chemistry education research and...

This journal is c The Royal Society of Chemistry 2013 Chem. Educ. Res. Pract.

Cite this: DOI: 10.1039/c3rp00111c

Rethinking chemistry: a learning progression onchemical thinking

Hannah Seviana and Vicente Talanquerb

Dominant educational approaches in chemistry focus on the learning of somewhat isolated concepts

and ideas about chemical substances and reactions. Reform efforts often seek to engage students in the

generation of knowledge through the investigation of chemical phenomena, with emphasis on the

development and application of models to build causal explanations and predict outcomes. However,

chemistry has been characterized as a technoscience that blends scientific pursuit and technological goals.

Besides searching for explanations, our discipline also involves the design of substances and processes to

address relevant problems, as well as the evaluation of social, economic, and environmental benefits,

costs, and risks associated with chemical knowledge and products. In order to develop authentic curricula,

instruction, and assessments that are better aligned with the core goals and practices of chemistry, we

need to understand how students’ chemical thinking progresses over time. We define chemical thinking

as the development and application of chemical knowledge and practices with the main intent of

analyzing, synthesizing, and transforming matter for practical purposes. In this paper we present a

blueprint of a theoretically sound and evidence-based foundation for an educational framework centered

on the idea of chemical thinking. Our investigations are focused on the development of a learning

progression that describes likely pathways in the evolution of students’ chemical thinking with training in

the discipline from grade 8 (age 13–14) through 16 (undergraduate completion).

Introduction

Dominant approaches to the teaching of chemistry in manycountries tend to present the discipline as a collection of somewhatisolated topics: atomic structure, chemical reactions, chemicalbonding, thermodynamics, kinetics, etc. Chemical ideas andpractices are also often introduced devoid of a clear practicalpurpose, beyond that of explaining the properties of matter(Van Berkel et al., 2000; Eilks et al., 2013). This approach toteaching chemistry is disconnected from the major purposes ofthe chemical enterprise; it fails to engage students in learninghow to pose and answer questions that reflect authenticchemical concerns: how do we identify, avoid, or mitigate theeffects of pollutants in our environment? How do we synthesizenew medicines? The topic-centered conceptualization of schoolchemistry stands in stark contrast to current visions for scienceeducation in the US (NRC, 2011), as well as in other countries(Waddington et al., 2007; Osborne and Dillon, 2008), in whichstudents are expected to ‘‘. . . over multiple years of school, actively

engage in science and engineering practices and apply cross-cutting concepts to deepen their understanding of each field’sdisciplinary core ideas’’ (NRC, 2011, p. 2).

A central challenge in science education is the developmentof coherent educational models that integrate central ideas andpractices within and across disciplines. One possible approachto facing this challenge is through the development of curriculaand instructional models that actively and meaningfully engagestudents in authentic domain-specific practices, meaning thespecific ways of reasoning, doing, and valuing that characterizeeach scientific domain (Bulte et al., 2006; Talanquer, 2013a). Inthe case of chemistry, those authentic practices certainly involvethe investigation of chemical substances and phenomena in thesearch for explanations for their properties and behaviors. How-ever, chemistry has been characterized as a technoscience thatblends the pursuit of scientific knowledge with technologicalgoals driven by human needs and conditions (Bensaude-Vincentand Simon, 2008; Chamizo, 2013). As such, chemical scientistsalso engage in practices that are akin to those of engineers,involving the design, application, and evaluation of methodsand strategies to analyze, synthesize, and transform chemicalsubstances (NRC, 2003).

The concept of chemical thinking can be used to capture theknowledge, reasoning, and practices that characterize the chemical

a Department of Chemistry, University of Massachusetts Boston, Boston, MA 02125,

USA. E-mail: [email protected] Department of Chemistry and Biochemistry, University of Arizona, Tucson, Arizona

85721, USA. E-mail: [email protected]

Received 25th September 2013,Accepted 26th September 2013

DOI: 10.1039/c3rp00111c

www.rsc.org/cerp

Chemistry EducationResearch and Practice

PERSPECTIVE

Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33

.

View Article OnlineView Journal

http://dx.doi.org/10.1039/c3rp00111c

http://pubs.rsc.org/en/journals/journal/RP

Chem. Educ. Res. Pract. This journal is c The Royal Society of Chemistry 2013

enterprise (Talanquer and Pollard, 2010). We define chemicalthinking as the development and application of chemicalknowledge and practices with the main intent of analyzing,synthesizing, and transforming matter for practical purposes.The core ideas and theoretical and experimental practices usedto identify substances, synthesize new materials, and transformmatter uniquely define chemistry as a discipline. They havealso had a profound impact on modern societies. They havetransformed the way we live and have helped increase the quality ofhuman life. They have also opened the door to serious environ-mental and social problems. Understanding the ideas and practiceson which chemical thinking is based is therefore critical forindividuals who will rely on their chemical understanding to makeimportant decisions in the practice of their profession. Thefoundations of this understanding are also critical for scientificallyliterate individuals who will not necessarily pursue career path-ways that involve the discipline of chemistry. Throughout theirlives, these individuals will have to make important decisionsthat rely on fundamental chemical thinking to answer questionssuch as how to dispose appropriately of material waste, whatcontainers are safe for storing different foods, and which fuelsimpart the least damage to the planet.

Common teaching approaches in chemistry tend to focus onengaging students in the description of the structure of chemicalsubstances and the characteristics of chemical processes (Van Berkelet al., 2000). Recent educational reform efforts seek to shift teachingpractices to actively involve students in authentic investigationsthat create opportunities for modeling and sense-making (NRC,2011). From our perspective, some of these efforts tend tooveremphasize the investigative and explanatory aspects ofchemistry at the expense of other core practices (Talanquer,2013a). A more authentic balance needs to be achieved by alsoopening spaces for students to engage in the design of materi-als or processes, and in the evaluation of the benefits, costs andrisks associated with the use and transformation of chemicalentities. Integrating these other aspects of chemistry wouldhelp us foster students’ abilities to make informed decisions,build justifications, evaluate outcomes, and test ideas in thecontext of personal, societal, and global concerns.

In recent years, chemistry educators across the world haveattempted to transform chemistry education based on the ideaof context-based learning (Bulte et al., 2006; Schwartz, 2006;Eilks et al., 2013). The associated curricula use everyday contexts andapplications of chemistry as starting points for developing studentunderstanding. Several of these approaches engage students inauthentic practices and seek to connect key ideas and skills inmeaningful ways (King, 2012). Nevertheless, in some cases corechemistry ideas and practices get buried under the specificcontexts used to motivate and drive learning in the classroom.When this happens, learning about the particular problemunder analysis (e.g., global warming) and the specific chemistryconcepts related to it (e.g., properties of greenhouse gases)become the central focus of attention, rather than the analysis,discussion, and reflection of the core chemistry ideas, ways ofthinking, and experimental strategies that can be applied tounderstand and solve such types of problems.

Given the above ideas and concerns, our work has focusedon building a theoretically sound and evidence-based foundation fora coherent educational framework centered on the idea of chemicalthinking, which can guide the transformation of curriculum, instruc-tion and assessment in chemistry. Our work is being informedby careful analyses of the nature of our discipline (Talanquer,2013a), the relevance of chemical thinking to modern citizenry(Sjostrom, 2013), and by research on how students learnchemistry over time. In this paper we present a blueprint ofthis foundation, together with a description of our researchefforts to actually build it. Our investigations are focused on thedevelopment of a learning progression (LP) that describes likelypathways in the evolution of students’ chemical thinking withtraining in the discipline from grade 8 (age 13–14) through 16(undergraduate completion).

Theoretical frameworkLearning progressions

In the past few years there has been a surge of interest in thedevelopment of learning progressions (LPs) of central ideas inthe sciences that can serve as curriculum models and assess-ment frameworks in educational settings. These LPs describesuccessively more sophisticated ways of thinking about a topic(NRC, 2007; Corcoran et al., 2009) and are based on educationalresearch about how people learn, existing pedagogical contentknowledge in the area of interest, as well as on the critical analysisof the structure of the associated disciplinary knowledge. To date,educational researchers have developed LPs in science for diversetopics such as atomic–molecular structure (Smith et al., 2006;Stevens et al., 2010), force and motion (Alonzo and Steedle, 2009),genetics (Duncan et al., 2009), the theory of evolution (Lehrer andSchauble, 2012), scientific argumentation (Berland and McNeill,2010), and energy (Lacy et al., 2012). However, there is still ampledebate on issues such as what constitutes progress in a given area,how more sophisticated ways of thinking are characterized, andwhether progress can be described as a series of successive levelsof understanding (Sikorski and Hammer, 2010). Debate also existson whether intermediate levels in a progression should onlyinclude simplified ideas that are scientifically sound, or whetherdescribing scientifically inaccurate intermediate understandingsmay also be productive (Duncan and Rivet, 2013).

The promise of LPs lies in the potential to guide thecoordination of teaching, instructional resources, and assessmentwith cognitive and metacognitive practices so that learningbuilds coherently. However, the field has not yet come toconsensus on a more precise definition of what an LP is.Recently, Duschl and collaborators (Duschl et al., 2011) con-ducted a comprehensive review of LP research in scienceeducation across the US and Europe over the past decades.These authors reviewed critical foundational domains that havecontributed to the development of LP research, including workin the areas of conceptual change, didaktik and teachingexperiments in the European tradition of education (e.g., Modelof Educational Reconstruction), and learning trajectories inmathematics education.

Perspective Chemistry Education Research and Practice

Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33

. View Article Online



In their review, Duschl and collaborators focused on describinghow LPs are being created, and how they are being validated anddescribed. They isolated four major aspects in which existing LPstend to vary: (1) their focus, as some LPs center on scientificknowledge without integrating science practices, while others focuson science practices without integrating domain knowledge; (2) thenature of their boundaries, as the definition of lower and upperlevels (also called lower and upper anchors, respectively) variesacross LPs; (3) the characteristics of intermediate states, as themanner in which different levels of understanding are studied,described, and related to instruction ranges from linear sequencesof steps to complex networks of ideas, with some LPs mostlyindependent of instruction and others tightly linked to it, and(4) the associated models of conceptual change, as some LPsare based on a misconception-based ‘fix it’ view, while otherstake an evolutionary ‘work with it’ perspective.

Although the current number of investigations on LPs aboutdifferent topics is relatively small, there are important studiesdirectly or indirectly focused on core ideas in chemistry. Such isthe case with the LPs associated with atomic–molecular structure(Smith et al., 2006; Stevens et al., 2010), properties of matter (Smithet al., 1985, 2010; Liu and Lesniak, 2005), the concept ofsubstance (Johnson and Tymms, 2011), and carbon cycling(Mohan et al., 2009). Many of the findings described in thesestudies provide foundational ideas for initial hypotheses for ourown chemical thinking learning progression (CTLP). Assessmentwork, such as that of Claesgens et al. (2009) on changes in studentreasoning about matter, energy, and change, also provides insightsinto how chemistry students’ ideas and ways of reasoning mayevolve over time. Before we describe our approach to the develop-ment and validation of the CTLP, which is a blended researchpursuit and educational development effort, it is important tolay out the fundamental theoretical commitments that form thefoundation under this work.

Theoretical commitments

Commitment 1: focus on disciplinary core practices and peda-gogies. Most chemical scientists working in either academic centersor industries engage in one or more of the following disciplinarycore practices: analysis, synthesis, and transformation. Analysisinvolves the development and application of strategies fordetecting, identifying, separating, and quantifying chemicalsubstances (Enke, 2001). Synthesis encompasses the design ofnew substances and synthetic routes (Hoffmann, 1993). Finally,transformation focuses on controlling chemical processes fornon-synthetic purposes, such as harnessing chemical energy(NRC, 2003). Theoretical concepts and experimental proceduresin chemistry are developed and applied at the service of the coregoals of these different practices. From this perspective, learningchemistry should involve making sense of these concepts andprocedures in the contexts in which they are relevant, and for thepurposes for which they have been developed. To foster andsupport the development of such understandings, we need to gobeyond uncovering and tracking students’ conceptions aboutsubstances and chemical processes. We actually need to explorehow learners think about and apply these concepts in the context

of analyzing substances, designing strategies to synthesize themor control their changes, or evaluating their effects on differentsystems. This type of exploration is at the core of our work in thedevelopment of the CTLP. We expect this progression to support andfacilitate a shift in chemistry education to emphasize the develop-ment of chemical thinking through pedagogical approaches thatengage students in authentic activities.

Various chemists and philosophers of science have arguedthat chemical scientists engage in well-defined approaches topracticing chemistry. Breslow defines the work of chemists asconsisting of two different types of activity: ‘‘investigat[ing]the natural world and try[ing] to understand it, while otherchemists create new substances and new ways to perform chemicalchanges that do not occur in nature’’ (Breslow, 1997, p. 2). Othersemphasize that, beyond increasing knowledge, chemists areobligated to engage in evaluative activity as well (Vilches andGil-Perez, 2013). In our view, the teaching of chemistry shouldrely on three major pedagogical approaches that reflect coreactivities in our discipline: investigation, design, and evaluation.Investigation refers to inquiry practices focused on the description,understanding, and prediction of chemical properties andphenomena. These types of activities involve the design andimplementation of investigations to answer relevant questionsabout systems of interests, as well as the development andapplication of models to make sense of observations, buildexplanations, or predict outcomes. Major curricular and reformefforts in both school (K-12) and university chemistry have beenlargely based on this type of pedagogical approach. For example,the science process skills movement in the 1980s (Padilla et al.,1983; Padilla, 1986), the inquiry emphasis in the 1990s and2000s (NRC, 1996), the 5E and 7E instructional design models(Bybee, 1997; Eisenkraft, 2003; Bybee et al., 2006), and mostrecently, the incorporation of science practices in the frameworkfor K-12 Science Education in the US (NRC, 2011).

The process of design in chemistry involves the creation andimplementation of strategies for analyzing, synthesizing,or transforming substances to address relevant problems.Engineering design (French, 1999) and technology design (Dugger,2001) have some overlap with how chemists engage in these typesof processes. Currently, a design-based pedagogy is largely absentfrom school chemistry in the US, but it becomes increasinglyimportant as students advance in their chemistry training at theuniversity level. In an educational setting, design activities shouldengage students in brainstorming ideas, identifying and analyzingconstraints, weighing the benefits and tradeoffs of alternativeplans, making decisions, testing ideas, and evaluating outcomes.Many of these aspects of engineering design have been includedin the Framework for K-12 Science Education (NRC, 2011), andincorporated into the Next Generation Science Standards (NRC,2013). Thus, as new standards become adopted in the US, theremay be an increasing emphasis on a design-based pedagogy inschool chemistry.

The third type of authentic activities, labeled evaluation, isconcerned with considering, weighing, and judging the social,economic, and environmental benefits, costs, and risks of chemicalproducts and activities. The development and use of chemical

Chemistry Education Research and Practice Perspective

Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




substances and processes demand the consideration of moraland ethical issues. In an educational setting, evaluation activitiesshould thus engage students in analyzing complex and controversialproblems or dilemmas, connecting ideas, building arguments, andmaking decisions while taking into consideration social, economic,political, and ethical factors. An evaluation-based pedagogy is alsolargely absent from school science in the US, though is incorporatedin such efforts as Science, Technology, and Society initiatives(Bybee, 1987; Solomon, 1996; Eilks et al., 2013), and Socio-Scientific Issues perspectives (Zeidler et al., 2005). Its presencein university chemistry education ranges from absent to consideredto some extent, depending on factors such as geography, valuesand research areas of faculty, mission of the chemistry departmentor higher education institution, and national political interests.For example, some chemistry departments include expertise ingreen chemistry, which emphasizes this pedagogical approachto chemistry research.

The development and validation of the CTLP described inthis paper is ultimately concerned with characterizing learningpathways through which students’ ideas and ways of thinkingabout synthesis, analysis, and transformation (i.e., their chemicalthinking) develop through formal education, in the context ofpedagogical approaches that involve investigation, design, andevaluation. This approach to the conceptualization of chemistryeducation shares common features with models of competenceused to develop chemistry education standards in some Europeancountries (Waddington et al., 2007). For example, ‘‘investigation’’and ‘‘evaluation’’ are areas of expected competence in Germanscience education standards (Schecker and Parchmann, 2007).

Commitment 2: focus on crosscutting disciplinary concepts.Work in chemistry LPs, as well as related research on students’alternative conceptions, has traditionally focused on exploringand tracking student understanding of fundamental concepts suchas atomic structure and chemical bonding. In the development ofour CTLP, we follow an alternative approach by analyzingthe evolution of students’ conceptions about crosscutting dis-ciplinary concepts judged to be critical in the understandingand practice of analysis, synthesis, and transformationin chemistry. Such crosscutting concepts include: chemicalidentity, structure–property relationships, chemical causality,chemical mechanism, chemical control, and benefits–costs–risks.As shown in Table 1, each of these concepts is linked to a corequestion driving chemical thinking. These crosscutting concepts areproposed as lenses through which to analyze students’ conceptualunderstanding of core elements of chemistry knowledge (e.g.,chemical bonding, atomic structure). Our goal is to characterizehow students think with such key elements as they reason about,for example, how to induce and control a chemical process(chemical control), or how to evaluate its potential impact on theenvironment (benefits–costs–risks). A core claim guiding our workis that the above crosscutting concepts should be used to analyzestudents’ understanding of chemistry as learners engage in corechemistry practices using different core disciplinary pedagogies.For example, we could explore how students use chemical bondingideas to make claims about structure–property relationships inthe context of designing a strategy to identify an unknown

chemical substance. As described in the following paragraphs,existing research suggests that the understanding of the selectedcrosscutting disciplinary concepts undergoes vast changes withtraining in the discipline. This research has allowed us to buildinitial hypotheses about lower levels (or lower anchors) andinitial steps in our CTLP.

Chemical identity. The development of analytical techniques toidentify and detect chemical substances in our surroundings, aswell as strategies to synthesize chemical compounds, is based onthe central assumption that each chemical substance has a differ-entiating property that makes it unique (Enke, 2001). However,understanding which properties may be used as differentiatingcharacteristics for chemical substances is not straightforward.Relevant research suggests that novice learners’ ideas anddecisions about identity and category membership are constrainedby the surface features and appearances of the systems of interest(Vosniadou and Ortony, 1989; Talanquer, 2009). Additionally,students tend to identify a sample of material by its history (originand what has happened to it) rather than physical and chemicalproperties (Johnson, 2000). Also, they often fail to understandthe difference between substance and object (Au, 1994), andbetween extensive (e.g., mass) and intensive (e.g., density)properties (Wiser and Smith, 2008).

Structure–property relationships. Making decisions about whattype of substance to synthesize, what reactant to use, or whatdetection technique to utilize critically depends on the under-standing of the relationship between properties and structure atdifferent scales. Many physical and chemical properties emergefrom the dynamic movement and interactions between the myriadsof particles that make up a macroscopic sample of a substance(Chi, 2005; Levy and Wilensky, 2009). However, research indicatesthat novice learners tend to rely on an additive versus an emergentframework in the prediction of the physical and chemical proper-ties of substances (Talanquer, 2008). Students implicitly presup-pose that substances directly inherit properties from theirindividual submicroscopic components (Talanquer, 2006; Taberand Garcıa-Franco, 2010). In general, students have difficultiesreasoning about systems and processes at multiple scales, focusingon surface features of a system to build predictions and makedecisions (Gilbert and Treagust, 2009; Cooper et al., 2012, 2013).

Chemical causality. The atoms, ions, or molecules that makeup a chemical substance adopt stable dynamic structures

Table 1 Crosscutting disciplinary concepts targeted by the CTLP and associatedcore questions to which each of them responds

Crosscuttingdisciplinary concept Core question

Chemical identity How do we identify chemical substances?Structure–propertyrelationships

How do we predict the properties of materials?

Chemical causality Why do chemical processes occur?Chemical mechanism How do chemical processes occur?Chemical control How can we control chemical processes?Benefits–costs–risks How do we evaluate the impacts of chemically

transforming matter?


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




determined by the nature of random interactions betweenparticles in the system. Stable structures may undergo trans-formations via random interactions with other particles. Thelikelihood of such transformations depends on both the stability ofthe new structures and the probability of successful interactionsbetween particles in the system (Atkins and de Paula, 2006).Educational research suggests that novice learners rarely adopt acausal framework in which structures and events result fromdynamic random interactions among multiple components. Onthe contrary, many students assume that phenomena areinduced by leading agents acting on passive entities (Andersson,1986; Grotzer, 2003), or consider that processes are driven by anagent’s desire or purpose (Taber, 2009, 2013; Talanquer, 2010).These ways of thinking constrain students’ ability to analyzechemical phenomena which frequently involve simultaneousrandom processes that occur with different probabilities.

Chemical mechanism. The prediction and control of the out-come of a chemical process depends on the understanding of thesequence of events triggered by the combination of chemicalreactants. The identification of reaction mechanisms relies ontheoretical models about the structure of matter that can be usedto explain and predict properties and behaviors (Carroll, 1998).Relevant educational research indicates that many novice learnersthink of chemical reactions as mixing or relocation processes,with no clear sense of what happens during a chemical change(Andersson, 1990; Hesse and Anderson, 1992). In general, theyhave a tendency to describe the evolution of a chemical process asa linear chain of events (story or chronology). Each variable isconsidered one at a time, assigning a preferred direction to theprocess (Talanquer, 2006). Sequence of events in chemicalreactions are built based on surface structural features ratherthan on the analysis of chemical properties and interactions(Bhattacharyya and Bodner, 2005).

Chemical control. The design of successful strategies toidentify or synthesize a chemical substance depends on theunderstanding of internal and external factors that affect thestability and reactivity of chemical substances. It is also critical tounderstand how such factors affect the rate of chemical processesand the extent to which these processes can go to completion(NRC, 2003). Research in this area suggests that many novicelearners implicitly assume that chemical processes always need tobe initiated by active agents and that, once started, they always goto completion (Johnson, 2000, 2002; Taber and Garcıa-Franco,2010). In general, students struggle to differentiate those factorsthat affect the rate of a chemical process versus those thatdetermine the final state of equilibrium (Gilbert et al., 2002). Ingeneral, one can expect novice learners to conceive chemicalcontrol as something achieved by modifying external parameters,such as temperature, neglecting internal factors such as thestructure of the particles involved (Kind, 2004).

Benefits–costs–risks. The analysis, synthesis, and transformationof chemical substances have many benefits for modern societies.However, there are also social, political, economic, and environ-mental costs and risks that need to be taken into account when

making decisions involving the application of chemical design(Bensaude-Vincent and Simon, 2008). Research in the area ofargumentation of socio-scientific issues suggest that individualstend to selectively credit or dismiss evidence of benefits, costs, andrisks based on personal values that they share with others ratherthan on scientific knowledge (Kahan et al., 2011). In the context ofscience education, science learners have been found to rely onemotive, intuitive, and rationalistic resources when analyzing socio-scientific issues, independently of their level of content knowledgeabout a subject (Sadler and Donnelly, 2006). Changes in students’ability to generate high-quality costs–benefits and risk analysesseem to vary in a non-linear fashion with content knowledgeacquisition (Sadler and Fowler, 2006).

Commitment 3: focus on mapping the cognitive landscape.We conceive the development of understanding of chemicalthinking as occurring in a multi-dimensional space, in whichsets of ideas become integrated in different ways as studentsdevelop more sophisticated understandings. The developmentof the CTLP requires the characterization of this complex dynamicknowledge space as represented by the analogy depicted in Fig. 1.The variables of the complex system represented in this figure(two variables shown) correspond to the progress variablesof the LP. Wilson (2009) describes progress variables as dimen-sions of student knowledge or competency along which pro-gress is expected to occur. Student understanding of a topic canbe thought of as a constrained dynamic system in constantinteraction with its environment (Brown and Hammer, 2008).As a result of these dynamic interactions, structures andpatterns emerge representing conceptualizations or ways ofthinking. These more or less stable cognitive structures maybe conceived as dynamic attractors or semi-stable states, whichmay represent radically different accounts of a system orphenomena. We see the construction of an LP as a process ofidentifying and characterizing both the evolutionary path ofsuch states from naıve to sophisticated ways of thinking, as wellas the internal constraints and external conditions (e.g., instruc-tion) that support such evolution. Our approach to developingthe CTLP follows the example of Wiser’s and Smith’s work

Fig. 1 Theoretical potential energy surface of a complex system as an analogicalmodel of the CTLP knowledge space. Variables (two shown) are progressvariables in the LP. The z-axis (E) represents the stability of cognitive constructsand ways of reasoning, with more stable constructs having a lower E. Wells in thesurface correspond to ‘‘cognitive attractors’’ or regions of relative stability thatmay be stepping stones in the learning progression.


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




(Smith et al., 2010; Wiser et al., 2013) in looking for inter-mediate attractors or states that are stepping stones in students’learning.

Stepping stones represent productive ways of thinking thatmay support important reconceptualizations with properinstruction. For example, the analysis of children’s ideas aboutmaterial kinds (chemical identity) reveals that at young agesmaterials are thought of as objects that can be categorized intodifferent classes based on perceptual features (Wiser and Smith,2008). It is not until children start thinking of materials as basic‘‘constituents’’ of matter that they can begin to develop anunderstanding of the concept of substance (Johnson, 2000). Thus,conceptualizing materials as fundamental constituents ratherthan as simple labels used to describe different types of ‘‘stuff’’can be seen as an stepping stone in the path to meaningfullyunderstand what a chemical substance is and what cues aremost relevant in assigning chemical identity.

The progress variables in our CTLP have been selected totrack students’ conceptual sophistication and modes of reason-ing about the six crosscutting disciplinary concepts describedin the previous section. Conceptual sophistication is determinedby the nature of students’ underlying assumptions about thestructure and properties of chemical entities and phenomena.Modes of reasoning refer to the complexity of student reasoningin terms of their ability to connect ideas, build justifications,make decisions, and construct sophisticated explanations. Tofacilitate the characterization of assumptions and modes ofreasoning we have identified a set of essential questions thatdefine the critical aspects of students’ chemical thinking thatwe propose to explore (see Fig. 2). These essential questions arefundamental queries in the work of chemical scientists as theyanalyze materials, synthesize new substances, and transformmatter. The answers given to these questions can be expected todepend on people’s conceptual sophistication and modes ofreasoning as related to the six crosscutting disciplinary conceptsrepresented in Fig. 2. These essential questions define theprogress variables along which chemical thinking is hypothe-sized to develop with training in the discipline.

Commitment 4: focus on ‘‘assumptions’’ to assess conceptualsophistication. Researchers have explored and tracked different

aspects of students’ understanding in the development of LPs.For example, some of them have focused on the evolutionof students’ mental models of fundamental systems or pheno-mena (Stevens et al., 2010). Others have paid attention to thenature of the explanatory accounts built by students at differentlevels in a given progression (Mohan et al., 2009). In our case,changes in conceptual sophistication are described in terms ofthe evolution of underlying assumptions that support, but alsoconstrain student reasoning about the relevant entities andprocesses in a domain (Talanquer, 2006, 2009, 2013b; Maeyerand Talanquer, 2013). A focus on ‘‘assumptions’’ facilitatesthe identification of cognitive resources that may supportproductive thinking at different educational levels (Taber andGarcıa-Franco, 2010). This approach also recognizes the dynamic,complex, and contextual nature of students’ ideas (Sevian andStains, 2013; Stains and Sevian, 2013).

A variety of researchers interested in conceptual change haveidentified diverse cognitive elements that seem to guide, support,but also constrain students’ reasoning in different domains.They have referred to these cognitive elements in differentways, such as implicit presuppositions (Vosniadou, 1994),phenomenological primitives (diSessa, 1993), ontologicalbeliefs (Chi, 2008), conceptual resources (Redish, 2004), andcore knowledge (Spelke and Kinzler, 2007). Nevertheless, thereis debate on the extent to which these cognitive elements arebetter described as coherent integrated knowledge systemsversus fragmented collections of cognitive resources (Brownand Hammer, 2008; diSessa, 2008; Vosniadou et al., 2008). Itis likely that their level of integration may vary depending onthe nature of the knowledge domain and the prior knowledgeand experiences of each individual.

Based on these ideas, our research has been based on atheoretical model that proposes that student reasoning in chemistryis often guided by implicit or explicit assumptions about the natureof chemical entities and processes (Talanquer, 2006, 2009; Maeyerand Talanquer, 2013). The nature of these cognitive elements maychange over time with development and learning; some of theseconstraints may lose or gain strength depending on existing knowl-edge, contextual features, and perceived salient cues and goals of atask. Nevertheless, these assumptions support the development andapplication of dynamic mental models of the systems of interest,and help us make decisions about what behaviors are possible ornot and about what variables are most relevant in determiningbehavior. They also support the development or application ofreasoning strategies to make predictions about how the object willbehave when involved in different processes or events. Thus, thecharacterization of progress in understanding may be facilitated bymapping the landscape of assumptions that most commonly guidestudent reasoning when engaged in disciplinary tasks. Talanquer(2009) has illustrated this approach in the analysis of students’evolving ideas about the particulate model of matter. For example,many learning difficulties in this area seem to be associated withlearners’ tendency to assume that substances are homogeneousentities, and thus their properties are the same at all scales(e.g., atoms and molecules have the same density or meltingpoint as the macroscopic sample of which they are part).Fig. 2 Essential questions that define the progress variables in our CTLP.


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




Commitment 5: focus on modes of reasoning to assess thecomplexity of student thinking. Student understanding inchemistry cannot only be assessed based on the nature of thecontent knowledge that students demonstrate, or on the validityof the assumptions that they make about the structure, proper-ties, and behaviors of the system under analysis. It is alsoimportant to explore how students use the information availabletogether with their prior knowledge to make decisions, buildarguments, generate explanations, and make predictions.A variety of researchers have introduced different approachesto characterizing the level of sophistication and the complexityin student reasoning. For example, in the SOLO taxonomy definedby Biggs and Collis (1982), student responses are allocated to ahierarchy of stages (e.g., prestructural, unistructural, multi-structural) depending on the number and level of integrationof the elements considered. This taxonomy has been used asa reference by other authors to develop scales to measure,for example, quality and complexity of students’ scientificreasoning (Brown et al., 2010), performance levels in students’understanding of chemistry (Claesgens et al., 2009), and levelsof achievement in science domains (Bernholt and Parchmann,2011). In contrast with our approach, some of these studiesdefine different modes of reasoning by taking into account boththe conceptual sophistication and the complexity of studentthinking. Other scales have been proposed to differentiate howlearners use knowledge of different complexity in various contexts(von Aufschnaiter and von Aufschnaiter, 2003), and work on theuse of heuristics for decision making has helped us identifyspecific patterns of reasoning that characterize intuitive versusanalytical reasoning in chemistry (Maeyer and Talanquer, 2010,2013). Based on the analysis of existing work on assessment ofstudent reasoning, we have identified a set of modes of reasoning

that serve as initial hypotheses for our CTLP. The basic featuresof each of these modes of reasoning are summarized in Table 2.

Development and empirical validation

Our research work in the development and validation of the CTLPis guided by the following overarching research questions:

(1) What core assumptions and modes of reasoning can beexpected to characterize novice (lower anchor) and sophisticated(upper anchor) chemical thinking as related to each of the sixcrosscutting disciplinary concepts: chemical identity, structure–property relationships, chemical causality, mechanism, chemicalcontrol, and benefits–costs–risks?

(2) What core assumptions and modes of reasoning mayconstrain students’ thinking about each of the six crosscuttingdisciplinary concepts at intermediate learning stages?

(3) What ‘stepping stones’ characterize learning pathways inthe progression from novice to sophisticated chemical thinking?

To illustrate our approach to developing and validating theCTLP, we briefly outline examples of the application of ourframework to the analysis of data collected using an instrument(GoKart interview, see Table 3) designed to explore students’conceptual sophistication and modes of reasoning as related tochemical identity, structure–property relationships, and benefits–costs–risks. The GoKart interview protocol is being used byteachers in middle and high schools in an urban school districtin the Northeast, and by researchers at two universities in theUS and one in Costa Rica. The instrument was designed touncover students’ assumptions and modes of reasoning whenengaged in an evaluation task that requires the selection of themost appropriate fuel to use in the design of a GoKart. The taskthus demands students to apply chemical thinking in the areas

Table 2 Hypothesized modes of reasoning describing different levels of complexity in student reasoning

Mode Description

Descriptive Salient entities in a system are identified or recognized. Explicit properties are described, verbalized. Functions or propertiesof entities are seen as sufficient explanation for their behavior. A phenomenon is seen as an instantiation of reality; it may bere-described by merely asserting that things are as they are without referring to causes. Reasoning mostly based onexperiences and knowledge from daily life. Strong influence of surface similarity and recognition on judgment and decisionmaking.

Relational Salient entities in a system are identified or recognized. Explicit and implicit differentiating properties are highlighted.Spatial or temporal relations between entities are noticed. Correlations between properties and behaviors are established butnot explained or justified. A phenomenon is seen as an effect of a single entity, the natural outcome of a single property or thelinear combination of several properties; no mechanisms are proposed. Reduction of variables and overgeneralizationconstrain reasoning.

Linear causal Salient entities in a system are identified or recognized. Explicit and implicit differentiating properties are highlighted.Spatial or temporal organization of and connections between entities are noticed. Relevant direct interactions betweenentities invoked. Although the influence of many factors may be recognized, phenomena tend to be seen as (reduced to)the result of the actions of a single agent on other entities; proposed mechanisms involve linear cause–effect relations andsequential chains of events. Reduction of variables and overgeneralization frequently constrain reasoning.

Multicomponent Salient entities in a system are identified or recognized. Explicit and implicit differentiating properties are highlighted.Spatial or temporal organization of and connections between entities are noticed. Relevant interactions between entitiesare invoked. Effects of several variables are considered and weighed.

(a) Isolated (a) Complex phenomena are seen as the result of the static or dynamic interplay of more than one factor and the directinteractions of several components. Effects of several variables are considered and weighed separately.

(b) Integrated (b) Complex phenomena are seen as the result of the dynamic interplay of more than one factor and the direct and indirectinteractions of several components. Explanations as interconnected stories of how different variables affect the entitiesinvolved. The effects of different variables are more thoroughly and systematically explained for multiple entities involved.


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




of chemical analysis and transformation. In the context of thisproblem, participants are asked a series of questions that probetheir reasoning about the use of different fuels based oninformation about their physical properties, chemical composi-tion, and molecular structure. An important feature of theGoKart instrument is that there is no one right answer to thechoice of fuel. The final decision depends on the relative weightthat people assign to different factors when making a decision.Even experts do not agree on the best answer.

We have interviewed individuals ranging from students ingrade 8 (age 13) through graduate school (e.g., fifth-year PhDchemistry students), and chemistry experts in academia, industry,and government/regulatory work. The examples that are analyzedbelow are representative of the full data set, and illustratedifferent levels of conceptual sophistication and the applicationof various modes of reasoning along progress variable 2 in Fig. 2(What cues are used to differentiate matter types?).

Let us first consider the case of a tenth grader (code nameMaria) taking an introductory chemistry course in high schoolwho was interviewed by her teacher. As illustrated by the followingexcerpt, her reasoning was strongly influenced by the assumptionthat natural substances were somehow better for the environmentthan fuels she perceived as not coming from nature:

Interviewer: so if all the fuels cost the same per gallon, whichfuel would you choose to power your GoKart and why?

Maria: well, I wanna consider the environment and somethingthat’s not so harmful but at the same time it’s efficient so I don’t knowall of them but I’m gonna go probably the natural gas, that looks

Interviewer: and why?Maria: because well the name of it, first of all, is convincing,

and it’s, I feel like, I know gasoline is, can be harmful so,

and I don’t know so much about ethanol; I think methanecould work

Interviewer: ok alright, so why did you decide then, it’s thenatural gas, because it has word natural in it?

Maria: yeahThe analysis of the above excerpt also reveals that Maria’s

reasoning was largely based on the direct association betweennaturalness of the source and level of pollution, withoutreference to a causal mechanism to justify such link. This isan example of ‘‘relational’’ reasoning as described in Table 2.

Maria also assumed that gaseous materials could causemore pollution, as gases get emitted into the air. As shownbelow, this association led her to question her initial decision:

Maria: well now after I see that the rest are liquid and this isthe only one provided by gas, I’m thinking it’s a good choice butthen when I think about pollution and I’m just like, I don’t know,there’s just so many factors, I don’t know

Interviewer: so you’re sticking with the natural gasMaria: yeahInterviewer: and does the fact that it’s a gas versus the other

liquids have any impact on your choice?Maria: it does, it has both, has a positive and a negative

effect, has the negative effect because I’m thinking about pollution,I’m thinking about the air being polluted, I’m thinking aboutgas in the air and it’s like, it’s not a good thing, and then I’mthinking about how it’s gas and it’s less resources I’m usingfrom, so yeah

Interviewer: ok so I just want to make sure I understand whatyou’re saying. So because the natural is a gas it’s gonna affect theair more?

Maria: yeah it could possibly affect the air more

Table 3 GoKart interview protocol. Scenario: an amusement park has asked you to design a GoKart (a small vehicle with an engine that kids can ride in). During yourdesign phase, you must decide which fuel will power the GoKart. You are considering four fuels. First is gasoline, also known as octane, derived from petroleum.Second is also gasoline, but derived instead from wood pellets. Third, is natural gas, also known as methane. Finally, there is E85, which is mostly ethanol. (Photo of achild driving a GoKart is provided)

Interview question Question intent

1 Which fuel would you use? Why? (list of fuels andeach one’s main chemical component is provided)

� Generate mental model of the scenario� Determine immediately accessible prior knowledge about thefuels and bases for decision-making� Determine participant’s thinking on whether and how octane frompetroleum vs. from wood pellets differ

2 Gasoline and E85 are liquids, while natural gas isavailable as a gas. Is this important? Why?

� Determine how the participant considers state of matter� Determine participant’s thinking on how state of matter influencesfuel usage, reactivity, outcomes, and consequences of use

3 E85 contains carbon, hydrogen and oxygen, whilethe other two fuels contain only carbon and hydrogen.Is this important? Why?

� Determine how the participant considers composition of matter� Determine participant’s thinking about how composition influencesproperties, outcomes, and consequences of use

4 Are the molecular structures of the fuels important?Why? (ball-and-stick drawings, with element symbols added,are provided)

� Determine how the participant considers molecular structure� Determine participant’s thinking about how molecular size, shape, andbonding/connectivity influence properties, outcomes, and consequences of use

5 In terms of how the fuels affect the environment,is one fuel better than the others? Why?

� Determine what economic, environmental, social, political, ethical,and moral factors the participant views as important to considerin decision-making� Assess how participant evaluates benefits and costs associatedwith the use of different fuels


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




As the third excerpt included below illustrates, althoughMaria had been exposed to introductory ideas about molecularstructure and had learned enough basic organic nomenclatureto associate names with key structural features, her final decisiononly involved considerations based on her two intuitive assump-tions about the polluting nature of general types of materials (i.e.,natural vs. artificial and gases vs. liquids):

Interviewer: when fuels are used in engines they can causepollution, kind of like what you were saying, so in terms of howthese four fuels would affect the environment, which of the fuels doyou think would be better than others to use? as far as the amountof pollution

Maria: this is when I’m like stuck because I feel like natural gas,I don’t even know so much about it, I feel like it, I don’t know, I feellike it’s good but then when I saw that it’s a gas, it’s the only onethat’s not a liquid, I, it’s kind of tricky because I don’t want it toaffect so much pollution like I don’t think it will affect as much, asthe other ones, so I still I’m persistent, still sticking to my answer,I still choose natural gas because I feel like there’s a clue in natural,in the way it, natural in the name I’m thinking it’s used outof natural resources, things that are not harmful, so lot of goodstuff in there

Interviewer: okMaria: yeah and E85 is ethanol, and you told us that the, when

it has -ol at the end it’s an alcohol and octane I’m guessing, isn’tthat it’s something sweet in it? no, no, wait, no, the ending, I don’tremember the ending

Interviewer: right we talked about it in [the] Smells [unit] okInterviewer: is there anything else that we should know about

in making this decision that I haven’t asked you yet?Maria: I still stick to my choice and I feel like there has to be a

clue in the word naturalMaria is a prototypical example of a student who does

not express mechanistic reasoning in her justifications orexplanations, but rather relies on simple associations to makeher decisions (relational reasoning). Her thinking alsoillustrates the influence of intuitive assumptions about theproperties of chemical entities. In particular, psychologicalresearch indicates that humans have a preference for what isperceived to be natural over what is perceived to be artificial (Rozin,2005). Natural substances and processes are often linked to asubjective impression of goodness, while the products of humanintervention are frequently judged more negatively. Positive feelingsabout entities or individuals are known to bias human decision-making (affect heuristic; Slovic et al., 2003), particularly in theabsence of prior knowledge, or under conditions of limitedtime or low motivation.

In our second example, we present the case of a senior(fourth-year) university student (code name Cartesian), who wasmajoring in chemistry and had just recently completed a physicalchemistry course. Cartesian’s decisions were strongly influencedby the assumption that the best fuel would be the most availableand easy to store and handle (because some sources werebecoming exhausted, and because transportation and storageof the fuel could incur significant costs). His first choice wasthen ethanol, judged to be more abundant than the other fuels.

He then switched to methane based on the assumption thatgases were easier to store and transport. When asked aboutwhether and how the molecular structures of the fuels wererelevant to the choice of fuel, Cartesian expressed knowledge aboutstructure–property relationships, but his reasoning was constrainedby the search for the fuel that would be easier to store. His lineof reasoning is illustrated in the following excerpt:

Interviewer: Okay. So now we’re going to give you the structuresof these molecules. Octane, methane, and ethanol. Do you thinkthat knowing the structure is important and does that changeyour decision?

Cartesian: . . .(long pause). . . I have to take a step back. I’m tryingto be unpartisan because I know these are all sources. I took a biggerstep back, knowing the structure would be an important element.

Cartesian: . . .(long pause). . . I don’t know how in-depth I should go,because we know that longer chains, you can look at the changes inboiling point and melting point and those types of areas. And you can tryand maybe do an analysis of these will go at higher or lowertemperatures and how much energy we would need to put in and outfor the reactions to occur to be more cost efficient. But these are all prettysmall, pretty similar compounds: octane, methane, and ethanol.

Interviewer: So from the structure you can determine melting pointand boiling point and that would be important for combustion?

Cartesian: Yeah. You could think of it as kind of an. . . youcould do it as an efficiency test. You could. . . I mean, longer chainsyou have more van der Waals and those types of forces, becausethese are all pretty similar compounds. Carbon chains, carbon chains,you have an alcohol group but that’s not going to do very much.So I don’t think any differences like that, for these compounds,would. . .I again don’t think that would make too much of a difference.I think these are all pretty similar compounds for what we’re usingthem for. I mean, you could look at the sizes and say . . .(pause). . . wecould look at the compounds and see the sizes and see how they stackif it would be more efficient for storage purposes. But methanebeing a gas, we could probably compress it a whole lot, so it’s stillprobably the most efficient to store. So again, I don’t really thinkthe structures are too useful for this type of question.

Cartesian exhibited a higher level of conceptual sophisticationthan Maria, as he recognized relevant cues in making the targeteddecision. Nevertheless, his chemical thinking was narrow andfocused on physical properties of the materials. This was a commonpattern observed in many of our interviews, where attentionto common practical issues in the handling of substances (e.g.,availability, storage, transport, safety, regular use) dominatedover the comparative analysis of chemical properties. Cartesian’smode of reasoning was also more complex than Maria’s. Forexample, in the previous excerpt his argument is multicomponent-integrated. He paid attention to a variety of factors, such aschemical composition, molecular size, and strength of inter-molecular forces to build a causal mechanistic argument thatcould help him differentiate among the substances. However,the assumptions that guided his reasoning constrained hisability to pay attention to more relevant cues.

Our third and fourth cases illustrate the chemical thinkingapplied by two different experts in the discipline. The first ofthem (code name Aimake), was a professor with specialization


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




in materials science and organic synthesis. In contrast with theprevious two examples, his differentiation of matter types wasbased on a chemical argument. In particular, he considered that ifcomplete combustion occurred, fuels could be differentiated bythe oxidation state of carbon atoms in their molecules. If thecombustion was incomplete, the outcome would depend onvarious other factors and would have to be empirically deter-mined. The excerpt below illustrates his ideas:

Interviewer: Ok, last question. If your goal was to consider theenvironment, what would be best for the environment, which ofthese fuels would be best to use?

Aimake: That is a question that I would first make an assumption.Assume I have the most efficient engine I can get out there.

Interviewer: What do you mean by efficient?Aimake: That it does complete combustion. If that’s my

hypothetical scenario, then I would still go for octane. Becausefor every carbon I use, I’m going to get more energy. And for everycarbon, I get one CO2. If you use ethanol, which has less energy,I still get the CO2, but I get less energy. So assuming, to me, thechemically speaking, the byproduct is the same. You’re going to getCO2. So, if I’m going to get CO2, I’d better get the maximum out ofit. So it’s a cost-benefit analysis. Because for every carbon there is,I’m getting the same molecule, the same product. It doesn’t matter.But, if I’m working now with an inefficient engine, then I have toworry, oh, am I getting complete combustion? Or am I getting theseother highly active molecules, sub-. . . not fully oxidized carbonproducts that could be really detrimental to the environment. Thenin that case, when I have inefficient engine, then I have to considerthe actual products themselves, which right now I don’t know. Butin a case where I have a complete ideal, I would still go for octane.

The second professor (code name Caesar), with specialization incatalysis and medicinal chemistry, and who had worked in thepaint industry, relied on different cues in his decision. For thisindividual, the source of carbon in fuels mattered. He assumed thatrecycling carbon atoms was less damaging to the environmentthan pumping new carbon into the atmosphere, thereforeethanol from a biological source (fermentation of sugars) wasbetter than octane from petroleum, but ethanol from petroleum(via cracking) would not be better than octane from the samesource. The excerpt below illustrates his argument:

Interviewer: Our last question is concerned with pollution.If the goal is to reduce pollution, which would be the best fuel?

Caesar: No question, ethanol.Interviewer: Why?Caesar: Because you are recycling carbon atoms. You see the

cane sugar, or whatever was the source, the biological source, whereethanol comes. That is, ethanol is from a biological, a primary biologicalsource, then we will be just recycling. If we, you would say, in petroleumwe have ethane (writes C2H6). It can be cracked to ethylene (drawsarrow, then C2H4 + H2). And part of it could be hydrated to ethanol(draws arrow from C2H4 to EtOH, writes H2O beside arrow). Ifthis is the source of ethanol, then we’re just pumping CO2 into theatmosphere. But if ethanol comes from a biological source, thenwe are just recycling carbon atoms. . . So it has to do with the originof the ethanol. If it comes from a petroleum source, we will keepon pumping carbon into the atmosphere. But if it comes from a

biological source, then we’ll just be recycling carbon. And so if youassure me that the ethanol would come from sugars, from thefermentation of sugars, that would be okay.

Both Aimake and Caesar cued on chemical compositionissues in their analysis, but focused their attention on differentaspects of the problem. Aimake paid attention to the ratio ofenergy to CO2 produced based on the analysis of the oxidationstate of carbon atoms in each fuel, while Caesar cued onthe origin of such atoms. Both professors displayed highconceptual sophistication in their chemical thinking, althoughmost of their reasoning was expressed at a relational level.For example, initially Aimake simply associated the amount ofenergy produced per carbon atom to the oxidation state ofsuch atoms, without reference to any causal mechanism tojustify his argument. Similarly, Caesar claimed that recycledcarbons were better than new carbon atoms for the environ-ment, without further explanation. The assumptions that theseindividuals made led them to rely on relevant chemical cuesin making their decision, but their expressed reasoningwas highly associative, relying on knowledge that was notexplicitly stated. Nevertheless, both of experts were able toadopt a multicomponent-integrated mode of reasoning whenprompted.

The examples presented in this section demonstrate that inorder to characterize differences in people’s chemical thinkingit is important to pay attention to both the assumptions(whether more intuitive or more academic) that they rely upon,as well as the modes of reasoning that they employ. Theprevious excerpts illustrate how a novice student cued onfeatures that were not centrally relevant, using relational orlinear causal reasoning to justify her decisions. Experts appliedsimilar modes of reasoning, but they cued on critical factors,building valid associations that were justified by knowledgethat can be made explicit. The student at an intermediatestage relied on multicomponent reasoning, but struggled todiscriminate among relevant and irrelevant cues in makinghis decisions. The assumptions that he made were hybrids,involving intuitive ideas mixed with academic constructs. Char-acterizing pathways of progression may thus be challenginggiven the diversity of ideas that intermediate students oftenbring into their analyses.

As the GoKart instrument illustrates, we are exploringstudents’ chemical thinking in terms of their conceptualsophistication and modes of reasoning, using semi-open inves-tigation, design, or evaluation tasks that engage students in oneor more core practices (analysis, synthesis, transformation) andpose questions that allows us to elicit how students apply theirknowledge in relation to relevant crosscutting disciplinaryconcepts. As seen in the examples, participants cue on differentfactors, often in relation to different crosscutting concepts. Forexample, Maria cued on potential harm to the environment(benefits–costs–risks), while Caesar cued on the method andsource of production (chemical mechanism). Our approachthus demands the design and development of an assortmentof research instruments that allow us to map the complexlandscape of chemical thinking (NRC, 2001).


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




Final comments

Our work in the development and validation of the CTLPrepresents, for us, a means to rethink both chemistry educationand chemistry education research in a way that is moreauthentic to our discipline and more meaningful and relevantto students. Our efforts provide an alternative lens through whichto analyze and think about chemistry teaching and learning(Talanquer, 2013c). We are carrying out this work involving andseeking the input of critical stakeholders. First, the CTLP must betrue to the discipline of chemistry. Therefore, we are askingexpert chemists in academia, industry, and government andregulatory sectors to help us validate important components ofthe LP, such as the essential questions that define our progressvariables and the associated upper anchors. Second, we areworking with secondary school science teachers who are notonly collaborating with us in the design of research instrumentsand the collection research data, but are also helping us groundour development in the authenticity of real classrooms. Third,our studies involve students from a variety of educational levelsand cultural and economic backgrounds. Our goal is to explorehow chemical thinking develops within existing learning condi-tions, as well as how it could progress when considering thealternative pedagogies that we advocate. Finally, we recognizethat our educational and research pursuits exist in the context ofnational frameworks and standards, and thus our CTLP should beresponsive to them but also challenge the status quo (Talanquerand Sevian, 2013).

We expect that our efforts will result in tangible products tosupport and facilitate chemistry teaching and learning drivenby our CTLP. We are currently developing detailed descriptionsof the learning progression from novice (lower anchor) to sophisti-cated (upper anchor) chemical thinking along all six of the cross-cutting disciplinary concepts. Such descriptions include carefularticulation of productive intermediate understandings (steppingstones) that support the reconceptualization of chemistry ideas.These descriptions are characterized in terms of core assump-tions and modes of reasoning that have emerged from ouranalysis. We are also collaborating with teachers in our researchteam to create instruments that can be used both as researchtools and as formative assessments in the classroom. Our intentis to produce instruments that will give teachers relatively rapidand valuable feedback on the assumptions and modes of reason-ing that constrain their students’ thinking, and provide guidanceon how to scaffold student learning. These instruments aretaking a wide range of forms, including two-tiered tests, ques-tionnaires with open-ended questions, cognitive interviews, andteaching experiments with specific prompts.

We are strong advocates of educational and research effortsthat seek to create conditions for students to actively engagein relevant, meaningful, and productive ways of thinking inthe science classroom. Nevertheless, we believe that sucheducational approaches should emerge from a critical analysisof the goals, practices, and ways of thinking that characterizeeach scientific discipline. They should also be responsive to therole and impact that the products of such human enterprise

have at the personal, societal, and global levels. Equally important,they should be based on the careful investigation and analysisof how students’ assumptions and modes of reasoning progressover time under various conditions. The work described in thispaper seeks to contribute to such types of efforts.

Acknowledgements

The authors wish to acknowledge the funding sources, US NSFawards DRL-1221494 and DRL-1222624, and a AAAS Women’sInternational Research Collaborations at Minority-ServingInstitutions award, that support our work. We thank all of thegraduate students and chemistry teachers who collaborate withus in the development of the CTLP. We would also like to givecredit to Jason Green for allowing us to use Fig. 1. Any opinions,conclusions, or recommendations expressed in this paper arethose of the authors, and do not necessarily reflect the views ofthe funding sources.

References

Alonzo A. and Steedle J. T., (2009), Developing and assessing aforce and motion learning progression, Sci. Educ., 93(3),389–421.

Andersson B., (1986), The experimental gestalt of causation: acommon core to pupils’ preconceptions in science, Eur. J.Sci. Educ., 8(2), 155–171.

Andersson B., (1990), Pupils’ conceptions of matter and itstransformations (age 12–16), Stud. Sci. Educ., 18, 53–85.

Atkins P. and de Paula J., (2006), Physical Chemistry, 8th edn,Oxford: Oxford University Press.

Au K. T., (1994), Developing an intuitive understanding ofsubstance kinds, Cognitive Psychol., 27, 7–111.

Bensaude-Vincent B. and Simon J., (2008), Chemistry: the impurescience, London: Imperial College Press.

Berland L. K. and McNeill K. L., (2010), A learning progressionfor scientific argumentation: understanding student workand designing supportive instructional contexts, Sci. Educ.,94, 765–793.

Bernholt S. and Parchmann I., (2011), Assessing the complexityof students’ knowledge in chemistry, Chem. Educ. Res. Pract.,12, 167–173.

Bhattacharyya G. and Bodner G. M., (2005), It gets me tothe product: how students propose organic mechanisms,J. Chem. Educ., 82, 1402–1407.

Biggs J. B. and Collis K. F., (1982), Evaluating the qualityof learning: the SOLO taxonomy (Structure of the ObservedLearning Outcome), New York: Academic Press.

Breslow R., (1997), Chemistry today and tomorrow: the central,useful, and creative science, Washington, D.C.: AmericanChemical Society.

Brown D. E. and Hammer D., (2008), Conceptual change inphysics, in Vosniadou S. (ed.), International handbookof research on conceptual change, New York: Routledge,pp. 127–154.


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




Brown N. J. S., Nagashima S. O., Fu A., Timms M. and WilsonM., (2010), A framework for analyzing scientific reasoning inassessments, Educational Assessment, 15(3), 142–174.

Bulte A. M. W., Westbroek H. B., De Jong O. and Pilot A., (2006),A research approach to designing chemistry education usingauthentic practices as contexts, Int. J. Sci. Educ., 28(10),1063–1086.

Bybee R. W., (1987), Science Education and the Science-Technology-Society (S-T-S) Theme, Sci. Educ., 71(5), 667–683.

Bybee R., (1997), Achieving scientific literacy, Portsmouth, NH:Heinemann.

Bybee R., Taylor J. A., Gardner A., Van Scotter P., Carlson J.,Westbrook A. and Landes, N., (2006), The BSCS 5E instruc-tional model: origins and effectiveness, Colorado Springs, CO: BSCS.

Carroll F. A., (1998), Perspectives on structure and mechanism inorganic chemistry, Pacific Grove, CA: Brooks/Cole.

Chamizo J. A., (2013), Technochemistry: one of the chemists’ways of knowing, Found. Chem., 15, 157–170.

Chi M. T. H., (2005), Commonsense conceptions of emergentprocesses: why some misconceptions are robust, J. Learn.Sci., 14(2), 161–199.

Chi M. T. H., (2008), Three kinds of conceptual change: beliefrevision, mental model transformation, and ontological shift,in Vosniadou S. (ed.), International handbook of research onconceptual change, New York: Routledge, pp. 61–82.

Claesgens J., Scalise K., Wilson M. and Stacy A., (2009), Mappingstudent understanding in chemistry: the perspectives ofchemists, Sci. Educ., 93, 56–85.

Cooper M. M., Corley L. M. and Underwood S. M., (2013), Aninvestigation of college chemistry students’ understandingof structure–property relationships, J. Res. Sci. Teach., 50(6),699–721.

Cooper M. M., Underwood S. M., Hilley C. Z. and KlymkowskyM. W., (2012), Development and assessment of a molecularstructure and properties learning progression, J. Chem.Educ., 89, 1351–1357.

Corcoran T., Mosher F. A. and Rogat A., (2009), Learningprogressions in science: an evidence-based approach to reform,Consortium for Policy Research in Education Report #RR-63.Philadelphia, PA: Consortium for Policy Research in Education.

diSessa A. A., (1993), Toward an epistemology of physics,Cognition Instruct., 10, 165–255.

diSessa A. A., (2008), A bird’s eye view of the ‘pieces’ versus‘coherence’ controversy (from the ‘pieces’ side of the fence),in Vosniadou S. (ed.), International handbook of research onconceptual change, New York: Routledge, pp. 35–60.

Dugger W. E., (2001), Standards for technological literacy,Phi Delta Kappan, 82, 513–517.

Duncan R. G. and Rivet A. E., (2013), Science learning progres-sions, Science, 339(6118), 396–297.

Duncan R. G., Rogat A. and Yarden, A., (2009), A learning progres-sion for deepening students’ understandings of modern geneticsacross the 5th–10th grades, J. Res. Sci. Teach., 46, 655–674.

Duschl R., Maeng S. and Sezen A., (2011), Learning progres-sions and teaching sequences: a review and analysis, Stud.Sci. Educ., 47(2), 123–182.

Eilks E., Rauch F., Ralle B. and Hofstein A., (2013), How to allocatethe chemistry curriculum between science and society, inEilks I. and Hofstein A. (ed.), Teaching chemistry – a studybook,Rotterdam, The Netherlands: Sense Publishers.

Eisenkraft E., (2003), Expanding the 5E model, Sci. Teach.,70(6), 56–59.

Enke C. G., (2001), The art and science of chemical analysis, NewYork, NY: Wiley.

French M., (1999), Conceptual design for engineers, London:Springer.

Gilbert J. K. and Treagust D. (ed.), (2009), Multiple representationsin chemical education, The Netherlands: Springer.

Gilbert J. K., De Jong O., Justi R., Treagust D. and van Driel J.(ed.), (2002), Chemical education: towards research-basedpractice, Dordrecht: Kluwer.

Grotzer T. A., (2003), Learning to understand the forms ofcausality implicit in scientifically accepted explanations,Stud. Sci. Educ., 39, 1–74.

Hesse J. J. and Anderson C. W., (1992), Students’ conceptions ofchemical change, J. Res. Sci. Teach., 29(3), 277–299.

Hoffmann R., (1993), How should chemists think? Sci. Am.,268(2), 66–73.

Johnson P., (2000), Children’s understanding of substances, part 1:recognizing chemical change, Int. J. Sci. Educ., 22(7), 719–737.

Johnson P., (2002), Children’s understanding of substances,part 2: explaining chemical change, Int. J. Sci. Educ., 24(10),1037–1054.

Johnson P. and Tymms P., (2011), The emergence of a learningprogression in middle school chemistry, J. Res. Sci. Teach.,8(8), 849–877.

Kahan D. M., Jenkins-Smith H. and Braman D., (2011), Culturalcognition of scientific consensus, J. Risk Res., 14, 147–174.

Kind V., (2004), Beyond appearances: students’ misconceptionsabout basic chemical ideas, 2nd edn, London: Royal Society ofChemistry.

King D., (2012), New perspectives on context-based chemistryeducation: using a dialectical sociocultural approach to viewteaching and learning, Stud. Sci. Educ., 48(1), 51–87.

Lacy S., Tobin R., Wiser M. and Crissman S., (2012), Lookingthrough the energy lens: a proposed learning progression forenergy in grades 3–5, Paper presented at the Energy Summit,2012, http://esummit-msu.net/content/looking-through-energy-lens-proposed-learning-progression-energy-grades-3-5, accessedon August 27, 2013.

Lehrer R. and Schauble L., (2012), Seeding evolutionary thinkingby engaging children in modeling its foundations, Sci. Educ.,96(4), 701–724.

Levy S. T. and Wilensky U., (2009), Crossing levels and repre-sentations: the connected chemistry (CC1) curriculum, J. Sci.Educ. Technol., 18(3), 224–242.

Liu X. and Lesniak K., (2005), Students’ progression of under-standing the matter concept from elementary to highschool, Sci. Educ., 89(3), 433–450.

Maeyer J. and Talanquer V., (2010), The role of heuristics instudents thinking: ranking of chemical substances, Sci.Educ., 94(6), 963–984.


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




Maeyer J. and Talanquer V., (2013), Making predictions aboutchemical reactivity: assumptions and heuristics, J. Res. Sci.Teach., 50(6), 748–767.

Mohan L., Chen J. and Anderson C. W., (2009), Developing amulti-year learning progression for carbon cycling in socio-ecological systems, J. Res. Sci. Teach., 46(6), 675–698.

National Research Council (NRC), (1996), The national science educa-tion standards, Washington, D.C.: National Academies Press.

National Research Council (NRC), (2001), Knowing what stu-dents know: the science and design of educational assessment,Washington, D.C.: National Academy Press.

National Research Council (NRC), (2003), Beyond the molecularfrontier: challenges for chemistry and chemical engineering,Washington, D.C.: National Academy Press.

National Research Council (NRC), (2007), Taking science toschool: learning and teaching science in grades K-8,Washington, D.C.: National Academies Press.

National Research Council (NRC), (2011), A framework for K-12science education: practices, crosscutting concepts, and coreideas, Washington, D.C.: The National Academies Press.

National Research Council (NRC), (2013), The next generationscience standards, Washington, D.C.: The National AcademiesPress.

Osborne J. F. and Dillon J., (2008), Science Education in Europe,London: Nuffield Foundation.

Padilla M. J., (1986), The science process skills, Researchmatters. . .to the science teacher, Archived by NARST as No.9004, March 1, 1990, http://www.narst.org/publications/research/skill.cfm, accessed on August 1, 2013.

Padilla M., Okey J. and Dillashaw F., (1983), The relationshipbetween science process skills and formal thinking abilities,J. Res. Sci. Teach., 20(3), 239–246.

Redish E. F., (2004), A theoretical framework for physicseducation research: modeling student thinking, in RedishE. F. and Vicentini M. (ed.), Proceedings of the InternationalSchool of Physics, ‘‘Enrico Fermi’’ Course CLVI, Amsterdam:IOS Press.

Rozin P., (2005), The meaning of natural, Psychol. Sci., 16(8),652–658.

Sadler T. D. and Donnelly L. A., (2006), Socio-scientific argu-mentation: the effects of content knowledge and morality,Int. J. Sci. Educ., 28, 1463–1488.

Sadler T. D. and Fowler S., (2006), A threshold model of contentknowledge transfer for socio-scientific argumentation, Sci.Educ., 90, 986–1004.

Schecker H. and Parchmann I., (2007), Standards and compe-tence models: The German situation, in Waddington D.,Nentwig P. and Schanze S. (ed.), Standards in science educa-tion, Munster, Germany: Waxmann, pp. 147–164.

Schwartz A. T., (2006), Contextualized chemistry education: theAmerican experience, Int. J. Sci. Educ., 28, 977–998.

Sevian H. and Stains M. N., (2013), Implicit assumptions andprogress variables in a learning progression about structureand motion of matter, in Tsaparlis G. and Sevian H. (ed.),Concepts of matter in science education, Dordrecht: Springer,pp. 69–94.

Sikorski T. F. and Hammer D., (2010), A critique of how learningprogressions research conceptualizes sophistication and pro-gress, in Gomez K., Lyons L. and Radinsky J. (ed.), Learning inthe Disciplines: Proceedings of the 9th International Conference ofthe Learning Sciences, Chicago, IL: International Society of theLearning Sciences, vol. 1, pp. 1032–1039.

Sjostrom J., (2013), Towards Bildung-oriented chemistry educa-tion, Sci. & Educ., 22(7), 1873–1890.

Slovic P., Finucane M., Peters E. and MacGregor D. G., (2003),The affect heuristic, in Gilovich T., Griffin D. and KahnemanD. (ed.), Intuitive judgment: heuristics and biases, CambridgeUniversity Press, pp. 397–420.

Smith C., Carey S. and Wiser M., (1985), On differentiation: acase study of the development of size, weight, and density,Cognition, 21, 177–237.

Smith C., Wiser M., Anderson C. and Krajcik J., (2006), Implica-tions of research on children’s learning for standards andassessment: a proposed learning progression for matter andatomic-molecular theory, Measurement, 14(1&2), 1–98.

Smith C. L., Wiser M. and Carraher D. W., (2010), Using acomparative, longitudinal study with upper elementaryschool students to test some assumptions of a learningprogression for matter, Paper presented at the annual meet-ing of the National Association for Research on ScienceTeaching, Philadelphia, PA.

Solomon J., (1996), Teaching science, technology and society:developing science and technology series, Bristol, PA: Taylorand Francis.

Spelke E. S. and Kinzler K. D., (2007), Core knowledge, Dev. Sci.,10, 89–96.

Stains M. N. and Sevian H., (2013), Uncovering implicitassumptions: a large-scale study on students’ mental modelsof diffusion, J. Learn. Sci., under review.

Stevens S., Delgado C. and Krajcik J. S., (2010), Developing ahypothetical multi-dimensional learning progression for thenature of matter, J. Res. Sci. Teach., 47, 687–715.

Taber K., (2009), College students’ conceptions of chemicalstability: the widespread adoption of a heuristic rule out ofcontext and beyond its range of application, Int. J. Sci. Educ.,31(10), 1333–1358.

Taber K. S., (2013), A common core to chemical conceptions:learners’ conceptions of chemical stability, change and bond-ing, in Tsaparlis G. and Sevian H. (ed.), Concepts of matter inscience education, Dordrecht: Springer, pp. 391–418.

Taber K. S. and Garcıa-Franco A., (2010), Learning processes inchemistry: drawing upon cognitive resources to learn aboutthe particulate structure of matter, J. Learn. Sci., 19(1), 99–142.

Talanquer V., (2006), Common sense chemistry: a model forunderstandingstudents’ alternative conceptions, J. Chem.Educ., 83(5), 811–816.

Talanquer V., (2008), Students’ predictions about the sensoryproperties of chemical compounds: additive versus emergentframeworks, Sci. Educ., 92(1), 96–114.

Talanquer V., (2009), On cognitive constraints and learningprogressions: the case of structure of matter, Int. J. Sci. Educ.,31(15), 2123–2136.


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33




Talanquer V., (2009), On cognitive constraints and learningprogressions: the case of ‘‘structure of matter’’, Int. J. Sci.Educ., 31, 2123–2136.

Talanquer V., (2010), Exploring dominant types of explanationsbuilt by general chemistry students, Int. J. Sci. Educ., 32(18),2393–2412.

Talanquer V., (2013a), School chemistry: the need for transgres-sion, Sci. & Educ., 22, 1757–1773.

Talanquer V., (2013b), How do students reason about chemicalsubstances and reactions? in Tsaparlis G. and Sevian H.(ed.), Concepts of matter in science education, Dordrecht:Springer, pp. 331–346.

Talanquer V., (2013c), Chemistry education: ten facets to shapeus, J. Chem. Educ., 90(7), 832–838.

Talanquer V. and Pollard J., (2010), Let’s teach how wethink instead of what we know, Chem. Educ. Res. Pract., 11,74–83.

Talanquer V. and Sevian H., (2013), Chemistry in past and newscience frameworks and standards: gains, losses, and missedopportunities, J. Chem. Educ., DOI: 10.1021/ed400134c.

Van Berkel B., de Vos W., Verdonk A. H. and Pilot A., (2000),Normal science education and its dangers: the case of schoolchemistry, Sci. & Educ., 9, 123–159.

Vilches A. and Gil-Perez D., (2013), Creating a sustainablefuture: some philosophical and educational considerationsfor chemistry teaching, Sci. & Educ., 22(7), 1857–1872.

von Aufschnaiter C. and von Aufschnaiter S., (2003), Theoreticalframework and empirical evidence of students’ cognitive

processes in three dimensions of content, complexity, andtime, J. Res. Sci. Teach., 40(7), 616-648.

Vosniadou S., (1994), Capturing and modeling the process ofconceptual change, Learn. Instr., 4, 45–69.

Vosniadou S. and Ortony A. (ed.), (1989), Similarity and analo-gical reasoning, New York: Cambridge University Press.

Vosniadou S., Vamvakoussi X. and Skopeliti I., (2008), Theframework theory approach to the problem of conceptualchange, in Vosniadou S. (ed.), International handbook ofresearch on conceptual change, New York: Routledge, pp. 3–34.

Waddington D., Nentwig P. and Schanze S. (ed.), (2007), Standardsin science education, Munster, Germany: Waxmann.

Wilson M., (2009), Measuring progression: assessment struc-tures underlying a learning progression, J. Res. Sci. Teach.,46, 716–730.

Wiser M. and Smith C. L., (2008), Learning and teaching aboutmatter in Grades K-8: when should the atomic-moleculartheory be introduced? in Vosniadou S. (ed.), InternationalHandbook of Research on Conceptual Change, New York:Routledge, pp. 205–239.

Wiser M., Fox V. and Frazier K., (2013), At the beginning wasamount of material: a learning progression for matter forearly elementary grades, in Tsaparlis G. and Sevian H. (ed.),Concepts of matter in science education, Dordrecht: Springer,pp. 95–122.

Zeidler D. L., Sadler T. D., Simmons M. L. and Howes E. V.,(2005), Beyond STS: a research-based framework for socio-scientific issues education, Sci. Educ., 89, 357–377.


Publ

ishe

d on

26

Sept

embe

r 20

13. D

ownl

oade

d on

08/

10/2

013

18:3

0:33



chemistry education research and...

Documents