning
in an organic chemistry laboratory.
The results of this study may reveal whether interactive computer simulations
can be effective in helping students overcome misconceptions and improve
understanding of organic chemistry concepts. In so doing, this study will fill an
important void in the conceptual change literature.

This chapter contains a review of the literature relevant to this research study.
This literature review is organized as follows: (a) misconceptions and alternative
conceptions, (b) conceptual change theory, (c) undergraduate science laboratory
pedagogy, and (d) interactive computer simulations.
Misconceptions and Alternative Conceptions
A misconception is conceptual knowledge that differs from commonly
accepted scientific consensus (Sanger, 1996). For Keig (1990), misconceptions are
deeply held beliefs that fail to provide a complete and accurate description of the
scientific world. The subtle difference between these definitions lies in the concept of
accepted versus accurate. In early scientific history, stating that the earth was
round, that atoms were composed of smaller subatomic particles (such as protons and
neutrons), or that bacteria caused disease would have been considered as
misconceptions, since these views differed from those espoused by the scientific
community. Nevertheless, misconceptions are very much viewed as deviations from
correct scientific interpretations.
Wandersee, Mintzes, and Novak (1993) describe two types of misconception
research studies: nomothetic studies are those that evaluate concepts as compared to a

standard knowledge base, while idiographic studies, being more qualitative in nature,
evaluate students' understanding on its own merits, with no comparison to external
standards. Using this terminology, most of the research described in this literature
review is nomothetic.
To some authors, the word misconception has a negative connotation
(Herron, 1996). Terms such as alternative frameworks (Driver & Easley, 1978),
preconceptions (Benson, Wittrock, & Bauer, 1993), spontaneous reasoning or
spontaneous conception (Viennot, 1979), alternative conception (Wandersee et
al., 1993), or naive conception (Caramaza, McCloskey, & Green, 1981), may be
preferred for two reasons. First, the terms imply their origins, something that the word
misconception does not. Second, several of the terms convey intellectual legitimacy
and imply that the beliefs are valid, rational, and could become more scientifically
accurate (Wandersee et al., 1993).
For the purpose of this literature review, the terms misconception and
alternative conception will be used interchangeably to indicate any stated belief
whose meaning deviates from that espoused by the scientific community. However,
the actual terms used in the original articles will be maintained.
The distinction between misconceptions and other type of errors is subtle, but
important. Keig (1990) differentiates misconceptions from simple slips and from on
the spot generative errors that dont arise from deep-seated beliefs. Misconceptions

also differ from non-conceptions or partial conceptions (Haidar & Abraham, 1991),
which involve incomplete or even non-existing knowledge.
Fowler and Bon Jaoude (1987) propose six categories of errors that students
commonly make, only one of which relates to deep-seated scientifically incorrect
ideas. The six categories are misfacts, mispropositions, misbeliefs, misconceptions,
mismanipulation, and processing mistakes. Misfacts occur when students memorize
factual information incorrectly, such as stating that the speed of light is 3.0 x 106 m/s
instead of 3.0 x 10 m/s. Mispropositions involve an incorrect relationship between
two variables, such as believing that liquids with high boiling points have high vapor
pressures. Misbeliefs are cognitive misinterpretations that affect attitudes, such as
believing that snakes are slimy or that toads cause warts. Processing mistakes occur
when an illogical or inaccurate thought process steers the learner to a faulty
conclusion. The last category, called mismanipulation, involves physical mistakes,
such as using laboratory equipment incorrectly (Fowler & Bou Jaoude, 1987). These
different types of errors suggest different causes and different conceptual change
Misconceptions in Chemistry
The cognitive field of misconceptions is extensive, covering virtually all grade
and experience levels from pre-school age children (Driver, 1989b) to college

students (Crosby, 1987; Osborne & Cosgrove, 1983; Sanger, 1996) to pre-service
teachers (Lee, 1999b) to university professors (Birk & Kurtz, 1999; Lin & Cheng,
2000). Similarly, misconceptions occur in all disciplines. Because of the highly
abstract nature of science (Gabel, Samuel, & Hunn, 1987a), the bulk of the
misconception literature deals with misconceptions in physics (Escalada & Zollman,
1997; White, 1993), biology (Cho, Kahle, & Norland, 1985; Treagust, 1988a),
mathematics (Jiang & Potter, 1993), and chemistry (Crosby, 1987; Driver, 1989b;
Sanger, 1996; Wandersee, Mintzes, & Novak, 1993; Zoller, 1990).
Some misconceptions identified by these studies are relatively harmless: A
student who believes that sulfur atoms are yellow, that naphthalene atoms smell like
mothballs, or that water molecules are made up of small droplets (Andersson, 1990)
may still be successful in college chemistry. However, other misconceptions are
much more problematic. Believing that electrons are shared equally in all covalent
bonds (Wandersee et al., 1993), that polarity determines molecular geometry
(Peterson & Treagust, 1989), that heat causes molecules to expand (Williamson,
1992), that boiling involves breaking covalent bonds (Henderleiter, Smart, Anderson,
& Elian, 2001), or that matter holds atoms together in a chemical bond (Boo, 1998)
leads to erroneous views of structure, bonding, equilibrium, kinetics,
thermodynamics, reactivity, and other chemical phenomena. For organic chemistry,
misconceptions about chemical reactivity, atomic and molecular structure, bonding,

polarity, and molecular geometry are particularly harmful, since these topics form the
core of the discipline.
A large body of literature has examined misconceptions in the physical and
biological sciences, such as force and motion, light, energy, physiology, circulatory
system, protein translation, photosynthesis, electricity, and food webs (Cho et al.,
1985; Escalada & Zollman, 1997; Kaplan, 1997; Treagust, 1988a; Wandersee et al.,
1993; White, 1993). According to these studies, general chemistry students hold a
number of misconceptions about electrochemistry (Roberts, 1993; Sanger, 1996), the
mole (Duncan & Johnstone, 1979; Griffiths & Preston, 1992; Harrison & Treagust,
1996), particulate and molecular views of matter (Novick & Nussbaum, 1978; Novick
& Nussbaum, 1981), chemical equilibrium (Crosby, 1987; Hackling & Garnett, 1986;
Tyson, 1996), thermochemistry (Boo, 1998), and kinetic molecular theory reactions
(Krajcik, 1991).
A smaller body of work pertains to chemical reactions (Andersson, 1986;
Ben-Zvi, Eylon, & Silberstein, 1987; Boo, 1998; Hinton & Nakhleh, 1999), atomic
and molecular structure (Griffiths & Preston, 1992), covalent bonding (Peterson &
Treagust, 1989; Peterson, Treagust, & Garnett, 1989), and molecular geometry and
polarity (Furio & Calatayud, 1996). Misconceptions of these latter topics will be
explored in detail, since these are the most troublesome for organic chemistry.

Misconceptions about Chemical Reactions
Understanding chemical reactions on a molecular level requires chemical
knowledge, the ability to apply the law of conservation of matter to chemical
reactions, and the ability to visualize reactions occurring on an atomic or molecular
scale (Hesse & Anderson, 1992). Unfortunately, many high school and college
chemistry students appear to lack one or more of these abilities.
Problems in Interpreting Chemical Reactions. An abundance of research
reveals that students at all developmental and experiential levels hold misconceptions
about changes that occur during chemical reactions (Andersson, 1986; Ben-Zvi et al.,
1987; Chandran, Hesse, & Anderson, 1992; Hesse & Anderson, 1992; Hinton &
Nakhleh, 1999; Lee, 1999b; Robinson, 1999). Much of the problem appears to be
students inability to interpret chemical reactions.
Hesse and Anderson (1992) found that, even though many high school
students were able to balance simple reactions involving burning, rusting, and
oxidation, these students could not provide complete and scientifically accurate
explanations for their observations of these reactions. The authors noted three
problems relating to students inability to visualize chemical reactions. First, students
cannot explain chemical reactions on an atomic or molecular level and instead
consider heat, cold, and decay to be physical entities. Second, students do not apply
conservation of matter principles when determining the masses of products formed in

a chemical reaction; instead, students tend to treat chemical reactions as though they
were merely physical reactions. Third, high school students do not relate chemical
reactivity to chemical theories, but instead make analogies to common, everyday
objects (such as comparing rust to a bathroom cleaner, because rust weakens and
removes the outer layer of a nail, just as bathroom cleaner removes the bathtub scum).
The authors conclude that students possess deeply held naive conceptions that modify
their perceptions of chemical reactions. These naive conceptions are incompatible
with scientific concepts and may arise from misinterpretation of analogies (Hesse &
Anderson, 1992).
Problems in Understanding Chemical Change. A common misconception
about chemical reactions is that products are latent, somehow hidden inside the
original compound (Keig, 1990). For example, students believe that rust is hidden
inside iron and that carbon dioxide and water are released into the air when wood
bums (Tiskus, 1992). This misconception, called displacement by Andersson (1986),
reveals a low level of understanding of chemical change.
Displacement is just one method by which students consider chemical change
to occur. Andersson (1986,1990) proposes five hierarchical categories of conceptual
understanding of chemical change: disappearance, displacement, modification,
transmutation, and chemical interaction. These categories recognize that new
substances are formed, but differ in how the change occurs and the fate of the

reactants. The author describes a study involving the combustion of phosphorous
from within a sealed flask to produce a white smoke. This demonstration was used to
elicit misconceptions about chemical reactions. Few high school chemistry students
(only 26%) were able to correctly apply conservational reasoning to determine that
the mass of the flask and contents would be the same before and after reaction. Even
fewer students were able to describe correctly the changes that occurred. The author
notes that in this study and others described in the research, students did not use the
concepts of atoms or molecules in their explanations of chemical reactions, but
instead used erroneous concepts of chemical change.
Problems Interpreting Chemical Symbols in a Chemical Reaction. An
interesting disparity exists between ability to balance the symbols represented by a
chemical equation and conceptual understanding of the meaning implicit in those
symbols. Students can balance chemical equations correctly without any conceptual
understanding of what is happening at the molecular level. This result has been
documented extensively (Andersson, 1986; Ben-Zvi et al., 1987; Mas, Furio, Perez,
& Harris, 1987; Nakhleh, Lowrey, & Mitchell, 1996; Puskin, 1998). When asked to
write a balanced equation for reaction of iron with oxygen to form iron oxide, most
students could provide the correct balanced equation. However, when asked to
explain the equation they had just written and to draw a picture of the reaction on an
atomic and molecular scale, students had great difficulty. Neither their drawings nor

their verbal explanations adequately explained how this chemical change occurred.
Students did not realize that oxygen-oxygen bonds were broken during the reaction or
that iron-oxygen bonds were formed (Andersson, 1990).
Problems Applying Conservation of Matter Principles. As indicated earlier,
students have problems applying conservational reasoning to conclude that total mass
of the reactants must equal total mass of the products. Even after students observe
that iron changes its physical properties after combustion with oxygen and even
though they recognize that a new productrusthas been formed, many students
continue to believe that the mass of the product will be the same or less than the mass
of the original metal before combustion occurred. Students neglect the mass of
oxygen in the reaction, even though they can write the structure of the product
(Fe2C>3) correctly and should therefore recognize that oxygen has been added to the
iron. Even so, students are often unable to interpret the information implicit in the
formula correctly to form a logical conclusion (Andersson, 1990; BouJaoude, 1989,
1991). Andersson (1990) concludes that balancing equations and writing correct
formulae do not imply conceptual understanding, a conclusion that has been reached
by others (Gabel et al., 1987a; Gabel, Samuel, & Schrader, 1987b; White & Tisher,
Problems Writing Chemical Formulas. Many students are unable to write
correct chemical formulas or relate symbols to structures. Robinson (1999) attributes

this problem to the symbolic nature of chemistry, where a symbol represents both
microscopic and macroscopic matter simultaneously. For example, the symbol C can
represent elemental solid carbon or an atom of carbon. Similarly, the word oxygen
can imply the element oxygen, with all its associated physical and chemical
properties, it can represent a molecule, or it can represent an isolated oxygen atom
within a compound. Students have difficulty interpreting what is meant by the symbol
and by the text, and often misapply and misinterpret macroscopic behavior from a
microscopic particle.
Subscripts in symbols can cause problems for both novice and experienced
chemistry students. For example, Ben-Zvi, Eylon, and Silverstein (1986) found that
only ten percent of Israeli high school chemistry students could accurately depict the
structures representing diatomic oxygen, elemental oxygen, and oxygen atoms (O?,
02 (g), and 2 O, respectively) in chemical equations. This problem is not unique with
high school students: Even college chemistry majors and engineering students have
difficulty interpreting subscripts to determine the number of atoms within a molecule
and are often unable to differentiate between a compound containing multiple atoms
and multiple molecules of a compound (Garnett, Garnett, & Hackling, 1995).
For organic chemistry students, the problem is exacerbated: Chemical
formulas represent not only the number of specific types of atoms within a molecule,
but specific functional groups. For example, a carboxylic acid has the general formula

RCO2H. From the formula, students must recognize both structure and functionality.
Structure implies that two oxygen atoms are bonded to the carbon atom, one through
a double bond, one through a single bond. Functionality implies that these atoms
behave as a single entity, a functional group. Herron and Nurrenbem (1999) call this
process chunking. Beginning students lack this ability and often misinterpret
chemical formula, treating all atoms as separate entities
Misconceptions about Dynamic Reactivity
Not unsurprisingly, students views of chemical reactions often differ from
those of scientists. One such difference relates to the dynamic nature of chemical
reactions. Ben-Zvi, Eylon and Silverstein (1987) compared students concepts of a
chemical reaction to a scientists interpretation with respect to a number of different
factors. To the scientist, the symbolic chemical reaction implies structural, interactive,
dynamic, and quantitative aspects of chemical reactivity. These concepts seem to be
absent in students interpretations: Students see reactions as being static, not dynamic.
They do not consider reactions as dynamic processes involving bond breaking and
bond formation, but instead believe that chemicals are simply added together to form
new products, called modification by Andersson (1990). As was found in other
studies, few students can differentiate between subscripts and coefficients.

Several key findings emerged from this research. First, misconceptions affect
all proficiency and academic levels. The students in this study were academically
superior Israeli high school students who had earned high grades in chemistry.
Similar results have been obtained in other studies (Pereira & Pestana, 1991;
Wandersee et al., 1993). Second, this study is important in demonstrating the static
view that students hold of chemical reactions. Chemical reactions start at the
molecular level. If students cannot visualize molecules correctly, they will be unable
to envision the process of breaking and reforming bonds in chemical reactions. The
authors conclude that few students demonstrate conceptual understanding of chemical
reactions and suggest that difficulties in conceptualizing chemical reactivity may
originate from misunderstandings of atoms and molecules.
Garnett, Garnett, and Hackling (1995) formed the same conclusions. These
authors found that students could not interpret chemical equations, believing that
subscripts represented atomic groups. For example, students considered calcium
chloride (CaCk) to consist of the elements Ca and CI2. Students in the study also held
the misconceptions that coefficients do not represent the numbers of particles
participating in a reaction, that chemical equations do not represent reactions at a
particulate level, and that chemical reactions are static, rather than dynamic,
processes. The authors conclude that students inability to visualize molecules as
particulate and dynamic severely handicaps their conceptual understanding.

Misconceptions about Atomic and Molecular Structure
Many descriptive misconception research studies use students conceptions of
water molecules to elicit misconceptions about atoms, molecules, structure, and
bonding. Water is an example of a deceptively simple structure with familiar physical
attributes, which are explained in terms of intermolecular attractions. Consisting of
only two elements and three atoms, the shape depends upon both bonding and non-
bonding interactions. Because students are so familiar with the structure of water,
their misconceptions are more easily revealed. The studies described below have all
used the structure of water as a tool to diagnose misconceptions.
A classic study by Griffiths and Preston (1992) identifies misconceptions
about the structure of water as a function of chemistry experience and academic level.
In this study, high school students were grouped into the following three categories
based upon previous chemistry experience and academic standing: academic science
(students having at least a 75% average in three science classes), non-academic
science, and non-science. Interviews revealed fifty-two commonly held conceptions
about water, which were subsequently grouped into seven categories relating to
structure, composition, weight, size, shape, bonding and energy of molecules.
Common misconceptions were grossly overestimating the size of atoms, believing
that atoms of the same substance varied in size, and considering the shape and mass
of atoms and molecules to depend on physical state (gas, liquid, or solid). One

revealing misconception was that atoms were alive. Students apparently interpret the
word nucleus biologically, where the nucleus of a cell refers to a living organism.
Overlapping terminology as a cause for misconceptions has been noted by others
(Harrison & Treagust, 1996).
Some misconceptions are common to all students, regardless of science
background. Interestingly, the amount of science instruction does not appear to
influence the number of misconceptions. In fact, experience may actually increase
misconceptions. Griffiths and Preston (1992) describe five misconceptions that were
held more deeply by academic science students than either students with weaker
chemistry background or students with no science background. Academic science
students were more likely to believe that: (a) water molecules consist of solid spheres,
(b) pressure affects the shape of molecules, (c) heat causes molecules to expand, (d)
the size of an atom depends only on the number of protons, and (e) the size of an
atom changes when it collides with another atom. The authors attribute these
misconceptions to overlapping terminology and analogies that students subsequently
interpret too literally.
Problems with Atomic Models. In an earlier study with university-level
chemistry students, Cros and colleagues (1986) reached this same conclusion. This
study demonstrated that although students knew the constituents of the atom very
well, they had little understanding of how the protons, neutrons, and electrons

interacted. Many students postulated an atomic structure based on the Bohr planetary
model, with electrons orbiting the nucleus, much like planets orbit the sun. In a
follow-up on the same population after students had completed a year of university-
level chemistry, the authors found little change in conceptual understanding (Cros,
Chastrette, & Fayol, 1988). Students still maintained misconceptions about atomic
structure even after a year of formal chemistry instruction. The authors suggest that
the models that professors use to explain abstract concepts may leave a lasting, and
perhaps incorrect, conception of atomic structure and other chemical principles.
Problems with Physical Models. The Cros studies mentioned above examined
the relationship between mental models and misconceptions. A study by Harrison and
Treagust (1992) examined the relationship with physical models. The authors note
that students have many misconceptions about atomic and molecular structure that
may be traced to inappropriate use of physical models and analogies. Although most
of the high school students in the study realized that matter was composed of atoms,
few had accurate concepts of atomic structure. Over half of the students envisioned
atoms as being very hard balls or spheres and neglected the role of electrons entirely.
The authors suggest that the words used in chemistry, words such as shell, cloud,
and orbit, have multiple meanings across science disciplines and that this
overlapping terminology may provide the learner with an inaccurate mental model of
atomic structure. Similar problems are encountered when students describe molecular

structure. Lewis structures, space filling models, and ball and stick models are types
of structural representations. While these models attempt to provide a physical
representation of the abstract concept of bonding, they may give students a false idea
of bond angles, bond lengths, and atomic size, leading to further misconceptions.
Ironically, most students who verbalized atoms as being spheres did not
represent them this way on paper, but instead drew them as circles. Few students
recognized the dichotomy between their mental models and pictorial representations.
The authors conclude that students have only superficial understanding of atomic
structure and little conception of particulate properties, a finding that parallels the
Cros study (Harrison & Treagust, 1996).
Problems with Composition of Matter. Garnett, Garnett, and Hackling (1995)
expanded upon the study of Harrison and Treagust (1996). In addition to examining
students views of atoms and molecules, this study used interview techniques to probe
students perceptions of the composition of matter. Six common misconceptions
identified in this study are: (a) all atoms are the same size; (b) matter is continuous,
with no space between particles; (c) atoms and molecules have different sizes and
shapes depending upon the phase; (d) atoms and molecules have macroscopic
properties, such as color and odor; (e) temperature affects the shape and size of
molecules; and (f) melting and boiling involve breaking covalent bonds. In addition,
the study revealed that students have difficulty interpreting composition of matter. A

matter. A significant number of students considered water to be a homogenous
mixture of elemental hydrogen and oxygen rather than a compound composed of
these elements. Even students who correctly described the composition of water
believed that bubbles in boiling water were elemental hydrogen and oxygen (Garnett
et al., 1995). This latter misconception appears to be held by a number of students at
different experience levels (Bodner, 1991).
Misconceptions about Covalent Bonding and Structure
In the studies of misconceptions about chemical reactions, many authors have
noted that students do not envision reactions occurring through a series of bond
breaking and bond forming reactions. A study by Ben-Zvi, Eylon, and Silverstein
(1986) suggests the problem may relate to students inability to visualize molecular
bonding. In this study, students were asked to draw a pictorial representation of the
decomposition of oxygen dichloride (OCI2) into elemental chlorine and oxygen. Even
though students were reminded that chlorine and oxygen were both diatomic, many
students represented elemental oxygen using only one atom of oxygen. Even students
who portrayed oxygen as being diatomic could not accurately depict covalent
bonding; many students used elongated bond lengths between the two atoms in
diatomic oxygen. These studies reveal that students lack conceptual understanding of

the properties of atoms and molecules and misinterpret covalent bonding (Ben-Zvi et
al., 1986).
Problems with Bonding Properties. As in studies discussed previously, Pereira
and Pestana (1991) found that students believe atomic size to differ in gas, liquid, and
solid phases. Moreover, the high school students in this study thought that bond
properties, such as bond length and bond angles, also varied with physical phase.
Students depicted drawings showing the longest hydrogen-oxygen covalent bond
length in the gas phase and the shortest in the solid phase. Students also drew bond
angles as varying between 30 and 180 depending upon the physical state of matter.
These misconceptions might be understandable in a novice chemistry student.
However, students in this study were academically gifted students selected to
participate in an international Chemistry Olympiad.
Problems with Intermolecular and Intramolecular Bonding. The studies
described above reveal a confusion between intermolecular (attractions between
molecules) and intramolecular (covalent) bonding. Few students are able to
adequately represent intermolecular attractions between neighboring water molecules.
Students in the Pereira and Pestana (1991) study discussed previously drew incorrect
pictures for both covalent bonds and hydrogen bonding interactions (hydrogen bonds)
between molecules of water. Drawings showed double bonds between hydrogen
atoms and hydrogen bonds being shorter than covalent oxygen-hydrogen bonds.

As stated earlier, students believe that covalent bond length varies with phase.
Conversely, many students erroneously believe that intermolecular attractions, such
as hydrogen bonding, do not vary with phase. Students represented all hydrogen
bonds as being identical in all phases, steam, liquid, and ice.
Even though this study was conducted with gifted students, the authors see
parallels with all levels of students. They conclude that all students have
misconceptions, that students have difficulty differentiating between intermolecular
and intramolecular bonding, and that students do not understand three-dimensional
structure. The authors suggest that better visual representation be used to help all
students comprehend structure and bonding (Pereira & Pestana, 1991).
Problems Visualizing Covalent Bonds. Boo (1998) corroborated these
findings, highlighting the confusion between intramolecular bonding and
intermolecular attractions and noting that students believe that covalent bonds are
broken when water changes from a liquid to a gaseous phase, a misconception noted
by other researchers (Henderleiter, Smart, Anderson, & Elian, 2001). However, an
even more important misconception identified from the Boo study concerns mental
models of bonding. Students mistakenly believe that matter exists between atoms in a
chemical bond, that this matter is a glue-like substance that holds the atoms together
in the bond, and that a chemical bond is a physical entity. Because of this
misconception, students also misinterpret thermochemical concepts. Using a concrete

model of chemical bonding, students rationalize that both bond-making and bond-
breaking require energy, just as constructing or destroying a building require energy.
Because of these misconceptions, students are unable to interpret thermochemical
processes, resulting in many misconceptions about thermodynamics.
Problems Interpreting the Octet Rule. Misconceptions about structure and
bonding may arise from students misinterpretation of the octet rule, an algorithm
used to rationalize covalent bonding and ion formation. Through interviews with
fifteen university students, Taber (1998) finds that students apply the octet rule
superficially, without considering the stability of the resulting structure. Using the
octet rule, students reason erroneously that Na7' (a sodium ion having gained eight
electrons) will be more stable than a neutral sodium atom. These alternative
frameworks influence how students learn chemistry. University students in this study
held dualistic views of bonding, with all bonds being either covalent or ionic. This
dualistic view may arise from anthropomorphic language used about atoms in a
molecule. When explaining their theories of bonding, students talk about sharing
between atoms, one atoms greed for another atoms electrons, or electrons
belonging to specific atoms in bonds. The author suggests that recognizing
students misconceptions can help teachers devise conceptual change strategies to
address and correct students alternative frameworks.

Misconceptions about Geometry, Bonding, and Polarity
Two companion studies from Curtin University of Technology examine the
relationships between students misconceptions of covalent bonding and other
molecular properties. Peterson and Treagust (1989) developed a two-tier test to
diagnose senior chemistry students conception of covalent structure and molecular
shape. The instrument was used to identify thirteen misconceptions held by eleventh
and twelfth year chemistry students (Peterson et al., 1989). Although misconceptions
were conspicuously reduced in the population of twelfth-year students,
misconceptions still persisted. Even after a year of studying chemistry, students still
held eight misconceptions about bonding, shared electrons, polarity, intermolecular
forces, the octet rule, and shape (Peterson & Treagust, 1989).
These studies suggest that misconceptions occur due to overlapping
terminology of scientific terms with everyday terms. For example, the origin of the
prevalent misconception that electrons are shared equally in covalent bonds may
derive from the dictionary meaning, where the word share implies an equal
division. In scientific terminology, electrons are common between two atoms, but the
electron density is not equally distributed. This problem with the word share was
also noted by Boo (1998).
Problems with Determining Polarity and Geometry. Geometry and polarity
appear to be difficult concepts for students, regardless of experience level. Furio and

Calatayud (1996) examined misconceptions of geometry and polarity occurring in
high school, college freshman, and upper division chemistry students. In general,
students were able to correctly assign geometries based upon three-dimensional
drawings, but had problems determining the geometry from the Lewis structures.
Similarly, students were frequently unable to predict polarity.
One of the most common mistakes that students made was determining
geometry and polarity of a molecule without considering lone electron pairs. This
reductionist tendency has been noted by other authors (Haidar, 1988; Haidar &
Abraham, 1991). Furio and Calatayud (1996) discuss two types of reductionism:
Geometric functional reduction refers to the tendency to consider only bonding
electrons when determining molecular geometry, while bonding functional reduction
relates to the tendency to consider only bond polarities when determining polarity.
Both forms of reductionism cause students to misinterpret Lewis structures and to
predict incorrect geometries and polarities.
Although the number of misconceptions generally decreases between high
school and first year college chemistry, misconceptions do persist. The authors urge
teachers to adopt a more constructivist approach to teaching and to use counter-
examples in the classroom to help students recognize and correct reductionist habits.
Problems Relating Electronegativities and Polarity. Birk and Kurtz (1999)
studied misconceptions relating to bond polarity, molecular shape, polarity of

molecules, crystal lattices, intermolecular forces, and the octet rule across four
academic levelshigh school, undergraduates, graduate students, and faculty.
Although experience did reduce the number of misconceptions, a number of
misconceptions were widely held across academic levels. Among the most pervading
misconceptions were that electrons are shared equally in a covalent bond, that non-
polar molecules must contain atoms of similar electronegativity, and that the presence
or absence of polar bonds determines whether molecule is polar or non-polar. The
authors speculate that many of these misconceptions should disappear after students
take organic chemistry, a discipline that emphasizes structure and bonding.
However, further chemistry experience has not been shown to eliminate
misconceptions. Birk and Kurtz (1999) demonstrated that many of the same
misconceptions persist in upper division chemistry students, even after they complete
organic chemistry. The authors find that misconceptions did decline sharply in
graduate students and faculty, leading the authors to speculate that having to teach
scientific concepts to others helps the learner to analyze and amend his or her own
misconceptions. Few studies have specifically analyzed misconceptions about
covalent bonding and structure held by sophomore-level organic chemistry students.
However, several researchers have noted the presence of misconceptions in this
population of students (Gonzales, 1998; Zoller, 1990).

Misconceptions about Organic Structure and Reactivity
Two additional studies relate specifically to organic chemistry, with one study
examining misconceptions of nursing students taking a one-semester general
chemistry, biochemistry, and organic chemistry course, and the other using a
sophomore-level organic chemistry course. Zoller (1990) observed many of the same
misconceptions of general and organic chemistry principles that have been noted in
other studies of college freshmen and new misconceptions relating to organic
reactivity. This qualitative study identified misconceptions from questions that
students ask in class. Common misconceptions of these first-year students concerned
the relative strength and reactivity of single, double and triple bonds, Lewis acidity,
nucleophile strength, electrophile strength, leaving group ability, optical activity, and
directive effects of electrophilic aromatic substitution reactions. The author suggests
conceptual change strategies that might be useful in inducing more scientifically
correct concepts for each of the misconceptions.
Gonzales (1998) examined organic chemistry students conceptions of terms
relating to nuclear magnetic resonance. From open-ended response questions, she
identified misconceptions dealing with aliphatic, aromatic, unsaturation, resonance,
symmetry, chemical shift, and several other terms. Categorizing responses as
understanding, partial understanding, no understanding, or misconceptions, she found
that few students had full understanding of these terms. Many students exhibited only

partial understanding or misconceptions. There appears to be little relationship
between the nature of the naive concepts and students problem-solving strategies.
Regardless of their level of conceptual understanding, many students used an
algorithmic process to solve structures. The author recommends that three-
dimensional molecular structure be emphasized earlier in the organic course.
Summary of Misconception Literature
The seven statements that follow summarize findings from the studies
discussed above and from numerous other descriptive studies of misconceptions:
1. Chemistry is a complex, abstract subject that lends itself to many
misconceptions (Tiskus, 1992).
2. Misconceptions are deeply held (Griffiths & Preston, 1992) and persist,
even after instruction (Gabel, 1999; Williamson, 1992).
3. Misconceptions are ubiquitous, occurring in all age and educational levels
(Birk & Kurtz, 1999).
4. Students inability to visualize at the molecular level leads to common
misconceptions about structure, bonding, and other chemical principles (Gabel, 1999;
Williamson, 1992).
5. Misconceptions are often internally consistent and coherent (Driver,

6. Misconceptions arise from a variety of sources (Hewson & Hewson, 1983).
7. Misconceptions are resistant to change (Posner, Strike, Hewson, & Gertzog,
The application of these last three findings to conceptual change theory will
be explored in this next section.
Conceptual Change
A students extant conceptual framework dictates future learning (Bodner,
1986). Recognition of the origin of misconceptions may suggest strategies for
modifying and amending misconceptions. Misconceptions arise from various factors,
including the students present conceptions, the nature of science teaching, and the
instructor. This next section discusses possible origins of misconceptions.
Origins of Misconceptions
One cause of misconceptions has been discussed previously: Overlapping
terminology and analogies may cause students to develop incorrect mental pictures of
atomic and molecular structure. However, this is not the predominant reason that
many science students have misconceptions. Constructivist learning theory suggests
that learning begins long before students ever step into a classroom and that students
learn science from interactions with the physical world. One of the problems in

science teaching may be that teachers use behaviorist teaching strategies to correct
misconceptions that originated from constructivist learning (Bodner, 1986; Herron &
Nurrenbem, 1999).
Herron and Nurrenbem (1999) use the metaphors of the microscope and
telescope to describe the difference between behaviorist and constructivist theories of
Behaviorist research attempted to narrow things down. It put learning under
the microscope in order to identify salient variables that could guarantee
improvement in performance. Constructivist-based research reverses that
focus, using a telescope to broaden the view of learning (p. 1354).
Chemistry faculty often teach from a behaviorist view, considering knowledge
to be transmitted from the expert teacher to a passive learner, who comes to class
without any preformed concepts. From a behaviorist perspective, learning occurs as a
consequence of repetitive programmed activities, designed to help students remember
and regurgitate information bestowed upon them by the teacher (Brooks & Brooks,
1993). From this perspective, it makes little sense that students should retain their
misconceptions after formal instruction.
From a constructivist standpoint, however, misconceptions are the natural
consequence of an individuals interaction with the physical, social, and cultural
environment (Brooks & Brooks, 1993). If knowledge is internal to the learner in a
Piagetian sense, then knowledge can only be influencednot determinedby

lectures, textbooks, computers, and other instructional media. The actual outcome of
instruction will be determined by the dynamic interaction of these media with the
individuals experiences within a social and cultural environment (Jaramillo, 1996).
Prior Knowledge as a Cause of Misconceptions
One of the leading causes of misconceptions, then, is the knowledge that
students bring with them to the classroom (Hawkins & Pea, 1987). Ausbel, as cited in
Bodner (1986, p. 873), recognized the importance of prior conceptions to the learning
process, stating If I had to reduce all of educational psychology to just one principle,
I would say this: the most important single factor influencing learning is what the
learner already knows.
Spontaneous Concepts. According to constructivist theory, students develop
spontaneous theories about science and the natural world even before starting school.
Spontaneous concepts arise out of childrens interactions with a concrete, physical
world. Conversely, scientific concepts evolve from formal, verbalized abstractions,
which are later applied to physical phenomena (Gonzales, 1998; Jaramillo, 1996).
Misconceptions occur when students attempt to reconcile their spontaneous concepts
with scientific concepts taught in school, when students try to replace incorrect
conceptual frameworks or to supplement incomplete mental models with
scientifically accurate concepts, and when students attempt to integrate two

incompatible concepts (Harrison & Treagust, 1996). Hewson and Hewson (1983) use
the term conceptual capture to refer to this process of integrating reasonable, but
contradictory concepts.
Conceptual Ecology. Conceptual ecology refers to the balance between a
learners conceptual framework arising from the physical and social environment and
that arising from formal instruction (Posner et al., 1982). Using a constructivist
emphasis, the authors use this phrase to emphasize the importance of a learners prior
knowledge in concept formation (Strike & Posner, 1985). Just as ecology involves an
interdependence of all constituent ecosystems, conceptual ecology describes a system
of concepts that is simultaneously both interconnected and interdependent (Krajcik,
A learners conceptual ecology is shaped by past experiences, prior
knowledge, epistemological views of science, spontaneous conceptions formed in
childhood, and analogies and metaphors internalized during the educational process
(Posner et al., 1982; Strike & Posner, 1985; 1992). Conceptual ecology, then,
describes a cognitive environment in which new information is continually being
integrated into an existing framework of conceptual connections. Since each learner
has a unique conceptual ecology, each learner assimilates information differently. The
individual learners prior knowledge thus acts as a template to incorporate new

concepts or as an inhibitor to prevent internalization of the new information (Keig,
1990; Tiskus, 1992).
This complex relationship between extant spontaneous concepts and
formalized scientific concepts explains why correcting misconceptions is so difficult
and why integration of a new concept is both learner dependent and erratic (Posner et
al., 1982). It also explains why students have an abundance of conceptual information
about many topics incorporated into their conceptual ecology before they ever step
into the classroom. Unfortunately, much of that information may be incorrect.
Unexpected Interpretations. Some misconceptions arise from
misinterpretation of observations of the natural world. For example, children
frequently believe that bubbles that appear in boiling water are heat. Other
misconceptions arise from what Gunstone and Champagne (1990) term the
unexpected interpretation of school learning. For example, learning that water is
composed of the elements oxygen and hydrogen and realizing that water can be
electrolyzed to produce those same elements, some students conclude that the bubbles
in boiling water are oxygen and hydrogen (Garnett et al., 1995). This misconception
persists well into college (Henderleiter, Smart, Anderson, & Elian, 2001) and
graduate school (Bodner, 1991). Even after exposure to correct scientific
explanations, students often maintain their misconceptions, choosing to either ignore
or modify this new scientifically correct information. Students often hold competing

incompatible conceptions, using a formal concept for answering test questions and
using their internalized informal conceptual framework to explain real life
phenomena (Gunstone & Champagne, 1990).
Instruction as a Cause of Misconceptions
Prior knowledge, naive concepts, and spontaneous concept formation provide
one explanation for the occurrence of misconceptions about the natural world. A
second explanation relates to the nature of instruction itself. Chemistry is a difficult,
abstract subject, made more so by the teaching methodologies commonly used.
Tiskus (1992) suggests four reasons why formal science instruction itself may
not eliminate misconceptions: (a) the subject matter is too compartmentalized, so that
students fail to make connections between related concepts in other topic areas; (b)
students lack prerequisite knowledge, so cannot understand basic scientific concepts;
(c) students misinterpret the complex, formal, and abstract language of science; and
(d) students apply algorithms rather than conceptual understanding when problem
solving. Similar rationale has been proposed to explain why laboratory work is
similarly unsuccessful in changing misconceptions. Students perceive lessons as
being isolated and disconnected from lecture concepts. They also do not understand
the purpose of doing lab work and therefore often misunderstand directions. In

addition, they often lack the requisite knowledge to understand the concepts behind
the experiments (Hegarty-Hazel, 1990a).
Instructor as a Cause of Misconceptions
A third cause of misconceptions is instructor-related (Renstron, Andersson, &
Marton, 1990). As mentioned previously, analogies and metaphors that teachers use
to describe abstract phenomenon can lead to incorrect scientific conceptions
(Harrison & Treagust, 1996). Phrases such as electron cloud, electron shell, and carry
charge (Garnett et al., 1995) may lead to incorrect mental models about the atom.
Misconceptions arise as students try to interpret metaphors and analogies too literally
(Haidar & Abraham, 1991). For example, students with more chemistry instruction
are more likely to believe that atoms are very hard spheres and that electrons
circulated around the nucleus in orbits than students with less formal instruction.
These errors arise from too literal interpretation of analogies used to describe atomic
structure (Griffiths & Preston, 1992). Although analogies can make abstract concepts
more concrete, Garnett, Garnett, and Hackling (1995) advise caution in using
analogies, viewing them as unintended sources of misconceptions.
Misconceptions may arise from students own attempts to simplify complex,
abstract concepts by under-differentiation, over-differentiation, and over-

generalization so as to make concepts become meaningless outside of a meaningful
context (Ben-Zvi et al., 1986; Keig, 1990).
Overlapping Terminology. In addition to unfortunate choices of analogies
and metaphors, terminology used the instructor and the textbook may also cause
students to develop misconceptions. Logan and Logan (1993) argue that semantic
differences may contribute to misconceptions. Many terms in chemistry have roots in
other disciplines: Words such as resonance hybrid, nucleus, and shells are
context dependent, with quite different meanings depending upon science discipline
(Keig, 1990). Many of these terms also have different commonplace meanings: For
example, the word particle in chemistry can designate an atom, a molecule, or an
ion, all invisible entities. In contrast, the common meaning of this word refers to a
visible object (Garnett et al., 1995). Confusion between the different semantic
meanings may cause a student to retrieve the incorrect meaning, thus resulting in a
misconception (Keig, 1990).
Instructor Misconceptions. The instructor may also contribute to
misconceptions in a more overt way: Research reveals that misconceptions are held
at all academic levels, including graduate students, pre-service teachers, secondary
school science teachers, community college instructors, and university professors
(Gabel, 1999; Lee, 1999b; Lin & Cheng, 2000; Tobin, Tippins, & Gallard, 1993;
Wittrock, 1994). Science teachers themselves may still exhibit some misconceptions

about the very subjects they are teaching (Lee, 1999; Lin, 2000; Perez & Carrascosa,
1987) and may be passing down their misconceptions (Griffiths & Preston, 1992).
Persistence of Misconceptions
Regardless of their origin, misconceptions are persistent. Science experience
itself does not eliminate misconceptions, although it generally reduces the numbers of
misconceptions and the extent to which they are held (Birk & Kurtz, 1999). However,
students with more science experience and better grades may actually hold more
misconceptions about specific topics than do less academically prepared students with
less science background (Griffiths & Preston, 1992; Williamson, 1992).
Lecture has not been a particularly effective method of inducing conceptual
change (Bodner, 1986; Diab, 1990; Grynkewich, 1994; Lazarowitz & Tamir, 1993;
Nakhleh et al., 1996). Paradoxically, the students in most need of constructivist
instruction may be the ones who get the least. The largest class sizes are in the
freshmen introductory level courses, where lecture is the most common instructional
mode (Johnstone & Su, 1994). Cohen (1994) sardonically summarizes current
pedagogy in undergraduate science education: To teach is to narrate, knowledge is
facts, to learn is to memorize (p. 27). Gallet (1998) reproaches current teaching
activities with fostering memorization and superficial understanding, rather than
conceptual change.

Fortunately, if constructivist theory explains why students have
misconceptions, it also suggests methods of inducing conceptual change. The next
section will discuss conceptual change theories and their applications to conceptual
change in science.
Conceptual Change Theories
Strike and Posner (1992) define conceptual change as an alteration of
conceptions that are in some way central and organizing in thought and learning, a
phenomenon that is analogous to Kuhns notion of a paradigm shift or Piagetian
accommodation (p. 148). The authors equate the magnitude of conceptual change
with the change between the quantum mechanical and Newtonian views of the
universe. In accordance with these perceptions of conceptual change, Zietsman and
Hewson (1986) suggest that conceptual change teaching should considered learning
as a change in students conceptions rather than simply adding new knowledge to
what is already there (p. 28).
A Piagetian Model of Conceptual Change
Conceptual change is arguably quite difficult, as the number of
misconceptions that persist through years of formal instruction can attest. Zietsman
and Hewson (1986) propose a two-step conceptual change model of learning, which

involves diagnosis of specific misconceptions, followed by remediation and practice
to show why the existing conception is unsatisfactory and why replacement with a
scientifically correct conception is beneficial. This sequence is in accordance with a
Piagetian model of conceptual change. According to this model, learning occurs as
learners interact and adapt to their environments through a process of disequilibrium.
When learners encounter a new experience that contradicts their current conception,
they experience mental conflict. Equilibrium is restored when the learner can
assimilate the discordant idea into existing conceptual frameworks or when the
learner is able to modify her extant conceptual schema to accommodate the discrepant
idea (Bodner, 1986). This Piagetian model suggests several strategies for conceptual
change teaching, which are described in the next section.
Conceptual Change Teaching Strategies
Tiskus (1992) describes conceptual change teaching as a rubric attached to
methods of instruction meant to uncover, elucidate, confront, and abandon naive
conceptions followed by the adoption of the scientific conceptions (p. 48).
Ironically, lecturethe predominant method of science teachingis ill designed to
bring about cognitive change (Wyckoff, 2001). Hewson and Hewson (1983) support
this allegation, claiming that conceptual change strategies are more effective than
traditional instruction in replacing misconceptions with scientifically acceptable

conceptions. In addition to the Piagetian model discussed above, four other models of
conceptual change are presented in this literature review.
Discrepant Event Model of Conceptual Change
For Posner, Strike, Hewson, and Gertzog (1982), dissatisfaction with the
existing conceptual framework is a critical first step in evoking conceptual change.
According to this model, dissatisfaction occurs when a learner encounters a problem
that he is unable to solve within the existing cognitive framework. (In Piagetian
terms, the problem is called a discrepant event.) The greater the cognitive dissonance
experienced, the easier a replacement concept can be accommodated.
The authors describe three criteria that must be met if replacement is to occur:
the replacement concept must be intelligible, plausible, and productive (Posner et al.,
1982; Strike & Posner, 1992). In other words, concepts must be easily
understandable, must appear to be logical, and must be capable of solving similar
problems. These criteria imply that the new concept must be consistent with the
learners own epistemological beliefs, consistent with other theories or knowledge in
the learners conceptual ecology, must be consistent with past experiences, and must
provide a rational view of the physical world, compatible with the learners
observations and experiences (Basili, 1988). The new concept must also be able to
solve a wide repertoire of new problems that might arise.

While these criteria are necessary for conceptual change to occur, they do not
imply that change will be inevitable or timely. Even if the new concept meets all of
the above criteria, the concept may not be incorporated into the extant conceptual
ecology. Learners may reject scientific conceptions for numerous reasons. They may
ignore the discrepant event, compartmentalize their knowledge so as to avoid internal
conflict, or may purposely misinterpret data so as to reconcile the new concept more
easily into their existing conceptual frameworks. Finally, accommodation may be
slow and incomplete, with students gradually accepting some components of the
concept, while rejecting others.
When Strike and Posner (1992) later revisited their initial model, they
acknowledged three limitations. First, the old model represented misconceptions as
though students could clearly identify and articulate their misconceptions, when often
they cannot. Second, the old model did not adequately convey the interconnectedness
and interactivity of conceptions, misconceptions, and conceptual ecology, although
this interconnectedness is the very crux of conceptual change. Third, the model
tended to be overly rational. The authors recommended several modifications to the
original model. The new model stresses the dynamic, complex nature of a learners
conceptual ecology, the influence of affective factors, such as motivation, goals, and
social surroundings, and the mutual influence of conceptual factors on an individuals
conceptual ecology (Strike & Posner, 1992). The authors recommend that teachers

create cognitive conflict to help students diagnose misconceptions and then provide
multiple learning strategies to help students interrelate new and extant concepts.
Constructivist Computer Learning Model
The Driver and Oldham (1986) constructivist model expands upon the earlier
Posner model and delineates a five-step approach to elicit conceptual change. The
five-steps are orientation, elicitation, restructuring, application, and review. In
orientation, instructional background provides an overview, establishes the goals of
instruction, and motivates students for future learning. In the elicitation step, the
instructor uses conversations, interviews, and other activities to extract students
existing conceptual ideas. Cognitive conflict is created in the third step (called
restructuring), with the instructor introducing discrepant events to cause students to
question their existing framework. Examples of discrepant events are thought-
provoking questions, group discussions, demonstrations, or tests. During the
restructuring phase, after a struggle to reconcile the discrepant event, the student
modifies or replaces existing conceptions. In the application step, students practice
with the new concept to solve additional problems, including problems which are
both familiar and novel. Finally, during the review step, students reflect upon what
they have learned (Driver & Oldham, 1986). Driver and Scanlon (1988) later adapted
this theory into a computer model for conceptual change using the constructs of

bridging analogies, dynamic pictorial representations, immediate feedback, and direct
Dynamic Model of Conceptual Change
Kracjik (1991) envisions conceptual change as being a dynamic, recursive
process. The Kracjik model compresses the five steps of the Driver and Oldham
model into four steps: recognition, reconstruction, application, and integration. In the
Kracjik model, students first become aware of their current level of understanding.
They then restructure their understanding, apply their new understanding to problem
solve, and then compare their new understanding with their own conceptions.
Exposing events (called discrepant events in the Posner and Driver models) and
concept maps are two methods that can be used to help students recognize gaps
between their concepts and scientifically accepted explanations. Exposing events help
students recognize the flaws in their conceptual framework, while concept maps
demonstrate missing and incorrect connections.
The main difference between the Kracjik model and other models of
conceptual change is that the Krajcik model emphasizes the recursive, dynamic, but
difficult and slow nature of conceptual change: Learners must constantly reexamine
their understanding, compare it to a correct scientific concept, then reapply the
modified framework to see how well it resolves the conflict of the exposing event.

This recursive, non-linear nature of the model is demonstrated by feedback loops,
which allow for multiple repeating cycles of the four steps. The Kracjik model also
provides the conceptual framework for computer-based instruction (Krajcik, 1991).
Learner Control Model
Based upon the Strike and Posner (1992) model, White and Gunstone (1992)
delineate three conditions that must be met before conceptual change can occur: (a)
the learner must recognize the nature of the misconception, b) the learner must decide
whether or not to reevaluate the current conception with respect to the new or
conflicting information, and (c) the learner must decide whether to reconstruct the
existing conceptual framework.
The two main differences between this model and the other preceding models
are the strong emphasis on learner control and the significance of prior conceptions to
the conceptual change process. At each step of the White and Gunstone model, the
learner is fully aware of his misconceptions and decides whether to incorporate the
new concept or to preserve the existing conceptual framework. Each step assumes
learner control over the conceptual change process and each step relates new
conceptions back to the existing framework (White & Gunstone, 1992).

Commonalities of Conceptual Change Models
The conceptual change models discussed have three traits in common. First,
each theory recognizes that the first step in conceptual change is identifying
misconceptions. Second, in each theory, conceptual change is catalyzed by
dissatisfaction with an existing conceptual framework. Third, all involve practice or
reflection using the newly formed concept.
Identification of Misconceptions. Most instructors are not aware of the
misconceptions that students bring to the classroom (Gabel, 1999). Diagnosis, if done
at all, may not occur until after the learner has already been unsuccessful on one or
more tests. Individual or group student interviews may be effective in diagnosing
misconceptions, but they are time-consuming and difficult and their validity often
depends upon interviewer interpretation. The development of two-tier diagnostic
tests, such as the Covalent Bonding and Structure Test, has greatly simplified the
process of identifying students misconceptions (Treagust, 1988a, 1988b). First tier
questions examine prior knowledge, while second tier questions examine conceptual
understanding (Birk & Kurtz, 1999). Scoring depends upon providing both the correct
answer and the scientifically correct explanation. This type of multiple-choice test has
been used successfully to identify misconceptions (Haidar, 1988; Haidar & Abraham,
1991; Schmidt, 1996; Zoller, 1996)

Discrepant Events. All the conceptual change models discussed involve an
exposing or discrepant event, an unexpected observation that the student is unable to
explain using his existing mental models. An example of a use of discrepant event
discussed earlier is the combustion of steel wool. This study has been used and cited
by a number of researchers to illustrate the common misconception that burning
transmutes chemicals to other elements and reduces mass (BouJaoude, 1991).
Students persist in thinking that the steel wool has been transformed into carbon and
that the mass of the residue will be less than the original mass. When a combustion
apparatus was connected to a triple beam balance, students altered their conceptual
frameworks to account for their observation that mass increased.
Other types of discrepant events may be peer-related. Novick and Nussbaum
(1981) used exposure of student misconceptions and student-generated drawings as
discrepant events. Students were interviewed about their perceptions of molecules,
then were asked to draw pictures representing gas phase behavior. Students used their
drawings to describe, explain, and justify their conceptual understanding to
classmates. During the application phase of the conceptual change strategy, students
applied their models to a new situation, which was usually a discrepant event
designed to point out the inadequacies of the students models. Based on the problems
they encountered in trying to reconcile their models, students refined or revamped
their models accordingly (Novick & Nussbaum, 1978). This approach, while time-

consuming, was moderately successful in modifying existing students
The Role of Practice. Practice applying the new conceptual framework to
new situations is a critical part of conceptual change. Lecture, a largely passive
process (Spencer, 1999), does not give students opportunities to practice. Concept
mapping, molecular level drawings, and computer simulations are three conceptual
teaching strategies that allow students to practice and reflect. Concept mapping and
molecular drawings are discussed briefly here, while conceptual change through
simulations will be discussed more fully in a later section.
Concept Maps. Concept maps, which represent links between related ideas
and concepts, have been reviewed extensively in the literature (White & Gunstone,
1992). They have been used to (a) explore understanding of a limited aspect of a
topic, (b) determine if students have conceptual understanding of a specific lesson, (c)
determine whether students can see relationships between distinct topics, (d) examine
whether students can differentiate between important and less important topics, (e)
identify perceived changes in relationships between topics, and (f) promote
discussion (White & Gunstone, 1992, p. 30-36).
Concept mapping has been widely implemented in secondary school settings,
but has had limited acceptance in general chemistry and organic chemistry lecture and
laboratory classes. When used in general chemistry (Nicoll, Francisco, & Nakhleh,

2001), the results have generally been positive. Similar results have been obtained
when used in organic chemistry. Nash, Liotta, and Bravaco (2000) found that students
formed more expert-like associations between concepts after using concept mapping
in an organic laboratory class. In Markows (1995) study, organic chemistry lab
students doing concept maps understood lab experiments better than students just
reading the textbook. Concept maps may also help students with lower verbal ability
better than those with greater verbal acuity (Kracjik, 1991).
Molecular Drawings. Redescription using molecular drawings is another
conceptual change strategy that has been used successfully. This strategy involves
having students describe chemical and physical phenomenon on a molecular basis,
before attempting to problem solve on a macromolecular scale. Students using this
approach significantly outperformed a control group of students on the Particulate
Nature of Matter Test, but their performances on the American Chemical Society
national high school exam did not differ significantly (Gabel et al., 1987b). The
authors conclude that students can improve their particulate views and problem-
solving abilities by describing reactions in terms of molecules before starting to solve
the problem.

Summary of Conceptual Change Teaching Strategies
Each of these conceptual change strategies provides opportunities for students
to practice using newly incorporated conceptions and each has met with some
success. However, each has been only minimally (at best) incorporated into the
college chemistry curriculum. Nor, with large class sizes and an emphasis on content
coverage, are they likely to be. The instructional strategy least likely to bring about
conceptual change is lecture (Gallet, 1998), yet this is the instructional pedagogy
most widely used in freshmen and sophomore level science courses. For these
reasons, Boud, Dunn, and Hazel-Hegarty (1986) suggest that the laboratory is the best
place to identify student misconceptions and bring about conceptual change
Conceptual Change in the Chemistry Laboratory
This section of the literature review will summarize the numerous goals and
objectives of laboratory teaching. The section will also describe the pervading
instructional pedagogies implemented in the undergraduate chemistry laboratory.
Goals of Laboratory Work
The undergraduate science laboratory is generally viewed as a panacea. It has
been likened to the role of sacraments in religious observances, an outward and
visible sign that students are acquiring scientific grace (Hegarty-Hazel, 1990b, p. 3).

The importance of the laboratory component to undergraduate science education can
be inferred by the number of goals ascribed to the laboratory. A comprehensive
review of laboratory instruction has identified 1,547 distinct objectives of the science
laboratory (Boud et al., 1986). The purposes of undergraduate science laboratory, as
compiled from the recent literature, can be broadly grouped into five categories as
Conceptualization and Illustrative Goals. Laboratory work illustrates
reactions, principles, and theories discussed in lecture (Abraham et al., 1997; Domin,
1999b), illustrates mechanism of reactions (Pickering, 1988), and makes abstract
concepts more concrete (Hegarty-Hazel, 1990a; Moody & Foster, 1997).
Cognitive Goals. Laboratory work helps students learn generalized, systematic
ways of thinking that can be transferred to other disciplines (Hegarty-Hazel, 1990b).
Lab also offers unique opportunities to diagnose and dispel misconceptions (Boud et
al., 1986), promotes critical thinking and problem-solving abilities (Kandel & Ikan,
1989; Kharas, 1997; Kirschner & Meester, 1988; Lazarowitz & Tamir, 1993;
Pickering, 1988), helps students learn knowledge and skills that can be transferred to
new and unfamiliar situations (Kirschner & Meester, 1988), and helps students
remember the central idea of an experiment over a significantly long period of time
(Hegarty-Hazel, 1990b).

Psychomotor Goals. Laboratory work teaches manipulative skills (Boud et al.,
1986) and provides experience using equipment and instrumentation (Hegarty-Hazel,
Processing Goals. Laboratory work helps develop scientific inquiry skills,
such as observation, description, estimation of measurements, data manipulation, and
evaluation of results (Edelson & O'Neill, 1994; Kirschner, Meester, Middlebeck, &
Hermans, 1993; Kirschner & Meester, 1988; Lazarowitz & Tamir, 1993; Rubin,
1996). Laboratory work helps students understand science as a process of scientific
inquiry and hypothesis generation (Hodson, 1996; Shiland, 1993).
Affective Goals. Laboratory work provides a model of scientific inquiry
(Hegarty-Hazel, 1990a) and fosters a sense of success, motivation, and control (Boud
et al., 1986; Lazarowitz & Tamir, 1993). Laboratory work promotes positive attitudes
towards science (Freedman, 1997).
Actualized Learning Outcomes from Laboratory
From these lofty claims, the undergraduate science laboratory would appear to
be an effective strategy for changing students misconceptions (Tobin, Tippins, &
Gallard, 1993). Unfortunately, little evidence suggests that these goals are being
achieved. Although 37 reviews of science laboratory instruction were published
between 1954 and 1990 (Lazarowitz & Tamir, 1993), surprisingly little quantitative

or qualitative research studies have evaluated the effectiveness of learning in the
laboratory. According to Boud, Dunn, and Hegarty-Hazel (1986), evaluative research
suffers from seven main shortcomings, which include: (a) unsuitable comparisons in
studies using dissimilar instructional methods, (b) inappropriate and poorly described
research methodologies, (c) inappropriate instrumentation used to measure
performance, (d) using factual data rather than practical skills to evaluate laboratory
performance, (e) lack of an equivalent control group, (f) insufficient length of
treatment, and (g) trivial nature of research study (Boud et al., 1986). Because of
these flaws, little real insight about the effectiveness of laboratory work has been
forthcoming. The research is even more sparse about the organic chemistry
laboratory, where most research studies involve descriptive studies that use anecdotal
evidence or compare non-equivalent classes (Browne & Blackburn, 1999; Duchovic,
1998; Gallet, 1998; Glaser & Poole, 1999; Moody & Foster, 1997; Neeland, 1999;
Ram, 1999).
From the available literature, however, Lazarowitz and Tamir (1993) conclude
that laboratory activities, as they are currently taught, do not enhance learning or
increase conceptual understanding of science (p. 120). Students retain little of what
they learn in lab and have difficulty applying what they do know (Gallet, 1998). In
fact, students who take laboratory with lecture do not perform better on conceptual
lecture based tests than students who dont (Burnett, 1983; Kirschner, Meester, &

Flansburg, 1972). Rubin (1996) concurs, noting that lab experiences do not develop
process and higher order cognitive skills. Laboratory work would thus appear to be a
poor return of knowledge in proportion to the amount of time and effort invested by
staff and students (Hilosky, Sutman, & Schmuckler, 1998). Kirschner and Meester
(1988) argue that lab is an ineffective method of promoting student learning and that
skills acquired through lab work might be better acquired elsewhere. In fact, the
authors find it a paradox that:
A degree in the natural sciences that does not include a rather large amount of
laboratory work is considered at best a second-rate degree. At the same time,
it sometimes seems that the only skills that this laboratory work appears to
excel in achieving are the lowly-regarded manipulative skills. Why then do we
insist on long hours of laboratory work? (p. 83)
Hegarty-Hazel (1990) concludes that science education without some
laboratory experience is unthinkable, but equally, that student laboratory practice is
not a general panacea, the universal end to a multiplicity of means (p. 55).
The Laboratory Curriculum
The chasm between the goals ascribed to the laboratory and the realized
outcome of laboratory work is vast. With such lofty goals for the laboratory, how can
it be that so little learning occurs? Domin (1999) proposes two reasons for the
ineffectiveness of undergraduate chemistry lab: (a) students spend more time trying
to get correct results than in thinking about how the science principles are being

applied in the lab, and (b) most of the experiments stress lower-order skills such as
rote learning, memorization, and algorithmic learning .
The main problems facing the chemistry laboratory would seem to be
attributable to the laboratory curriculum, which has alternatively been criticized as
either too trivial or too complex and abstract (Hegarty-Hazel, 1990b; Landis et al.,
1998). Duchovic (1999) argues that laboratory experiences are little more than simple
demonstrations and tedious exercises. The type of laboratory experience a student
receives depends in large part upon the pedagogy and instructional methodology used
in the laboratory. Expository, inquiry-based, discovery, and problem-based
laboratories are four types of curricula used in undergraduate chemistry labs.
Expository Laboratory Instruction. In expository lab instruction, students
follow directions to arrive at a predetermined outcome, the purpose of which is to
illustrate an important reaction or to verify a principle or theory. The expository
approach is also known as verification or traditional (Domin, 1999b), formal
(Kirschner & Meester, 1988) or, more derogatorily, cookbook (Pickering, 1988), the
later term implying that undergraduate expository laboratory work requires little
thought and is little more than following directions to a recipe (Pickering, 1988).
Using Blooms taxonomy to classify general chemistry lab texts in terms of higher-
order and lower-order thinking skills, Domin (1999a) finds that the majority of the lab
texts appeal to the lowest three cognitive levels of knowledge, comprehension, and

application. None of the lab text activities he investigated required students to operate
at the highest three cognitive levels: synthesis, analysis, or evaluation (Domin,
1999a). Since the lab text dictates curriculum and pedagogy, a natural inference is
that high-level learning is not occurring in the lab. Rubin (1996) confirms this
assumption. He notes that lab students spend typically spend only two percent of their
time in lab asking higher-order questions. The problem is not unique with chemistry.
Similar analyses have been conducted on biology and physics lab texts, with
comparable results (Germann, Haskins, & Auls, 1996; Tamir & Lunetta, 1978).
Assessment of expository lab activities reveals that this type of experience does not
foster conceptual change (Gunstone & Champagne, 1990), does not foster critical
thinking skills (Gallet, 1998), and does not enhance student learning (Hofstein &
Lunetta, 1982; Tobin & Gallagher, 1987).
Inquiry-Based Laboratory Instruction. Inquiry and discovery labs are often
classified together because they are more open-ended than expository type of
laboratory experiments. However, they have quite different outcomes and different
procedural pathways. Inquiry labs have an undetermined outcome and require
students to propose hypotheses and design their own analytic procedures, while
discovery labs have a predetermined outcome, which the learner must discover
through a series of guided or open-ended activities. Inquiry labs stress higher-order
cognitive levels, including synthesis, analysis, and evaluation, and have been shown

to enhance formal operational thought and critical thinking (Bailey & Chambers,
1996; Chiou, 1995; Domin, 1999b; Koschmann, Feltovich, Myers, & Barrows, 1992;
Savery & Duffy, 1995; Tobin et al., 1993).
Discovery-Based Laboratory Instruction. In contrast to inquiry labs that may
have multiple outcomes, discovery-based labs (or guided inquiry, as they are
sometimes called to distinguish them for their more open-ended namesakes) have a
predetermined outcome. The purpose of discovery-oriented labs is to allow students
to explore specific examples and deduce from the results a general understanding of
the inherent scientific principle (Domin, 1999b; Greenbowe & Burke, 1995;
Pickering, 1988; Tobin & Gallagher, 1987; Tobin et al., 1993). Discovery labs are
most successful when based upon a single concept and provide students with a
procedure to arrive at predetermined, yet unspecified outcome. Discovery labs should
also provide opportunities for individual reflection, followed by class discussion
(Bodner, Hunter, & Lamba, 1998).
Problem Based Laboratory Instruction. Problem-based instruction, also called
project-based instruction, has been used extensively in medical and dental schools,
but has not been widely implemented in undergraduate chemistry laboratory courses
(Collis, 1997; Lebow, 1993; Savery & Duffy, 1995; Savery & Duffy, 1996; Williams,
1992). In problem-based labs, students analyze real-world problems, propose
hypotheses, acquire relevant knowledge, and then design a procedure to solve the

problem. Although multiple solutions may exist, the outcome is predetermined and
requires deductive reasoning to arrive at the solution (Domin, 1999b). The premise of
problem-based learning is that students will be motivated to learn requisite
knowledge in order to solve a challenging task that is compelling, authentic, and
complex (Ram, 1999).
Because problem-based laboratory experiments use a deductive approach,
chemistry students must be familiar with the underlying scientific principle before
doing the experiments (Domin, 1999b). More time is required for both the instructor
and the students. Because of these problems, problem-based learning, when used in
undergraduate science labs, seems to be relegated to upper division courses, such as
physical chemistry. However, project-based learning has been implemented in some
organic lab courses and results are generally positive: Empirical evidence suggests
that students become more responsible learners, display more curiosity, develop a
sense of community, are able to transfer knowledge to other disciplines (Ram, 1999),
and develop more positive attitudes about the laboratory (Kharas, 1997)..
Analysis of Laboratory Instruction
Although research on the effectiveness of alternative laboratory instructions
has generally been positive (Bodner et al., 1998; Domin, 1999b; Rubin, 1996), there
is a strong perception among faculty that too much time is spent on process at the

expense of content (Domin, 1999b). Problems may also arise when students are
cognitively unprepared for the lab: These students may not be able to discover the
intended scientific outcome of the lab or to suggest solutions to problems of which
they have no conceptual understanding (Hodson, 1996).
In spite of the benefits that appear to accrue from inquiry-based, discovery,
and problem-based laboratory curricula, the vast majority of undergraduate chemistry
labs still use expository methods. In a recent survey of laboratory teaching, over
ninety percent of reporting colleges and universities indicated that experiments are
generally expository, where students follow step-by-step directions (Abraham et al.,
1997). Critics conclude that in a typical undergraduate chemistry lab, students
perform the same experiments that were done by their parents and synthesize
compounds without having any conceptual understanding of the chemistry involved
(Gallet, 1998; Pickering, 1988).
Pedagogical Benefits of the Laboratory
With all of the problems associated with the laboratory, it might be tempting
to conclude that laboratory work is unimportant to learning. However, this is not so:
Hands-on activities are found to correlate with improved achievement, particularly
for younger students (Hegarty-Hazel, 1990b). The work of Stohr-Hunt (1996)
demonstrates that frequency of hands-on laboratory use increases achievement on a

standardized science achievement test. Furthermore, laboratory activities provide
unique possibilities for conceptual change strategies that can help students develop
better conceptual understanding (Boud et al., 1986).
Even the most vehement critics of laboratory instruction do not endorse doing
away with the laboratory, but rather making it more responsive to student learning
and more conducive to conceptual change (Bodner et al., 1998). One solution to the
need to offer a high quality laboratory experience and the increasing difficulty in
doing so is to use computer-based laboratory simulations.
Computer Simulations and Conceptual Change
Computer simulations have the potential to address misconceptions, promote
conceptual understanding of molecular processes, improve visualization, and effect
conceptual change (Bobbert, 1982; Burchfield, 1995; Clariana, 1989; Gammon, 1995;
Hiltz, 1994; Kemer, Penner-Hahn, Berger, & Dershimer, 1997; Lorson, 1991; O'Neill
& Gomez, 1994; Owston, 1997; Prosser & Tamir, 1990; Reeves & Reeves, 1997;
Roberts, 1993; Teles & Collings, 1997; Vaidyanathan & Rochford, 1998). This part
of the literature review will examine the attributes of computer simulations, describe
research where simulations have been used in laboratory experimentation, and discuss
attributes of computer simulations.

Computer simulations encompass a wide range of technologies. Lunetta and
Hofstein (1981) categorize simulations into several levels, including videos or films
of actual events, three-dimensional models, recreations of actual events, and
hypothesis generation (Eisenkraft, 1987; Lunetta & Hofstein, 1981). Conversely, Lee
(1999) classifies simulations into two categories, based upon the extent of learner
control and degree of interactivity. The two categories are presentation and practice,
with the latter being more effective in promoting student learning. To Prosser and
Tamir (1990), simulations are revelatory computer-assisted learning in which the
user is guided through a process of discovery so that subject matter and underlying
theory are progressively revealed (p. 272).
Research studies of computer simulations have tended to focus on the delivery
platform rather than the pedagogy behind the simulation. As such, studies have been
organized around interactive videodisks (Katkanant, 1990; Strauss & Kinzie, 1994;
Zirkel & Zirkel, 1997), web-based simulations (Dede, 1996; Kinzie, Larsen, Burch, &
Baker, 1996), and computer-based simulations (Clariana, 1989; Freeman, 1995;
Joolingen & Jong, 1991; Windschitl & Andre, 1998). Based on Clarks (1983, p. 455)
oft-quoted allegation that media can never influence learning any more than a truck
bringing groceries influences nutrition, the pedagogical importance of computer
simulations should not depend upon whether the lab is delivered via videodisc,

computer screen, or the World Wide Web, but on the underlying instructional
methodology and pedagogy.
For the purposes of this research, then, the terms computer simulation,
simulation, interactive computer simulation, and interactive computer laboratory
simulations are used interchangeably and refer to multimedia computer programs
that seek to represent real experiments, using animation, interactivity, learner control
and feedback. The importance of these factors to the learning process will be explored
later. What is not included in the definition is strict adherence to any delivery
platform: Simulations will be described that are delivered via the computer, the
World Wide Web, or CD-ROM technology.
Regardless of the platform by which instruction is delivered, computer
simulations play many different roles in the undergraduate science laboratory,
depending upon the perceived goals of the laboratory. Interactive computer
simulations may be used as supplements to, replacements of, or preparation for a
traditional laboratory experience (Bobbert, 1982; Grosso, 1993). This section of the
chapter discusses the perceived benefits and limitations of computer simulations in
these various roles.

Benefits of Computer Simulations
Regardless of the role played, computer simulations have substantial and often
unique advantages as compared to traditional laboratory instruction. Twenty attributes
of computer simulations have been identified from the recent literature. Computer
simulations can:
1. Replace experiments that use hazardous materials (Smith, Jones, & Waugh,
2. Reduce cost (Fletcher, Hawley, & Piele, 1990).
3. Replace experiments that occur too quickly or too slowly to be done in a
regular laboratory period (Herron & Nurrenbem, 1999).
4. Reduce cognitive noise, so that students can concentrate on the concepts
involved in the experiments (Clariana, 1989).
5. Provide feedback to enhance conceptual understanding (Chickering &
Ehrmann, 1987).
6. Provide dynamic animations to emphasize the molecular level of chemical
reactions (Williamson, 1992; Williamson & Abraham, 1995).
7. Permit students to collect a multitude of data quickly (Vining, 2000) so as
to generate and test hypotheses (Joolingen & Jong, 1991).
8. Permit students to generate and test hypotheses.
9. Engage students with high level of interactivity (Grosso, 1993).

10. Present a simplified version of reality by distilling abstract concepts into
their most important elements (White, 1993; Windschitl & Andre, 1998; Zietsman &
Hewson, 1986), making abstract concepts more concrete (Rieber & Parmley, 1995).
11. Provide a complex, contextualized, dynamic environment (Park &
Hannafin, 1993).
12. Standardize instructional pedagogy, teaching, and content across multiple
lab sections (Abraham et al., 1997; Hilosky, Sutman, & Wang, 1997).
13. Actively engage students in scientific inquiry (Grosso, 1993).
14. Reduce ambiguity and help identify cause and effect relationships in
complex systems (Clariana, 1989).
15. Serve as pre-laboratory preparation (Bobbert, 1982).
16. Foster problem-solving skills (Katkanant, 1990; LaJoie, 1993).
17. Promote critical thinking skills (Bonk & Reynolds, 1997).
18. Foster deeper understanding of physical phenomenon (Baird, 1989) and
chemical reactivity.
19. Help students learn about the natural world through seeing and interacting
with underlying scientific models that would not be readily inferred through first-
hand observation (Krajcik & Lunetta, 1987).
20. Enhance concept formation (Bertrand, 2000) and promote conceptual
change (White, 1993; Windschitl & Andre, 1998).