Education

The Crossroads between Biology and

Mathematics: The Scientific Method
as the Basics of Scientific Literacy

ISTVAN KARSAI AND GEORGE KAMPIS

Biology is changing and becoming more quantitative. Research is creating new challenges that need to be addressed in education as well. New
educational initiatives focus on combining laboratory procedures with mathematical skills, yet it seems that most curricula center on a single
relationship between scientific knowledge and scientiflc method: that of the validity of knowledge claims, judged in terms of their consistency with
data. Collecting data and obtaining results (however quantitative) are commonly part of science, but are not science itself We envision that the
operative use of the complete scientiflc method will play a critical role in providing the necessary underpinning for the integration of math and
biology at various professional levels.

Keywords: scientiflc method, BIO2010, hypothesis testing, validation, modeling and simulation

D 'Arcy Thompson began his book On Growth and Form

(1917) by quoting Immanuel Kant: "...chemistry
was a science, but not a Science...for that the criterion of
Modern scientific advances have transformed the life
sciences, but until recently they had little infiuenee on under-
graduate training, leaving an unprecedented gap between
true Science lay in its relation to mathematics." Thomp- teaching and research (McComas 1998, Abell and Lederman
son went on to explain how chemistry was elevated to the 2007, Kerfeld and Simon 2007). In 2003, the National Research
level of "Science," whereas biology had not yet reached Council issued BIO2010, a report that suggested that biology
that level. The search for universal quantitative laws of should become more quantitative, an issue to be addressed
biology can be daunting, given the extraordinary com- both in research and education (NAP 2003). Biological
plexity of biological systems (West and Brown 2004). Yet researchers and educators today are closely collaborating
when such an interdisciplinary approach proves success- with mathematicians and scientists from other fields. Even
ful, it often provides a major breakthrough in biology (as if tomorrow's biologists had a more extensive mathematical
was first exemplified by the discovery of the Mendelian and computational background, a single person could not, in
laws of inheritance). We are now starting to understand general, pursue all these fields in depth; thus, the formation
that the amazing complexity and diversity of living or- of interdisciplinary collaborations continues to be essential
ganisms commonly stems from ultimately simple rules in the pursuit of biology (Karsai and Knisley 2009).
that can be explored by computational and mathematical These days, mathematics enters at every stage of science: in
means (Kauffman 1993). Although Thompson's attitude designing an experiment, seeking response patterns, and in
is widely present in the scientific culture today, we main- the search for underlying mechanisms. The interdisciplinary
tain that mathematics alone will not turn a field into approach is at the heart of many research areas, such as in
science, but the application of the scientific method does. genomics, where the size and complexity of the data sets and
To illustrate this, we intend to show how mathematics, the scale of the problems require the joint expertise of com-
inquiry-basedlearning, and the application of a modern puter scientists, statisticians, and biologists. However, while
philosophy of science could produce pedagogy to better this type of collaboration is becoming more common in
teach biology as a science. In this article, we define the research, the real challenge seems to begin at the undergradu-
"scientific method" as the sorts of things scientists do, ate level (Musante 2005), where we must train a workforce
and contrast this with the demonstrative methods more that is able to do collaborative work efficiently. According
frequently used in current instruction. to major surveys, American students graduate from college

BioScience 60: 632-638. ISSN 0006-3568, electronic ISSN 1525-3244. © 2010 by American Institute of Biological Sciences. All rights reserved. Request
permission to photocopy or reproduce article content at the University of California Press's Rights and Permissions Web site at www.ucpressjournals.com/
reprintinfo.asp. doi:10.1525/bio.2010.60.8.9

632 BioScience • September 2010 / Vol. 60 No. 8 www. biosciencemag. org

 Education

poorly prepared to enter the math, science, and technology Most undergraduates never meet anything close to real
workforce (Stukus and Lennox 1995, NAP 2003, Abell and science, and are exposed to "researcb activities" in a class-
Lederman 2007). We argue that simply introducing more room setting only. Simple investigation or inquiry, altbougb
mathematics to biology majors will not solve this problem. called "research," is in reality not researcb at all (Ortez 1994,
We believe tbat for matbematics to make sense in biology Mrosovsky 2006). Using the Google searcb engine and
education, science should make sense first. Tbe issue bas two collecting (i.e., literally "researcbing") information into a
interrelated aspects, and we deal with them separately: They report, or following a cookbook lab protocol are examples of
are the understanding of science and tbe understanding of "researcb" tbat in fact do not use the methods of science. By
matbematics for science (in this case, for biology). concentrating exclusively on sucb exercises, it is easy to lose
focus of wbat real researcb is all about. Pedagogy sbould not
Inquiry-based approach give in to logistical problems. Attacking a smaller number of
Inquiry-based investigation is widely publicized as a basis problems in greater deptb could be a solution tbat is closer
of science instruction. The National Research Council's to real researcb. For example, Felzien and Cooper (2005)
National Science Education Standards (1996) encourage developed an "Introduction to Research" course in which
teachers to focus on inquiry, where students are expected students especially valued the assignments to write grant
to formulate their own questions and devise ways to answer proposals. Students said this was tbe toughest task, but that
them. Tbese are generally inductive activities that require it helped thetp most to understand tbe biological research
students to work out their own procedures, collect tbeir data, process. Demers (2003) developed a well-rounded, student-
present and analyze tbose data, and derive conclusions from driven, inquiry-directed course: It begins with epistemologi-
the results they obtain (DeBoer 1991). Because data collection- cal definitions, discusses science and nonscience, investigates
and processing are quantitative ways to study biology, fresh- etbics, and develops critical thinking, all witb an overview of
man inquiry-based labs are commonly tbe first venues wbere tbe basic model of tbe scientific process. Then, students as
students study biology using statistics or simple math. a group are asked to apply tbeir newly learned awareness to
tbe discussion of selected researcb problems. In our opinion,
As a pedagogical approach, of course, inquiry is full of
tbese approacbes provide better intellectual preparation to
merits, and inductive activities tend to bring undergraduates
learn about bow to do science tban attacking different bio-
closer to understanding science (Kerfeld and Simons 2007).
logical problems eacb week.
However, the integration of this approach into tbe curricu-
lum commonly has flaws, and tbe true nature of inquiry is "A lab is wbere you do science" (Tbornton 1972). Tbis
often forfeited in this process (Edwards 1997). Publisbed view is often heralded in tbe literature, but we disagree.
researcb material is frequently too heavily structured and Ratber, it is in tbe investigative mind wbere we do science;
too complex for undergraduates, so instead of an original tbe laboratory offers only an opportunity to test scientific
researcb project, preselected substitutes—so-called real-life bypotbeses tbrougb predictions (tbe products of tbe mental
problems—are investigated in tbe classroom, in a seriously process that constitutes science). By contrast, most science
canned manner. Understandably, tbere are common tecbni- curricula tend to focus on a single relationsbip between
cal and' logistic constraints (one of tbem is time), but tbe scientific knowledge and scientific metbod: the validity of
real danger is that tecbnical difficulties in implementing an scientific knowledge claims, judged exclusively in terms
inquiry can cbange tbe whole pedagogy: Using scientific of tbeir consistency, witb observable evidence (Hodson
inquiry without first teaching the proper scientific metbod 1998). Most instructional laboratories incorporate only a
may generate a complete misunderstanding of how science few selected steps of inquiry even in bands-on experiments
works. Asking questions, collecting data, and obtaining an (Harker 1999). Harker argues tbat full participation in eacb
answer from tbe latter are parts of the scientific method, step of tbe scientific metbod would be necessary instead,
but do not wbolly constitute tbe scientific metbod itself and be also presents a positive example from his microbial
(Musante 2009). pbysiology course (Harker 1999; see box 1 for an example of
Going a step furtber, it is true that the scientific method our own approacb to inquiry).
can be best learned tbrougb researcb (Rotb 1995). So again, Scientific literacy doesn't necessarily call for deep under-
witb rigbt sentiment, many argue tbat getting undergradu- standing of difficult concepts such as the Nernst equations or
ates involved in researcb is important (McComas 1998). tbe precise conditions of tbe Hardy-Weinberg equilibrium,
However, we argue tbat there is too much of a difference but it does require a general understanding of basic scientific
between engaging in researcb (sucb as a single project tbat notions and tbe nature of scientific inquiry (Gross 2006).
continues for years) and doing "researcb projects" in tbe Becoming a successful researcher requires tbe learning of
classroom. Many students are initially uninterested in sci- many skills. However, focusing on any special skill, wbetber
ence and some are actually afraid of it (Demers 2003). Many quantitative or not (sucb as doing BLAST [basic local
science majors are attracted to bealtb professions but lack alignment searcb tool] searcbes for comparing genomes),
awareness of bow science operates in general, or bow this will provide only tbe tools required by tecbnicians, not
knowledge is important for tbeir chosen career (Felzien and scientists. To be a successful researcber, the most important
Cooper 2005). skill to have is tbe self-sufficient use of the scientific method.

www. biosciencemag. org September2010/Vol. 60No. 8 • BioScience 633

Education

Box 1 . Is body weight inherited or acquired? An exampie of the use of the scientific method in a freshman bioiogy ciass.

. '.At East Tennessee State University, we have developed'a' lab exercise that riot only addresses the understanding of an important bio-
fcgical ççncept but also specifically fosters the use of a particular aspect of the scientifi¿ method, the testing of multiple hypotheses'
Ijohnson et al. 2Ö06).
Heritability is the proportion of phenotypic variance in a population that can he attributed to genetic differences among individuals
rather (han to environmental factors. The main steps we follow to introduce the concept are:
Stage 1: Turn the question ("Is weight/height inherited?") into a series hypotheses and [predictions (HW denotes heritability of
weight). Then, discuss the differences between statistical and scientific null hypotheses.!
H^: HW does not depend on the sex (it is not different for males and females)
|1^: HW depends on the sex (it is different for males and females)
Using logic (such as inductive and deductive reasoning) as a tool, testable claims (predictions) are derived from these hypotheses:
F,,: HW is not different for the two sexes
F,: HW is smaller for males
F^: HW is smaller for females
. Ijiote: We also create a similar hypothesis-prediction tree for the heritability of height. Our goal is to formulate at least one specific
prediction derived from each hypothesis.
Stage 2: Experimental design. -
What needs to be measured and how? What are the costs and benefits of different, experimental approaches? A discussion is con-
ducted and a plan is formulated for selecting basic data for the investigation, using the weight of parents at the time when their
. age was similar to that of the student today. • •
.Stage 3:.Data collection'and processing".
• "We eniploy an online data collection form, previously developed for submitting data arionymously to a database. Data submission is
; «jheniewark and is voluntary. Data from several parallel classes and many years, accumulated in the database, are used for a regression
< i*ialj;sis by the students. This requires the use of a preprogramhied Excel worksheet that contains explanations and instructions On
^ hbw; to process the data. •. ' ' ' ,
.''lStágé-4.: évaluation of the results. • ,
'• *This'¿esíults" we obtained in a particular class are: male weight-?!^ = 86% and female wei'ght h^ = 57% (Johnson et al. 2006).
' •We|:pinpare these results to our original hypotheses and predictions and conclude thai the results supported P^ (and did not sup-
'!%ort'P|,-ór P',).' Wfe carry out the same steps for the heritability of height and reach similar conclusions. We compare our results to
'• .ítSfié professional literature (Brown et al. 2003) and discuss the differences in methodology and results.
'tStag^ 5i «Planning for the next step of investigation. • •
'¡|Jow, we discuss possible reasons for the significant difference between males and females. We construct new testable hypotheses for
. |jrtheristudies (e.g., how peer pressure may influence teenage girls and boys to control their weight differently, etc.).

Although this is one of the most difficult skills to acquire, which students are exposed is one of the major reasons many
university education tacitly expects students to pick it up students turn away from science at an early age (Holton
on the fly, and it is also assumed that faculty mentoring will 1992). It is important to emphasize to students that every
help this process. However, experience says otherwise. The experiment is set up within a theoretical, procedural, and
scientific method is no trivial matter, and it appears that sci- instrumental matrix, and it is this theoretical understanding
entists are less interested in it than are philosophers (Salmon that gives a purpose and a form to the experiments (Hodson
1989, Norton 1998). Studies such as that by Lombrozo and 1998)—in short, that an experiment outside the matrix of
colleagues (2006) indicate that students of various institu- the scientific method is junk. It is also imperative to demon-
tions carry several misconceptions about science and lack a strate that alternative hypotheses can generate identical pre-
sufficient understanding of how scientific views differ from dictions, so there is no crucial experiment to decide between
everyday opinions or even religious claims. them, and that obtaining negative results and anomalous
The misrepresentation of science and the incorrect use data is a natural feature of science. The challenge of science
of the scientific method has generated various myths and is exactly how to make progress despite these complexities
distorted views of science that are strongly rooted in the sci- and others. To face them, future researchers must be trained
entific mindset of the 1960s and early 1970s (Hodson 1998). in a more targeted way.
For example, experiments are often thought to be decisive At the level of the "big picture," such issues are commonly
and universally essential for testing hypotheses, whereas in dealt with in the context of nature of science (NOS) dis-
reality, no theory-independent experiments are possible; the cussions (McComas 1998, Abell and Lederman 2007). The
method of data collection used in testing a given hypothesis NOS deals with an important, general overview of science
and also the formulation of the hypothesis itself are dictated education that we do not intend to address here. Instead, in
by the very theory under review (Hodson 1988). There is the following sections, we hope to focus more concretely on
ample evidence that the current distorted image of science to the particular issue of why the quantitative approach is not

634 BioScience • September2010/Vol. 60No. 8 www. bioscienceniag.org

 Edueation
identical to the scientific approach, and how mathematics out true") is very different in quality: A positive outcome
asid statistics can (and cannot) play an important role in merely indicates that the available evidence is not against the
teaching biology as science. hypothesis. To say more than that, we would need to com-
pare predictions from various alternative hypotheses.
Can mathematics heip students better There are further issues to consider here. Hypothetico-
understand bioiogy? deductivism, often taught as the "right way" to do science,
Is focused quantitative training in biology a solution? Virtu- maintains that the task of scientific theories is to explain and
ally all educators agree that otir goal should be to develop predict facts about observed data. We rarely explain to stu-
students' capacity for critical thinking and problem solving dents that such empirical adequacy is insufficient by itself,
(Gross 2004, Abell and Lederman 2007). As a pioneering or that consistency with data and the validity of a hypothesis
example, the University of Tennessee implemented a process do not grant "truth" status to a theory. This important mat-
called "multiple routes to quantitative literacy" {www.tiem. ter is sometimes overlooked even by some scientists, who as
utk.edu/bioedf), in which quantitative topics were intro- individuals fight fervently for their pet theories (Mitroff and
duced to general biology courses, and math courses were Mason 1974). In reality, consistency with data signifies that
redesigned to provide more varied and informative topics the theory may he correct, but numerous other theories may
for students of the life sciences (Gross 2004). Including bio- also apply (a classic source is Duhem 1962).
logical data and biology-loaded topics in math courses and Next, a failure to draw a distinction between data and
quantitative training in biology courses has become a suc- phenomena is particularly deceptive (Haig 1996). In sci-
cessful and applauded recent trend, with many more recent ence, it is phenomena and not the bare data that we want
examples, but there has been no comparable success in inte- to decipher. Haig (1996) asserts: "Phenomena are relatively
grating calculus and statistics with biology at the freshman stable, recurrent general features of the world that we seek to
level (Karsai and Knisley 2009). explain.... Data, by contrast, are idiosyncratic to particular
May (2004) pointed out some of the dangers practic- investigative contexts. They are not as stable and general as
ing biologists face when attempting to use computational phenomena." Data are important, of course, because they
methods without a good understanding of the underlying serve as evidence for the phenomena. When we want to
mathematics; there is a similar danger when mathematicians render phenomena from data, we often employ some kind
try to do research in biology. The interdisciplinary bridge of data-processing methods such as provided by statisti-
between biology and chemistry, or biology and physics, has cal tools. However, we should be aware that these methods
been smoother. Such projects may require collaborations help detection, but not explanation (Haig 1996, Maret and
from scientists with different backgrounds, yet they all use Ziemba 1997). Therefore, implementing statistics in biology
the same approach, or "language," housed within the sci- courses without understanding the scientific method actu-
entific method they share. The interaction of mathematics ally misrepresents how science works. Paradoxically, this
with science has never been smooth (exemplified by the danger is even higher when a statistical course uses genu-
difference between Baconian and Cartesian science and ine biological data, because although the students obtain
the resulting fight between empiricism and rationalism in the vital statistical skills, the lively essence of the scientific
the 17th and 18th centuries). We believe this is because knowledge process is easy to miss, since data are taken at face
mathematics and science have significantly different roots value, and are taken as "given."
and approaches. Mathematics is a language that develops Similar problems occur also when mathematics (calcu-
its own internal structure through proving theorems. lus, for example) is integrated into biology and tnath takes
Mathematics and statistics do not use the methods of the over the course. If the course changes track, concentrating
empirical sciences. Understanding science is a first step in on mathematical tractability and skill training instead of
clarifying this; understanding the role of mathematics in scientific meaning, it will not progress in the same way that
science is a second. a typical science course would; that is, through the study of
Sometimes things get mixed up. For example, the testing alternative hypotheses or learning about historical shifts of
of scientific or statistical hypotheses, although they may scientific paradigms. For example, the well-known Lotka-
sound very similar, are in effect profoundly different activi- Volterra predator-prey model generates regular oscillations
ties, mainly because of the different epistemological ground- in population size. The model is typically, considered one
work involved. Their differences can generate confusion, of the most important mathematical equations in biology
especially in inquiry-based labs or when biological examples (Jungck 1997), although the system is known to be math-
are used in math classes. Students commonly fail to under- ematically unstable against modifications, spatial variations,
stand the distinction between a scientific hypothesis and a and stochasticity, and is therefore unlikely to be biologically
statistical hypothesis (Maret and Ziemba 1997). In short, relevant (Murray 2002). Beyond doubt, the Lotka-Volterra
in math, the acceptance of a hypothesis implies that our equations provided very important inspirations to ecol-
conjecture is true by itself or within a well-calculated degree ogy and other fields of science, such as complexity science.
of certainty. By contrast, in science, the certainty obtained However, there are virtually no biological data that fit the
when a hypothesis is confirmed (its predictions "come predictions of this model. The biggest problem, nevertheless.

www.biosciencemag.org September2010/Vol.60No.8- BioScience 635

Education

is not the lack of such predictive value but the biologically (1988) urged us to change our traditional view: "In the old
unacceptable assumptions behind the model. The Lotka- stereotyped school curriculum view of science, scientifk
Volterra model is on the basis of assumptions that do not knowledge exists 'out there' and scientists carefully, system-
hold in biology, such as aggregated food resources. In atically and exhaustively collect information that reveals it."
engineering and the computational sciences, trusted master The hypothetico-deductive method is still heralded as the
models (such as the Hamiltonian or Maxwell equations) are main model for science, and other approaches such as the
available, and the question is how to best exploit them. A grounded theory (Glaser and Strauss 1967) are not even
similar approach is misinformative in biology, where such mentioned, although they have been around for decades.
master models are not present, and analysis is subordinate to "Logico-deductive theorizing," as Glaser and Strauss (1967)
discovery—an integral part of the knowledge process. Even pointed out, exaggerates the significance of theory testing
well-known constructs such as the Lotka-Volterra model (which is not concerned with the theory's origin or develop-
cannot be used as master equations: Tweaking and twisting ment, only with its validation ) and denies the role of inductive
them to give better predictions, as in engineering problems, reasoning. However, in reality, most hypotheses and theories
does not present as much new insight as dealing with the tend to be underdeveloped, and as a consequence research-
phenomena and trying to build better models. ers usually submit "low-content" theories to early empirical
We think that the problems mentioned above are rooted testing (Haig 1996). The presupposition of the hypothetico-
in the epistemological differences between the fields of deductive method, that theories arise in a full-blown form,
mathematics and biology, and that these differences need to should be shifted to a more dynamic perspective on theory
be exposed. In mathematics, theories are laid out explicitly construction, in which a theory becomes an ever-developing
and in advance, as in the theory of equations or the theory entity, interwoven with data and hypotheses.
of complex variables. Results are obtained analytically; that When doing research, we often do not possess actual
is, by proving properties. The model's description is typically knowledge of the causal mechanisms that we abductively
complete and the standard of correctness is mathematical probe. Constructing models by analogy, drawing on
proof. Biology obtains results in a very different way. Here, mechanisms we already know, helps researchers construct
just as everywhere in science, the basic mental construct new theories. Reality is commonly simulated in a concrete
takes the form of a hypothesis. Hypotheses cannot be proven visual image, such as a stock-and-flow model, where the
or disproven, only supported or unsupported through tests causal mechanisms are drawn from the domain of previ-
of their predictions. Theoretical and computational models ous experience in other disciplines. Biology in particular
(in the same way as laboratory experiments or observa- uses many mechanical and electric circuit analogs. This
tions) can serve as additional tools to test the predictions kind of "abductive explanatory inferentialism" suggests
of a hypothesis and to perform limited experiments in that the theory of the scientific method is centrally
the sense of "what if" scenarios. Besides, in biological concerned with generating theories in a "backward" sense.
investigations, simulations are often preferred to formal The approach is also very close to how scientists gener-
models, because their assumptions can be more realistic ate models and work with them in practice, in terms of
(e.g., for prey-predation interaction, local models typical in what philosophers have come to call "inference to the best
simulations can be favored over aggregate models typical in explanation" (Harman 1965).
equations). Models in biology are seldom proved mathemat- We can safely predict that the strongest effect of math
ically; instead, model predictions are compared with natural on biology and biology education will be the extensive use
findings, and sensitivity analyses check how the model varies of models and simulations, as has happened in other fields
with a few selected parameters. of science (Clement and Rea-Ramirez 2008). Since 1998,
In other words, instead of looking for a complete proof, studies using models have increased more than fourfold in
the biologist marshals evidence to present the claim of a scientific literature (Keeling and Rohani 2007). Statistics
hypothesis beyond reasonable doubt. A hypothesis that is continues to have an important role in providing tools for
well supported and whose alternatives do not receive supe- testing predictions and constructing statistical descrip-
rior evidential support may eventually become a theory. tions, but on the other hand, the role of mathematical and
For example, evolution is a successful theory, but there is computational modeling becomes ever stronger and will
no way to prove that the theory of evolution is correct in a infuse biology in all phases, from hypothesis abduction to
mathematical sense; however, the scientific method does not ' the testing of alternative "what if" scenarios. Although more
require that. Focusing on formalism, truth, and proof thus formal models (e.g., differential equations) will continue to
misinforms students about science. inspire quantitative thinking, we foresee that their roles will
be increasingly augmented and partly replaced by simula-
Philosophy of science, modeling, and simulations tions. Computer simulations can be used as effective tools
Clarifying science, and clarifj'ing the role of math in science, for collaborative research and education as well, and many
are important steps. But there may be more to come. Model- biology researchers and students can access these tools today
ing and simulation might offer a farther step forward, while without extensive special mathematical training. Collabora-
bearing on the former two. To see how, consider that Hodson tion between a mathematician and a biologist through a

636 BioScience • September 2010 / Vol. 60 No. 8 www.biosciencemag.org

 Education

simulation platform can help a project stay focused as a Acknowledgments

genuinely scientific enterprise that uses the full power of the The authors thank Wayne G. Basier for establishing the
scientific method. In such a system, the biologist can keep Chair of Excellence for the Integration of the Arts, Rhetoric
track of biological assumptions, and the mathematician can and Science and East Tennessee State University for sup-
help the biologist avoid using naive math, thus inspiring porting George Kampis as Basier Chair and Istvan Karsai as
the biologist to understand the importance of embedded Basier Fellow in 2007.
assumptions (such as the assumptions about distributions,
randomness, and so on; Brent 2004, May 2004).
References cited
We believe that simulation platforms using individual- Abell SK, Lederman NG. 2007. Handbook of Research on Science Education.
based approaches especially will provide excellent educational Lawrence Eribaum Associates.
tools for biology. Today, integrated platforms and informa- Brent R. 2004. Intuition and innumeracy. Cell Biology Education 3: 88-90.
tion systems such as Visual Cell, the Virtual Physiological Brown WM, Beck SR, Lange HM, Davis CC, Kay CM, Langefeld CD, Rich
Human, and others provide templates where data, models, SS. 2003. Age-stratified heritability estimation in the Framingham
Heart Study families. BMC Genetics 4. doi:10.1186/1471-2156-4-
sitnulations, theories, mechanisms, and qualitative and S1-S32
quantitative knowledge are represented in a uniform and Clement )J, Rea-Ramirez MA. 2008. Model-based Learning and Instruction
transparent fashion that makes it easy to experiment, form, in Science, vol 2. Springer.
and evaluate models and alternative hypotheses (see Sauro DeBoer GE. 1991. A History of Ideas in Science Education: Implications for
2003 for earlier examples). This integration is highly instruc- Practice. Columbia University.
Deniers NE. 2003. Issues in science and technology. Student-driven inquiry
tive as it suggests a style for doing science by merging several
directed by the scientific processes. Journal of College Science Teaching
aspects of the scientific method into a uniform representa- 32: 332-337.
tional system while perhaps also serving as a useful template Duhem P. 1962. The Aim and Structure of Physical Theory. Atheneum
for science education. Although some of these integrated Press.
research tools are not easily accessible to all students, many Edwards CH. 1997. Promoting student inquiry. Science Teacher 64: 18-21.
open-source or commercial software platforms exist already Felzien L, Cooper J. 2005. Modeling the research process. Alternative ap-
proaches to teaching undergraduates. Journal of College Science Teach-
(e.g., Netlogo, or Starlogo TNG for agent-based modeling, ing 34: 42-46.
and Sage and Vensim for mathematical modeling). Using Glaser B, Strauss A. 1967. The Discovery of Grounded Theory. Aldine.
these accessible simulation platforms, we managed to suc- Gross LJ. 2004. Interdisciplinarity and the undergraduate biology curricu-
cessfully engage freshman biology students with minimal lum: Finding a balance. Cell Biology Education 3: 85-87.
math skills both in classroom exercises and in interdisciplin- . 2006. Scientific illiteracy and the partisan takeover of biology. PLoS
Biology 4: e 167. doi: 10.137l/journal.pbio.0040167
ary research that involved strong mathematical components
Haig BD. 1996. Grounded Theory as Scientific Method. Philosophy of Edu-
(Johnson et al. 2009). cation 1995 (Current Issues). University of Illinois Press.
Math and mathematics education can also gain from Harker AR. 1999. Full application of the scientific method in an under-
graduate teaching laboratory. Journal of College Science Teaching 29:
the infusion of biology, and modeling or simulation
97-100.
can be of help there, too. There is already a long list of Harman C. 1965. The Inference to the Best Explanation. Philosophical
mathematics problems that have arisen from biological Review 74: 88-95.
studies, ranging from the age structure of stable popula- Harre R. 1976. The constructive role of models. Pages 16-43 in Collins L, ed.
tions to qualitative calculus, which applies formalism to The Use of Models in the Social Sciences. Tavistock.
incomplete information (Kuipers 1994). The teaching Hodson D. 1998. Science fiction: The continuing misrepresentation of sci-
ence in the school curriculum. Curriculum Studies 6: 191-216.
of the concept of "rate of change" has traditionally been Holton G. 1992. How to think about the "anti-sciehce" phenomenon. Public
founded in analytical geometry, focusing on parabolas Understanding of Science 1: 103-128.
and ballistics. In biology, new bases can be found for the Johnson D, Levy F, Karsai I, Stroud K. 2006. Turning the potential liability
same concepts, such as muscle tension or proprioception of large enrollment laboratory science courses into an asset. Journal of
in locomotion (Brent 2004). We hope that new areas at the College Science Teaching 35: 46-51.
Johnson M, Gilliam T, Karsai I. 2009. The effect of infectiousness, duration
crossroads of biology and math will soon develop to a level of sickness and chance of recovery on a population: A simulation study.
where biologists can use them intuitively and they can also BIOS 80: 99-104.
be incorporated into education. To make it useful for sci- Jungck JR. 1997. Ten equations that changed biology: Mathematics in
ence and the scientist, an integrated or infused curriculum problem-solving biology curricula. Bioscene 23: 11-36.
should be based on a solid understanding of the scientific Karsai I, Knisley J. 2009. The role of institutes in interdisciplinary research
and education: An example from quantitative biology. Journal of Col-
method. We envision that successful programs in research
lege Science Teaching 38: 32-37.
and education in the United States and worldwide will Kauffman S. 1993. The Origins of Order: Self-organization and Selection in
promulgate the scientific method to stress that biology is Evolution. Oxford University Press.
science, and at the same time stimulate the use of common Keeling MJ, Rohani P. 2007. Modeling Infectious Diseases in Humans and
simulation platforms to enhance collaboration between Animals. Princeton University Press.
mathematicians and biologists in their efforts to generate, Kerfeld C, Simons R. 2007. The undergraduate genomics research initiative.
PLoS Biology 5: el41. doi:10.1371/journal.pbio.0040167
test, and analyze alternative hypotheses—and to educate a Kuipers B. 1994. Qualitative Reasoning: Modeling and Siinulation with
skilled workforce. Incomplete Knowledge. MIT Press.

www. biosciencemag.org September 2010 / Vol. 60 No. 8 • BioScience 637

Education

Lombrozo T, Shtulman A, Weisberg M. 2006. Tbe intelligent design con- Roth WM. 1995. Authentic School Science: Knowing and Learning in Open
troversy: Lessons from psychology and education. Trends in Cognitive Inquiry Science Laboratories. Kluwer Academic.
Science 10: 56-57. Salmon MH. 1989. Introduction to Logic and Critical Thinking. Harcourt
Maret TJ, Ziemba RE. 1997. Statistics and hypothesis testing in biology. Brace College.
Journal of College Science Teaching 26: 283-285. Sauro HM, Hucka M, Finney A, Wellock C, Bolouri H, Doyle J, Kitano H.
May RM. 2004. Uses and abuses of mathematics in biology. Science 303: 2003. Next Generation Simulation Tools: The Systems Biology Work-
• 790-793. bench and BioSPICE Integration. OMICS A Journal of Integrative
McComas WF. 1998. The Nature of Science in Science Education: Ratio- Biology 7: 355-372.
nales and Strategies. Kluwer Academic. Stukus P, Lennox JE. 1995. Use of an investigative semester-length labora-
Mitroff II, Mason RO. 1974. On evaluating the scientific contribution tory project in an introductory microbiology course. Journal of College
of the Apollo moon missions via information theory: A study of the Science Teaching 25: 135-139.
scientist-scientist relationship. Management Science: Applications 20: Thompson D. 1917. On Growth and Form. Cambridge University Press.
1501-1531. Thornton J. 1972. The Laboratory: A Place to Investigate. Commission on
Mrosovsky N. 2006. The marginalization of methods. BioScience 56: Undergraduate Education in the Biological Sciences.
790-791. West CB, Brown JH. 2004. Life's universal scaling laws. Physics Today 57:
Murray JD. 2002. Mathematical Biology, vols. 1 and 2. Springer. 36-42.
Musante S. 2005. Learning the nature of science. BioScience 55: 833.
———. 2009. A dynamic alternative to scientific method. BioScience 59: 15.
[NAP] National Academies Press. 2003. BIO2010: Transforming Under- Istvan Karsai (karsai@etsu.edu) is an associate professor in the Department of
graduate Education for Future Research Biologists. NAP. Biological Sciences, East Tennessee State University, in Johnson City. He studies
[NRC] National Research Council. 1996. National Science Education Stan- self-organization mechanisms of biological systems. George Kampis (gkampis@
dards. National Academy Press. colbud.hu) is professor and chair of Philosophy of Science, at Eötvös Loránd
Norton JD. 1998. How Science Works. McGraw-Hill. University, and fellow of Collegium Budapest, Institute for Advanced Study, in
Ortez RA. 1994. Investigative research in nonmajor freshman biology Budapest, Hungary. He studies evolutionary modeling, complex systems, and
classes. Journal of College Science Teaching 23: 296-300. various methodological and philosophical questions.

Cenozoic Mammals Protecting Mapping the Mind

of Africa Life on Earth RITA CARTER
Revised and Updated Edition
LARS WERDELIN, and An Introduction to the Science
"One of the clearest and best-illustrated
WILLIAM JOSEPH SANDERS, Editors of Conservation
attempts to explain the virtually
"A truly remarkable volume which, in its MICHAEL P. MARCHEHI and
inaccessible—the human brain."
scope, freshness, and value, provides far PETER B. MOYLE
—Sunday Times, London
more than anyone could have hoped for. "I have never seen a more comprehensive "Within the hardware of the brain there
Cenozoic Mammals of Africa raises the and succinct presentation of conservation lurks a software called the mind. Carter
bar in 'palaeo-publishing,' but particularly biology-the most important topic of our has brilliantly conveyed the thrill of
so, because it will be of use and in use by time. It bridges the gap between science discovering what this software can do."
scholars for years to come." and people with easy to read, non-technical
—Matt RIdiey, autbor of Genome:
—Richard Leakey, Stony Brook University prose. Conservation is everybody's busi- Tlie Autobiography of a Species In 23 Chapters
$95.00 cloth ness, and Protecting life on Earth is a must Finalist, Rhône-Pouienc Prize for Science Books
read for everyone who cares about the $29.95 paper
future of the planet."
—Mike Dombeck, University of Wiscpnsfn-
Stévëns Point Universily, System Feiiow of
Giobai Conservation, former Ciiief,
U.S. Forest Service
$49.95 paper
At bookstores or wvvw.ucpress.edu

U N I V E R S I T Y O F C A L I F O R N I A P R E S S

638 BioScience • September2010/Vol. 60No. 8 www.biosciencemag.org

Copyright of Bioscience is the property of American Institute of Biological Sciences and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.