.Q: 3’. 3— I... . .
.e. . .4 .p
. a .

        
       

 

a.n
£51.23?

:.t. '

9... o. 5‘2.» 3.

. I .3 ’i‘.v.‘h_

.15.).13.‘

I330...

         

 

 

                            

.- . .
. It...
1.

.. . ..
... to.

2 ‘z.

 

          

O

 

.4 V.
3. \b‘o ._ as)

a... \.
\qlu. .L‘O.‘“

 

Huygens”
V;‘ 5

           

6.. i.
v '40.!

 

nuns

. .¢. .
.. n .n. .031.
to

_ #9".

‘a'l‘fi

fl‘!9!19!%1\

 

      
                                   

. 4 .
g... 4... - o . I
u. . .. o. ... .

 

...

92"“

 

 

”a r .

. a. . . » . . . . .. . . . ‘ . . . ... . . . . . . . ... ‘ . . . . .. . . ..‘ a . . ‘ \ ..5~ “u.‘
Q . . . V . . . ‘ . , ., . . . . . . . . .V , ... . . .. .. .. . . . . . . boo...l!.l.l.o.
a ... . . . .. . ,.. A . . . u, a: ’0.‘. I.
. .a. .. . . ‘ , .

O. L 0..

\. . .

D

Q ..

o ... .

. .

Q. ..

D ..

O. .

9"?!”

.

. _ a . . . . . . . . A . , , .. . . . . . . .. y . . ..., . , . . . . . . . . . .. . . . C v V . .‘ A.v o‘c.
. . . . . ‘ ,.. . . . . . . . .. .V . . . .. . .. . .‘ . :0..bc‘.au"§l
‘ . ‘ , ‘ , . , . . . . . . . , I . . . v.a50:’3...

.‘uct. .Qo\‘C'_' .-
v

n g

'9'-

.1-

fut-o #yLvry u ~vr: v! ‘

.t!.... s. .

9-”.‘gjv‘gld

0.9:

.‘C’C.

.‘P’Q

'9H~°‘.:“~;‘1’
..E '

 

.4

   

o .L.

0%.; T‘Ot‘,

 

.v. c. ‘
. n .
, ... . . .. .
o l I ‘) It! . -
n . O | 3-. ..II I
.1 .. . . .. L . ..
I ' ' _ l
. 1. .u a.

9. 3.. ... .. . . p
O . \ ' . . . .o I. . u
I . . l

n .4 .. .
- 'l ‘
Q o ‘
.vt<. i i
. . o .. .
. Op. ‘ .
o ' I . A o .
I I V l ‘
. o . . . .. .
..

 

 

.- n u n u .. n . i

' .. .én . ;. . . ... . . .. ..... ..p q 1:... . . ”J, ..r .f. ,I- ..'L.

O Vufi... .. _ . . .. gm o . .. . . . . . . . 3. .o . . . 34 . . u . . .o . .| - .. ‘ A. . . c. o . V. g .1 .mfi”..n...n.1m‘ Aim.“ 9-8.3”.memrom‘M-hfiéns .

. .. H . . . . .n . ‘ . . .. o. .a . . T . .._.. —. . J 7 .~ I ..r.. o

- . . . . .. . .. . .. i1 . (. arr. 9.31.95. Lu .5 .
I ‘ . . . bot... . . . . \3 LI. .. .

 

 

2 J , LIBRARY
Michigan State
University

 

 

This is to certify that the
dissertation entitled

MATTER AND ENERGY TRANSFORMATION: AN
INVESTIGATION INTO SECONDARY SCHOOL STUDENTS’
ARGUMENTS

presented by

Kennedy M. Onyancha

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Curriculum, Instruction and
Teacher Education

 

 

 

Maw

Major Professor’s Signature

i/Z‘I/[o

 

Date

MSU is an Affirmative Action/Equal Opportunity Employer

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l 5108 K:IProj/Acc&Pres/CIRCIDateDuo.indd

 

MATTER AND ENERGY TRANSFORMATION: AN INVESTIGATION INTO
SECONDARY SCHOOL STUDENTS’ ARGUMENTS

By

Kennedy M. Onyancha

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Curriculum, Instruction and Teacher Education

2010

ABSTRACT

MATTER AND ENERGY TRANSFORMATION: AN INVESTIGATION INTO
SECONDARY SCHOOL STUDENTS’ ARGUMENTS

By
Kennedy M. Onyancha

Arguments are important to the construction of scientific knowledge and practices
including the development of skills and tools for assessing that knowledge. Whereas
research on arguments continues to accumulate, there is little evidence that this work
focuses on the development of both instructional and assessment tools to support students
in using empirically verifiable data and make connections of data to claims about natural
phenomena. In this dissertation study, I use a modified version of Toulmin’s (1958)
model of argument analysis to examine the kinds of Data and Warrants, and sometimes
Backing (elements of argument) students use to support the Claims they make about
matter and energy (e. g. see J in & Anderson, in preparation) in their oral arguments about
the Carbon Transforming Processes (CTPs) of Tree Growing (TG), Flame Burning (PB),
and Car Running (CR).

Findings from this study suggest that students use different kinds of elements to
support their Claims. In particular, more sophisticated responses tend to be characterized
by those elements that appeal to scientific principles. However, less sophisticated
responses tend to include elements that are, for example, analogical, and/or tautological,
as well as personal beliefs to support the Claims made about these processes, and thus
tend toward force-dynamic reasoning (Pinker, 2007). Implications for teaching, learning

and research in science education are included.

COPYRIGHT BY
KENNEDY M. ONYANCHA
2010

Dedication

I dedicate this dissertation study to my loving family. First, to my understanding, patient,
and lovely wife Stella, who has endured long years of conceiving and writing this study.
Second, to my adorable children Bitutu, Nyanchera, and Onyancha who have born my
long absence with a child’s unbelievable humility-they are the joy of our family. Third, to
my mother Josephine B. Onyancha for her motherly love from which I have always
drawn full and unconditional support. Fourth, is a piece of dedication to my siblings for
generously sharing their childhood dreams with me and urging me to reach for the skies.
Finally, to my late father Benson Onyancha who instilled in me a sense of respect, hard

work, truthfulness, and pursuit of knowledge.

iv

Acknowledgements

I owe my deepest gratitude to my adviser and Dissertation Director Dr. Charles W.
Anderson without whose sustained help and guidance during the conception and writing
of this dissertation, it would have been next to impossible to finish it. I wish to thank my
committee members, Drs. Thomas Bird, Angela C. Barton, and Cynthia Carver, for their
invaluable feedback during the writing of this Dissertation.

I would like to thank my graduate colleagues whose contributions made this Dissertation
a reality today: Li Zhan who helped me with reliability checks, besides data collection;
Hui Jin, Jing Chen, Jonathon Schramm, Hamin Back, Jennifer Doherty, and Dante

Cistema who helped me with Data collection and organization.

I will remain forever indebted to you all. O.M. Kennedy

Table of Contents

 

 

 

 

 

 

 

 

List of Tables ............ - ............ u - u - ,_ -- viii
List of Figures - - - ix
Chapter 1: Introduction and Rationale 1
Purpose of the study“ - 2
Practices of Responsible Citizenship .............................................................................. 3
Argumentation as Inquiry and Research Questions ........................................................ 8
Dissertation Overview .................................................................................................. 11
Chapter 2: Research Methods - 13
Context .......................................................................................................................... 13
Participants .................................................................................................................... 15
Data Collection: Clinical Interview Protocol ................................................................ 15
Role of interviewers ...................................................................................................... 18
Data analysis for Research Question 1 ......................................................................... 20
Step 1: Analyzing individual arguments ................................................................... 20
Examples of analysis ................................................................................................. 22
Step 2: Developing coding rubrics for Data and Warrants ....................................... 29
Step 3: Reliability Checks ......................................................................................... 38
Step 4: Finding patterns of association among Claims, Data, and Warrants ............ 39
Data Analysis for Research Question 2 ........................................................................ 39
Chapter 3: Findings 40
Characteristics associated with Data, Warrants, and/or Backing-Research Question 140
Types of Data ............................................................................................................ 41
Types of Warrants ..................................................................................................... 49
Associations of Data and Warrants to Claims- Research Question 1 ........................... 56
Patterns of association among Data, Warrants, and Claims ..................................... 56
Argument Levels of Achievement ............................................................................ 63
Comparison of characteristics of elements at two different points in time-Research
Question 2 ..................................................................................................................... 69
Summary ....................................................................................................................... 77
Chapter 4: Discussion and Conclusion w - ..... t 80
Limitations of the study ................................................................................................ 80
Assumptions .................................................................................................................. 81
Implications for Research and Practice ........................................................................ 83
Implications for science teaching and learning ............................................................. 84
Future directions in research ......................................................................................... 86
Appendix A: Interview Protocol -- 88
Environmental Literacy Carbon Interview - - 88

 

vi

 

 

 

 

 

 

Appendix B: Tools for Reasoning 97
Tools for reasoning: Matter and Energy ....................................................................... 99
Molecular Model Kits ............................................................................................... 99
Appendix C: Examples of Color coding - 101
Appendix D: Exemplar Worksheets for Reliability Checks - - -- 103
Appendix E: Proportions of Data and Warrant types 107
Appendix F: Pre-Post Comparisons ‘ - ............... 110
References , ‘ - 116

vii

List of Tables

Tabel 1: Rubric for coding for elements of an argument ...........................
Tabel 2: Levels of Claims ...................................................................
Table 3a: Example coding for specific Data in transcript ..............................
Table 3b: Example coding for specific Warrant in transcript ........................
Table 4a: Types of needs and exemplar responses ......................................
Table 4b: Types of results and exemplar responses ....................................
Table 5: Types of Warrants and exemplar responses .............................................
Table 6a: Overall Percentage (%) of types of Data at each Level of Claim .........
Table 6b: Overall Percentage (%) of types of Warrants at each Level of Claim. . ..
Table 7: Descriptions of characteristics associated with levels of achievement. .
Table 8a: Overall (%) Pre/Post Data comparison ........................................
Table 8b: Overall (%) Pre/Post Warrant comparison ...............................................

Table 9a: Overall (%) Pre/Post Data comparison with partial data excluded. . . . .

Table 9b: Overall (%) Pre/Post Warrant comparison with partial data excluded. . ..

viii

22

31

35

38

43

44

50

58

6O

62

72

74

75

77

List of Figures

Note: “Images in this dissertation are printed in color.”

Figure 1: Practices of responsible citizenship ............................................. 2
Figure 2: Frequency of indicators against levels of achievement ...................... 66
Figure A.l: Oak tree .......................................................................... 89
Figure A.2: Match and flame ............................................................... 92
Figure A.3: Car running ..................................................................... 95
Figure 3.1: Water and oxygen molecules ................................................. 99
Figure 32: Carbon dioxide molecule ...................................................... 99
Figure 3.3: Butane molecule ................................................................ 100

ix

Chapter 1: Introduction and Rationale

Important educational documents on reform-based science education (e. g.
National Science Education Standards, 1996) have focused on and advocated for helping
students to achieve scientific literacy. Research on science literacy, especially in learning
progressions (e. g. Alonzo & Steedle, 2008: Mohan, Chen & Anderson, 2009), is
expanding. In addition, school curricula have been developed partly in response to calls
focused on helping students to achieve proficiency in science (NRC, 2007). This regards
knowing, using, and interpreting scientific explanations of phenomena (NRC, 1996).

My study aligns with these goals for science teaching and learning. This study is
part of our larger environmental science literacy project that focuses on the quality of
students’ accounts (Claims) of natural phenomena: in this case Carbon Transforming
Processes (CTPs). In the project, we analyze claims they make relating to the role of
matter and energy in individual processes, such as tree growing, baby girl growing, girl
running, tree decaying, flame burning, car running, lamp lighting, and cross processes
and how these connect to claims they make about larger environmental issues, for
instance, global climate change.

The primary cause of global warming is the current worldwide imbalance among
three classes of carbon transforming processes: (a) organic carbon generation
(photosynthesis), (b) organic carbon transformation (biosynthesis, digestion), and (c)
organic carbon oxidation (cellular respiration, combustion). Mohan et a1. (2009) have
analyzed students’ accounts of these processes. This study is focused on the nature of
arguments (Carlsen, 1997; Erduran et al., 2004; Gotwals et al., 2009; Newton et al.,

1999) students construct in support of their claims. This is from a learning progression

perspective which is described as sequenced and successively more complex ways of
thinking about a topic that learners master and investigate over a broad span of time
(Mohan, Chen, & Anderson, in press; NRC, 2007; Popham, 2007; Smith, Wiser,
Anderson, & Krajcik, 2006).

Recent research on learning progressions (e. g. Alonzo & Steedle, 2008; Covitt et
al., 2009; Jin & Anderson, 2008; Mohan, Chen, & Anderson, 2008) has shown that
students have difficulties with the practice of tracing matter and energy in socio-
ecological systems. We view socio-ecological systems as the intertwining of the social
and ecological systems. Often, and as Mohan et al. report, students have matter and
energy disappearing in their accounts of processes involving changes in states or forms. If
research has to serve the goal of achieving science literacy for all students, then the
practices relating to student reasoning about matter and energy should be explored in-
depth as a way of informing both research and instruction. This way, it is possible to
make sense of some of the challenges students face in learning science and use or design
matching programs for supporting them in their efforts to overcome these challenges.
This study is focused on making sense of students’ arguments regarding matter and
energy in socio-ecological processes such as Tree Growing, Flame Burning, and Car

Running.
Purpose of the study

The purpose of this dissertation study is to seek to understand how students use
evidence in constructing arguments. This involves analyzing elements of arguments
(Toulmin, 1958) such as Data and Warrants, treated more fully under the analysis section,

to support Claims regarding scientific processes about matter and energy transformation

as a way of learning to talk science (Lemke, 1990). The view of learning to talk science
encompasses “observing, describing, comparing, classifying, discussing, questioning,
challenging, generalizing, and reporting among other ways of talking science” (p. 1).
The idea of learning to talk science in educational settings presupposes that,
besides helping learners to learn how to use scientific practices in their specific forms, it
is important too that their use does not impede such learning. This view tends to align
with the Practices of Responsible Citizenship proposed by Covitt et al. 2009. I use this
framework and argument as inquiry to contextualize my study within the larger
environmental literacy project. That is, I bound my investigation within two scientific
discourse community contexts: a) environmental literacy as described in the Practices of
Responsible Citizenship fiamework which I briefly discuss next and b) the literature
review relating to argument as inquiry.
Practices of Responsible Citizenship

I use Covitt et al.’s (2009) Practices of Responsible Citizenship framework, which
lays emphasis on the practice of inquiry and argumentation, to help me bind this
dissertation study within the larger environmental literacy project. In other words, this
study is a slice of the larger environmental science literacy project. Covitt et al.’s
theoretical framework relates to student involvement in intellectual work in the sense that
it advocates for, to illustrate, students’ engagement in socio-ecological issues in ways that
likely lead to making environmentally responsible decisions. In their own words, Covitt
et al. (2009) have argued that “when we judge that we don’t know enough to make an
informed decision, we investigate the problem, by inquiring directly into a situation or by

relying on inquiry conducted by others” (p. 8).

Covitt et al. thus conceive supporting students in engaging in evidence-based
scientific investigation and argumentation as practices of responsible citizenship and use
them to frame our understanding of students’ work regarding socio-ecological issues. I
present this conceptual framework in figure 1 below. This view presupposes that students
do not necessarily make decisions about socio-ecological systems based primarily on
scientific reasoning. Rather, that they do so based on “many other factors—students’
family and personal values, their common family practices, their identities, economic and
social considerations, etc...” (p. 5). This framework lays emphasis on four domains:

Investigating, Explaining, Predicting, and Deciding.

Discourses: Communities of practice, identities, values, fund of knowledge

Explaining and Predicting

(Accounts)

What is happening in this
H situation? What are the

likely consequences of

different courses of

action?

'\ /'

Deciding
What will I do?

Investigating

What is the problem?
Who do I trust?
What’s the evidence?

\WMJ
Figure I: Practices of responsible citizenship
Although I make frequent reference to the accounts students make in their
reasoning about Carbon Transforming Processes, this study mainly focuses on the
domain of investigating. This is because our larger environmental literacy project work
has covered the explaining and predicting domains as it relates to, for instance, water

(6. g. Gunckel, Covitt & Anderson, 2009) and carbon (Mohan et al., 2009). Specifically,
4

my focus is on how students use Data to defend the Claims they make in their oral work
about transformations in matter and energy.

Research has shown that, although scientific practices are advocated in major
science education documents [e. g. American Association for the Advancement of Science
(AAAS), 1993 & NRC, 2007], students and other people as well, face challenges in
carrying out this practice (Covitt etal., 2009; Lee & Songer, 2003). In this study, I use
the inquiry domain of responsible citizenship to inform data analysis and interpretation.
My hope here is to work toward contributing a possible solution to the challenges of
practice that students face. This domain has three main constructs: Identifying a problem,
deciding whom and what information to trust (information source and trust) and
evaluating evidence. In this study, I use the term “construct” to specifically refer to
concept. I briefly discuss these three constructs next.

Identifying a problem: is the first construct and its function is to guide the
proposed investigation relating to environmental issues. This could be in terms of teasing
out information about an issue at hand by asking such questions as what the problem is,
what is known about that problem and what needs to be known about it. Covitt et al.
(2009) have noted that students struggle with this construct when, to illustrate,
investigating socio-ecological issues. This could be because they lack the necessary skills
for using scientific information (Duschl et al., 2007) and therefore merely resort to, with
little/no questioning, using social information sources and in effect treating these as
authority.

Deciding whom and what information to trust: is the second construct and it

regards reasoning about sources of information. This may be in terms of identifying,

teasing out and selecting relevant sources of information needed for solving the identified
problem. This amounts to making decisions about what sources of information to trust.
Important education documents (e. g. NRC, 1996) recognize that encouraging students to
be skeptical and engaging them in critically evaluating sources of information is
important in making “personal and community decisions about issues in which scientific
information plays an important role” (Duschl, et al., 2007 p. 7). However, pedagogy,
curriculum and standards unlikely help students to achieve this goal because they tend to
treat science as consisting “of solved problems and theories to be transmitted” (p. 3).

Evaluating evidence: is the third construct and it regards evaluating and using
evidence in support of the claims made about the identified problem (Covitt et al., 2009):
That is, in carrying out investigations about a clearly identified problem, this ought to be
in concert with making decisions about what sources of data should be trusted, as well as
how compelling the evidence is for use in solving the identified problem. My study
focuses in particular on this aspect of investigating: using and evaluating evidence. In this
study therefore, I examine students’ reasoning in relation to argumentation as inquiry in
their responses to questions about Carbon Transforming Processes. But before I specify
the Research Questions that guide this study, T wish to note the link between
argumentation and inquiry.

There is an association between Covitt et al.’s (2009) account of practices
associated with decision-making in citizens’ roles and argumentation as inquiry. As
Covitt et al. inform us, people tend to ignore experts’ perspectives on important issues
such as global climate change because they either do not understand, for instance, the

practices resulting in necessary decisions or simply tend to perceive the decisions as

uncomfortable. Additionally, some individuals may base their decisions on sources of
information they believe to be reliable with little/no regard for investigation. A
consequence of this would be two or more individuals/ groups with opposing viewpoints
regarding environmental decisions with far reaching environmental implications.

On the one hand, if decisions about environmental issues are narrowly conceived,
they are likely to lead to negative individual and citizenship choices. The likely narrow
conception of important environmental issues points to possible challenges to science
education. For instance, how might educators prepare all learners to work toward making
environmentally responsible decisions now and in the future? For example, individuals,
especially those in influential positions, may make or influence others to make little/no
data-based decisions regarding, say Biofuel production (6. g. Gerbens-Leenes, et al.,
2009) with a likely result of planting certain crops that are unlikely to deliver the results
as claimed. To illustrate, these decisions may potentially lead to serious food shortages in
the long run (e. g. see Wadhams, 2009). On the other hand, if well conceived, decisions
' are likely to lead to scientifically responsible citizenship choices, for example, why
choosing energy efficient appliances over those that are energy inefficient as it relates to
carbon footprint makes sense.

While it is important to focus on source of information and/or data as an aspect of
making environmental decisions, it is equally important to see beyond source and
consider quality of arguments based on those data. Thus, Covitt et al.’s (2009) study
about the process of decision-making regarding socio-ecological issues suggests an in-
depth analysis of students’ claims about these processes and the quality of arguments

they construct in support of those claims. An important step toward scientifically literate

citizenry is to engage students in constructing arguments as they make sense of the world

around them.
Argumentation as Inquiry and Research Questions

Literature on science education (e. g. Driver, Newton, & Osborn, 2000; McNeil],
2009) presents scientific argumentation, as it does explanations, as a practice of scientific
inquiry. Indeed, the treatment of argument as a practice of inquiry is emphasized in
current reform-based science (NRC, 1996): That is, with a focus on promoting scientific
literacy among students, reform efforts point to the idea that in order to support inquiry,
science instruction and learning should be anchored in argument and explanation.
Moreover, researchers (e. g. Berland & Reiser, 2009; Clark & Sampson, 2007) view
argumentation as being a central practice of science upon which inquiry and instructional
goals are developed. Berland and Reiser, as well as other researchers (e. g. McNeil], 2009;
Sandoval & Millwood, 2005) contend that argument and explanation are interrelated
scientific practices of inquiry in that these not only emphasize building toward sense-
making and articulation, but also toward persuasion regarding phenomena.

The view of argument as an aspect of inquiry or investigation (I use these terms
interchangeably in this study) points to an age-old notion that substantial scientific
knowledge is gained and developed through argumentation (Clark & Sampson, 2007).
Indeed, Kilboum (2006) contends that studies that aim to contribute to knowledge tend to
“make claims. . .that are supported by argument and evidence.” And that these are
“opposed to claims based on unwarranted opinion, ideology, dogma, power, and

authority” (p 531). Again, this is suggestive: That argument is an integral part of inquiry.

The NRC (1996 & 2000) emphasizes the need to support students in the practice
of developing deep understanding of scientific knowledge and skills. Zembal-Saul (2009)
views this emphasis as supporting students in engaging in evidence-based scientific
arguments, a shift from merely engaging them in a less effective exploration and
experimentation focused on ascertaining scientific ideas which might be already known
to students. Additionally, Zembal-Saul notes that this shift signals a “relationship
between the goals of scientific inquiry and the practice of argumentation, constructing
and evaluating scientific arguments as an aspect of engaging in school-based scientific
inquiry” (p. 691). This is in line with other literature (e. g. Duschl, et al., 2007; Songer,
Lee, & McDonald, 2003; White & Frederiksen, 1998) which indicates that students who
engage in the practice of scientific inquiry of, say, identifying a problem, gathering data
and evaluating it, as well as drawing data-driven conclusions demonstrate higher gains in
science learning. These students too, according to literature (e. g. Mercer et al., 2004), are
likely to engage in scientific arguments and in effect are likely to learn the practice of
constructing data-driven arguments.

Furthermore, literature has indicated that learners who are engaged in the
practices of scientific inquiry are likely to be motivated to learn science (Mercer et al.,
2004; Mistler-Jackson & Songer, 2000; Okhee & Brophy, 1996; Tobin et al., 1999). To
illustrate, Mercer et al.’s study about teacher scaffolding of student argumentation
reported that those students who were engaged in argumentation contributed more to
discussions and collaborated to reach consensus (based on scientific reasoning) than
those who were not. Moreover, and as Bell and Linn (2000) inform us, students who

engage in the practice of inquiry-based arguments are likely to promote knowledge

integration. Additionally, these students’ belief of science as dynamic would likely be
related to the development of more complex arguments.

An equally important finding from the literature is that students who delve into
the practice of scientific inquiry are not only likely to improve their metacognitive skills
but also to experience conceptual change (e. g. Yore & Treagust, 2006; Duschl et al.,
1999). Additionally, these students are likely to engage in intellectual development
(V ygosky, 1986) based on, say, analytical (Touhnin, 195 8) rather than rhetoric arguments
(e. g. Driver, Newton, & Osborn, 2000). Consequently, these and other reasons arguably
provide the impetus to use inquiry practices in both science learning and instruction.

For purposes of expanding on what is known about how students use arguments, I
examined the quality of secondary students’ arguments in their oral responses to
questions about matter and energy transformation. To do this and to guide this study, I
used the following research questions:

Research Questions
1. What is the nature of secondary school students’ arguments about Carbon
Transforming Processes (CTPs) such as photosynthesis (e. g. Tree Growing),
and combustion (e. g. Flame Burning and Car Running)? How do these
arguments align with the already established Claims (e. g. Mohan et al. 2009)
about CTPs?

2. What is the nature of these arguments at two different points in time?

Before describing participants and data sources I include in this study (see
Chapter 2), I would like to briefly explain the specific arguments I address in this study.

Scientific studies present the nature of science knowledge as attempts to persuade others

10

of the validity of their claims, rather than consensus based on democratic processes (e. g.
Tippett, 2009). Indeed, other studies refer scientific argumentation to as the language of
science in which claims are made in one way or another, and supported by data of some
sort (Duschl, Ellenbogan & Eduran, 1999, in Tippett, 2009).

Whereas there are different forms of arguments, for instance, rhetorical/didactic
arguments which present one point of view (Driver et al., 2000), and
dialogical/dialectical arguments which explore different viewpoints during debate or
discussion (Tippett, 2009), this study is focused on (I will return to this in the data
analysis section) analytical argument as proposed by Toulmin (1958). This form of
argument follows the rules of logic and is advocated for in reform-based science (e. g.
Duschl & Osborn, 2002): That is, it is opposed to opinions and/or ideology. This study,
therefore, lays emphasis on the quality of arguments students construct and diverges from

the traditional rhetorical arguments characteristic of classrooms (Y ore, 2003).
Dissertation Overview

This dissertation study consists of four chapters. In this chapter I identify and
introduce the problem of the study by locating it within two contexts: First within our
larger Environmental Science Literacy project relating to Practices of Responsible
Citizens conceptual framework; and second within literature review relating to
argumentation as inquiry. I also briefly discuss the purpose of the study and identify
Research Questions that guide this study.

In Chapter 2, I present methods of data collection by first discussing research
context, participants, and data sources. Then, I discuss data analysis for Research

Question 1 in four steps: step 1 focuses on analyzing individual arguments by offering

11

examples of analysis; step two focuses on developing coding rubric for Data and
Warrants; step 3 discusses reliability checks; and step 4 focuses on finding patterns of
association among Claims, Data, and Warrants. Additionally, I discuss data analysis for
Research Question 2.

In Chapter 3, as an attempt to respond to Research Questions presented in Chapter
1, I present findings of this dissertation study culminating in a proposed Argument Levels
of Achievement.

Finally in Chapter 4, I discuss limitations of the study as well as contribution of

the study to the teaching and learning of science.

12

Chapter 2: Research Methods

Note: Images in this dissertation are presented in color.
In this chapter I describe the following aspects of my research methods:
0 The context of the study
0 The research participants
0 The protocol for clinical interviews that were my data source, administered at
two different points in time
0 Role of interviewers

0 Data analysis procedures for Research Questions 1 and 2
Context

This study is part of a larger multi-year Environmental Literacy Project work that
draws from a learning progression perspective (Mohan, Chen, & Anderson, 2009; NRC,
2007; Popham, 2007; Smith, Wiser, Anderson, & Krajcik, 2006). The goal of the project
has been to document students’ reasoning about events in socio-ecological systems
(Mohan et al. 2009). This includes their reasoning, at both the macroscopic levels (we
can see with our eyes) and microscopic levels (we cannot see but can use a microscope to
see), about carbon-transforming processes: Tree Growing, Flame Burning, Car Running
(all 3 are the focus of this study), Decay and so on. The broader focus of this project
therefore, is to develop a fiamework that describes students’ reasoning as it specifically
relates to Carbon Transforming Processes about the principles of matter and energy and
broadly scientifically responsible citizenship.

Whereas the larger project includes such strands as water cycle, Biodiversity, and

citizenship, this dissertation study is part of a carbon strand. For over five years now,

13

work in this strand has focused on how students reason about Carbon Transforming

Processes (CTPs) and developed a carbon cycle learning progression framework. This

framework is characterized by 4 Levels of Achievement (For a more complete

description, see Table 2, this section) as described by Mohan et al. (2009). Work on this

framework suggests that each student’s reasoning tends to fall into one of the four levels

briefly described as follows:

Level 1 students generate accounts that are characterized by force-dynamic reasoning
(Pinker, 2007). These students view, to use Pinker’s words, enablers (e. g. air, wood,
gasoline, & sunlight) as needs that satisfy natural tendencies of actors (e. g. trees,
flame, cars).

Level 2 students similarly develop accounts that are characterized by force—dynamic
reasoning but begin to recognize hidden mechanisms as driving carbon transforming
processes. However, this recognition is limited to, for instance, viewing enablers as
conditions for processes to happen rather than reactants in the same processes

Level 3 students tend to generate accounts that begin to recognize transformations in
matter and/or energy. However, they inconsistently apply the principles of matter
and/or energy by, for example, reasoning that matter is converted to energy (and vice
versa) during Carbon Transforming Processes

Level 4 students generate accounts characterized by a recognition that Carbon
Transforming Processes are driven by hidden mechanism: That transformations are
constrained by scientifically established principles such as those of matter and energy

conservation

14

In an attempt to make sense of students’ reasoning about Carbon Transforming
Processes, this study builds on this work by focusing on the kinds of data students present
in developing their accounts and how these (data) align with the four levels of
achievement just described.

Participants

In this study, I followed 16 secondary school students (6th to 12th grade) from four
secondary schools in rural and suburban southwest Michigan. This included 4 males and
4 females for each grade level, a total of 8 males and 8 females. These students
participated in a one-to two-month long learning progression intervention as they were
taught using designed instructional tools about carbon transforming processes. The
schools included two public middle schools, one public high school, and a math and
science center for gifted high school students. All of the four teachers were science
majors with at least a bachelor’s degree. Using Michigan Science Curriculum Standards,
these teachers tough units that that included Plants (e. g. needs of plants to grow and what
is food for plants); Animals (e. g. what makes up food we eat and What happens to food
in our bodies); Systems and Scale (6. g. Key Principles for reasoning about environmental
processes-Scale, Matter and Energy) and Decomposers (e. g. What happens to dead plant
and animal materials? and how decomposers work). Selected teachers and students came
from school districts with a largely Caucasian student population (approximately 88%).

In these schools, an average of 37% of the students received either fi'ee or reduced lunch.
Data Collection: Clinical Interview Protocol

I used data from clinical interviews regarding eight carbon-transforming

processes including photosynthesis, biosynthesis, digestion, food chains, cellular

15

respiration and combustion (Mohan, Chen, and Anderson, 2009; see also interview
protocol below). This data was from student pre-post clinical interviews conducted
during the 08-09 academic year. Student interviews lasted for approximately 40 minutes.
My specific focus here was on students’ use, if at all, of elements of arguments (Toulmin,
1958) described in detail in the analysis section, this chapter.

My analysis focused on the portions of the interviews that addressed three
processes: Tree Growing (TG), Flame Burning (PB), and Car Running (CR). Although
the interview protocol itself for these three processes is included in Appendix A, I offer
its brief description here below. This is for purposes of illustrating the sort of questions
we asked and how the interview progressed.

For both our large project on learning progression and this study, we developed an
interview protocol to elicit students’ understanding about eight focus environmental
events: tree growing, baby girl growth, girl running, tree decaying, flame burning, car
running, lamp lighting, and cross processes. All of these events involve organic carbon
generation, transformation, or oxidation.

The interview protocol contains a set of semi-structured questions for each focus
event. For each event, we started the interview with a set of general questions—questions
that use everyday language to ask about the actor and its enablers. For example, the major
general questions for Tree Growing are:

0 What does a tree need in order to grow? How does sunlight help the tree to grow?

0 Do you think that water will change into other materials inside the tree’s body?

16

However, these general questions are not effective in eliciting higher-level
accounts. Hence, we also ask follow-up higher-level questions which are more specific
about matter, energy, and processes. One example is:

0 You said that the tree needs Carbon dioxide and breathes out oxygen. Where do the

carbon atoms of C02 go?

Teaching experiments (for details, see e. g. Jin & Anderson, in preparation):
Before the intervention, the selected students responded to these kinds of questions.
Depending on class schedules, the start of the intervention varied from school to school.
During the intervention, teachers of these students used designed instructional tools
(Tools for reasoning) to help them (students) work toward constructing scientific
explanations of what happens to carbon during the aforementioned processes (Mohan et.
a1, 2009). An introduction to and an example of these tools is included in Appendix B.

After the intervention, the selected students responded to the same pre—interview
questions. The purpose here was to seek students’ reasoning about the same processes
before and after more targeted instruction. The pre-posts provided students with two
opportunities to explain their reasoning about Carbon Transforming Processes. This also
allowed for what I view as a wide enough range of responses from which I could examine
students’ arguments. I analyzed pre-post interview data fi'om 16 students (a total of
approximately 32 interviews) for the Carbon Transforming Processes of Tree Growing,
Flame Burning, and Car Running (a total of 83 arguments-some students did not
complete all the events in their interviews).

Choices about elementary students’ work and level 1 accounts: The fact that I

was specifically interested in how students justified Claims about matter and/or energy, I

17

did not include elementary students’ work in the analysis. This was because most of these
students’ accounts did not include specific Claims about matter and/or energy. For the
same reason, I did not consider Level 1 accounts from the selected secondary school
students in this study. That is, my focus was on acquisition of scientific forms of
argument or socialization into scientific practice, and not on analysis of a full range of
arguments developed by students at all levels. On the basis of these decisions, my
analysis will be limited to middle and high school students’ arguments. This is with a
likely consequence that I will be unable to establish a more complete picture of the nature

of arguments across grade levels.
Role of interviewers

Before I briefly discuss the intended role of the interviewers during data
collection, I would like to briefly provide a background in terms of planning for
interviews. This likely sets the stage for the development of and use of the interview
protocol, described in detail under Data Collection: Clinical Interview Protocol (see the
preceding sub-section). To move toward interviewing identified respondents, the research
team (PI, Co-PIs, and research assistants) planned for the interview. I should note here
that I was one of the research assistants working on the project. Our focus was to first
establish a common goal: To seek students’ ideas about such processes as Tree growing,
Flame Burning, Car Running, Lamp Lighting, and so on. Our overall goal was to use
these ideas to design classroom tools/materials for use in teaching and learning science
(see also introductory part of the interview protocol, Appendix A).

Second, we focused on constructing items around our overall goal, guided by the

idea that each question was designed in such a way that it clearly sought as much

18

information from the interviewees as possible. The constructed items were semi-
structured and rank ordered from general (e. g. needs of each event) to specific (high level
order) regarding matter and/or energy.

Throughout the planning sessions, we tried to make clear the intended role of the
interviewers as including establishing a working relationship with the respondents
through, for example, introducing her/himself to the respondent, explaining the general
purpose of our study, and explaining that there were no “right” or “wrong” answers and
that the respondent was free to ask any questions during the interview. Additionally,
interviewers were to ask questions in the interview protocol (see Appendix A) and record
data as verbatim as possible by videotaping the interview sessions.

Despite the identified efforts to assure consistency in interview approaches,
interviewers were not entirely consistent during the interviews. To illustrate, some
interviewers asked unintended leading questions. Although it may be true that the
questions these interviewers asked were those on the interview protocol, some of the
questions might have been worded in a way to suggest a likely sought for response: That
is, for the most part, some interviewers paraphrased respondent’s answers to the extent
that this reflected the interviewer’s views. Together, such questions and paraphrasing
yielded responses characterized by either “Yes”, “No” or both.

To minimize errors in my analysis therefore, I was not interested in “Yes” or
“No” responses. Rather, I specifically focused on analyzing students’ responses that had
little suggested interviewer responses—I focused on what I viewed as respondents’

verbal answers only but not those of the interviewer. The reason for focusing on students’

19

responses was to make sure that the Data and Warrants I was coding came from students

rather than from interviewers.
Data analysis for Research Question 1

My analysis for Research Question 1---What is the nature of secondary school
students’ arguments about Carbon Transforming Processes (CTPs) such as
photosynthesis (e. g. Tree Growing), and combustion (e. g. Flame Burning and Car
Running)? How do these arguments align with the already established Claims about
CTPs (e. g. those by Mohan et al. 2009?) included four main steps. First, I used color
codes to identify the key elements of Toulmin’s framework in each student’s arguments
about Carbon Transforming Processes. Second, I developed coding rubrics to classify
students’ Data and Warrants according to their nature and level of sophistication. Third, I
worked with a colleague to establish reliability for the codes. Fourth, I looked for patterns
of association among Claims, Data, and Warrants. These steps are described below.

Step 1: Analyzing individual arguments

In this study, I used student interview texts (transcribed verbatim), to examine
how they used data to defend their claims about how matter and energy are involved in
Carbon Transforming Processes (CTPs). To help me to follow students’ reasoning, I used
Excel to organize data based on each CTP. Then i color-coded students’ utterances for
elements of Toulmin’s analytical framework (e. g. see Examples 1 & 2, this Chapter).

Research Question I and T oulmin ’s analytical framework: For purposes of this
analysis I used a modified version of Toulmin’s (1958) model of argument to help me
code transcripts in terms of students’ arguments. The interviews were designed to elicit

students’ accounts or Claims (C). In particular, I was interested in the claims that students

20

made about transformations of matter and energy during CTPs. I therefore sought to
understand how students supported their Claims with Data (D) and usually Warrants
(140. These three elements constitute what Toulmin calls a basic argument.

I also examined how students used Backing (B) if at all, to construct arguments
relating to Carbon Transforming Processes. But the interview protocols rarely elicited
what Toulmin refers to as complete arguments (constituting Claims, Data, Warrants and
Backing). I provide descriptions of these elements in Table 1 below, which I first
generated from preliminary data analysis. A possible reason why the interview protocol
rarely elicited complete arguments could be the nature of questions we asked-«we rarely
challenged students to justify their Claims in detail. Although my analysis focuses on
basic arguments, I include analysis based on the element of Backing.

Table 1 has three columns: the first column shows the type of element present in
an argument as identified by Toulmin (1958). The second column indicates Toulmin’s
descriptions of those elements. The third column shows a modified version of Toulmin’s
descriptions of the same elements. I included this modification to align with data for this
study. In this study, I did not consider the elements of Qualifier and Rebuttal. This is
because these did not emerge from data analysis.

Most students’ accounts of Carbon Transforming Processes included the elements
of Toulmin’s basic argument. For example, through the questioning process, we began by
asking students to provide what Chi (1997) calls “messy” data in the sense that it is in the
form of verbal observations. In this study, I operationalize (Feest, 2005) Data to refer to
information relating to, for instance, visible observations (inputs and outputs) about given

Carbon Transforming Processes.

21

Table 1: Rubric tor coding (or elements at an argument

Element Toulmin’s Description

My Description

 

 

 

 

Claim The conclusion whose Statement(s) students make about how
(C) merits the proponent of matter and/or energy are involved in
the claim seeks to CTPs: Relate to hidden mechanisms
establish
Data (D) Evidence that the Visible observation(s) about CTPs,
proponent of the regarding a claim that students may
argument clearly appeals make: May include verbal observations--
to as a basis for the - typically statements about needs of
identified claim organisms or conditions for processes to
occur and statements about visible
results of processes.
Warrant General, hypothetical Universal premises students make that
(W) statements, which can link either one type of data and/or
act as bridges, and different types of data to the claim
authorize the sort of step regarding specified CTPs.
to which our particular
argument commits us
Backing The credentials which Universal premises students make that
(B) are designed to certify link warrants to theoretical fi'ameworks

the beliefs of the warrant

which explain hidden mechanisms of
CTPs

Then, we probed students to explain how that Data linked to the Claim they made

(Warrants). Whereas I was particularly interested in Warrants that mention principles

such as conservation of matter and energy or hidden mechanisms, I was also interested in

analyzing other types of Warrants that students generated. Following this, we asked them

to provide further information that supported the connection they made between Data and

Claim (Backing). Here are two examples to illustrate the sort of analysis I did.

Examples of analysis

Example 1 illustrates a student’s work (transcript) that uses Data and Warrants to

support the Claims made in ways consistent with scientific standards of argument. By

contrast, example 2 illustrates a student’s work (transcript) that uses Data and Warrants in

22

a more analogical sense to support the Claims made. Both transcripts focus on the
process of Flame Burning (FB). In this process, we provided students with two pictures:
one represented a match burning, and another represented a candle burning. For other
examples of color coded Data and Warrants, see Appendix C, Tables C1 and C2
respectively.

Example 1: More sophisticated response. In the following interview transcript
about match burning, I demonstrate how the dialogue between an interviewer (I) and a
student (ANW) proceeded. In addition, I show how this dialogue likely aligns with
Toulmin’s analytical framework and how data analysis likely proceeded. The color
codes, in the two examples below, represent specific elements as shown in the analysis
after each transcript:

1. I: What does a flame need in order to keep burning?

2. ANW: It needs oxygen, wood, wax and wick in order to keep burning

3. I: What is in wood that makes it burn?

4. ANW: Wood has chemical energy and that’s what makes it burn. You have to
5. have stored up energy to make it burn.

6. I: So, talk about chemical energy of the wood. So, when wood is burning,

7. where does that chemical energy to go?

8. ANW: It’s what’s being burned.

9. I: So, do you think the chemical energy still exists or somewhere or changing
10. to some other types of energy, or just burn up?

11. ANW: It Changes into heat and light energy.

12. 1: Oh, so chemical energy changing to heat and light energy. Very good. So,
13. how about wood?

14. ANW: When wood burns the, it gives off the same things from the candle

23

15. burn, carbon dioxide and water (inaudible).

Data: From this dialogue, ANW offers, what she considers to be needs or inputs
of the process of Flame Burning that it “needs oxygen, wood ...in order to keep burning”
(line 2, blue highlight). Given the descriptions in Table 1 above, this chapter, I consider
this to be Data. Noticeably, in this interview, the interviewer influences the direction of
the dialogue in, for instance, focusing it on wood only with no mention of other needs for
Flame Burning (FB) that ANW identifies. For this reason, the interviewer probes about
the specific premise of the need for wood for the flame to keep burning. Wood, as a need
for flame burning, is therefore presented both as a source of chemical energy and as a raw
material for matter transformation. Indeed, when further probed about the material of
wood, ANW proceeds to account for it saying that it is given off in the form of carbon
dioxide and water (lines14-15). I regard these two products as Data in the sense that they
are visible results of F B which ANW uses to make claims about energy and matter.

Claim: When the interviewer further probes about energy, ANW provides
information about the energy of wood, that it “changes into heat and light energy” (line
11, green highlight). That is, ANW seems to suggest that when a match burns, the
chemical potential energy of the wood is transformed into other forms of energy, in this
case, heat and light. Based on the descriptions in Table 1 above, I consider this to be a
Claim about how energy is involved in Flame Burning (FB).

In addition, she points to the idea that wood, on burning, chemically transforms,
implicitly though, into water and carbon dioxide (lines 14-15, green highlight) in the
argument she makes: That is, ANW suggests that some hidden mechanism happens to
wood with the resultant observable water and carbon dioxide. I consider these two

statements as Claims, one about energy (from the preceding paragraph), and another

24

about matter, in the sense that the argument develops around both energy and matter:
That is, these are the main parts of the argument around which the interview is
developed.

Warrants and Backing: After being probed by the interviewer about how wood
helps the flame to burn, ANW points to the idea that wood has chemical energy (line 4,
Pink highlight). This statement suggests the notion that wood is a source of firel
necessary for the process of Flame Burning. This way, the statement would serve as a
universal premise to link wood to the process of PB. Thus, this statement would be part
of the Warrant she provides to support the need for wood in this process. Moreover, she
seems to more fully offer a universal premise in support of the idea that wood is needed
for the flame to keep burning saying, “You have to have stored up energy to make it
bum” (lines 4-5). This seems to be what Toulmin calls personal knowledge that wood has
indeed energy necessary for the flame to keep burning. ANW therefore successfully links
Data to the claim she makes regarding both energy and matter transformations. ANW’s
work contrasts with J M] ’3 work which I present in example 2 below.

Moreover, ANW correctly suggests the idea that both energy and matter are
neither created nor destroyed during flame burning (lines 11, 13-15). Rather, though
implicitly, that these are conserved during this process. That is, ANW seems to point to
the idea that the energy and matter of wood are constrained by the laws of conservation
of energy and matter which explains the hidden mechanisms, in this case, relating to
Flame Burning (FB). 1 interpret this implicit understanding and use of universal

(scientific) laws in support of the identified warrant (see Transcript above) as implicit

25

Backing. This is on the basis that it supplies more information about not only warrants

but also the claims ANW provides and in the process, validating them.

Example 2: Less sophisticated response. In this example, I provide and analyze

an interview dialogue between an interviewer (I) and a student J M] for the same process

as in example labove ---F B. This analysis pertains to how J M] attempts to both use Data

and link it to the Claims, about matter and energy, she attempts to make in the interview.

Here is the interview dialogue:

l6

l7.

l8.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33

I: What does a flame need in order to burn?

JMJ: It needs the gas that they put on it. Like...

I: What gas?

JMJ: The gas that burns like for the candle or the match ...like the wood on the match
1: Ok. So what happens to the air when the flame uses it to keep burning?

JMJ: The air like gum Ill/it’ll over by all lllc’ gii.\'t’.\‘ in the flame. And then it uses the air.

I: Now what do you mean by take over?

JMJ: It like Imean if ci/i‘t'cirlt‘ is (1 gas ltiii it makes ii like a burning gas.

I: Ok. ...Why does the flame need wax and wood ...? What happens to them?

JMJ: It will disappear because ...u‘m' uml mmu’ (ll't’ kind (if/17ccflames '_/iiml... nil/mill it,
they 'lljiisi die (ii/f

1: Oh. Ok. ...And then do you think the wax and the wood are kind of used up?

JMJ: Yes.

1: Ok. So do you think this burning is kind of related to energy?

JMJ: Yes.

I: Could you give me more explanation about that?

JMJ: burning using cncrgyjust to stay ali\‘C...H‘il/i(ml ...cm'rgv, it '3' going die

I: So do you think the energy is created... [or] comes from like a wax or a wood or air?

26

34. JMJ: Ithink it comes from — it '5‘ created. So it 's kind ofclmnisnjn So like when two
35. things come together, there‘s that energy to burn.
36. 1: Ok. So energy is created.

37. JMJ: Yes.

Data: As in example 1 above, J M] was provided with two pictures: one
represented a match burning, and another represented a candle burning. The dialogue
above is therefore based on questions about Flame Burning in the two pictures. From this
dialogue, and like ANW, JMJ provides what she considers to be an observation (Data)
that the flame needs “gas” (line 17, blue highlight) which she likens to the candle, match,
and wood (line 19, blue highlight) to keep burning. From this interchange, the interviewer
takes it that JMJ is talking about air, candle, and wood (line 19) as needs, and therefore,
in this case, constituting Data for the flame to keep burning. Moreover, after being
probed about what happens to wax and wood (line 24), the student seems to think that
they disappear. Thus, although JMJ ’s idea of wax and wood disappearing reveals her
thinking, it suggests that this reasoning does little to conserve matter. With the
assumption that J MJ is treating air, candle, and wood, as needs, the interviewer shifts his
questioning from seeking ideas for more needs to focus on more information regarding
these three needs.

Although the interviewer later uses wax in place of candle (line 24), he probes for
J MJ ’3 understanding regarding how (see Dialogue above) the three needs relate to flame
bunting. Thus, air, wax, and wood are seemingly treated as raw materials in the sense that
without, for example, wood and air (Oxygen), the process of flame burning will not
proceed. The shift in focus seems to be about seeking to understand J M] ’s thinking about

how matter and energy are involved.

27

Claim: To further understand how J M] reasons about Flame Burning (F B), the
interviewer explicitly focuses the student’s attention on both matter (lines 27 & 28) and
energy (lines 29 & 30). When J M] responds to the interviewer’s questions, she suggests,
implicitly, two points. First, that the matter of wax and wood are used up in flame
burning, and in effect implying that that is how it should be, a view that is force-dynamic
(e. g. Mohan et al. 2009) in nature. Unlike ANW who perceives F B as constrained by
transformations of matter and energy, JMJ perceives the flame as needing wax and wood
(matter) to keep it alive (lines 25-26). Compared to ANW’s perception, I consider J MJ ’3
perception as constituting a different kind of Claim: That matter undergoes some
mechanism with the result that, rather than change of form, it ceases to exist.

Second, J M] acknowledges that burning is somehow related to energy (line 30).
Nonetheless, when asked for firrther information regarding this relationship (line 31),
rather than focus on energy transformation, she contends that burning uses energy “to
stay alive” without which the flame will “die” (line 32). In contrast to ANW who treats
energy as one of the constraints of FB, J MJ perceives energy as causing this process to
happen and helping the flame to stay “alive.” In addition, J MJ maintains, as she similarly
did regarding wax and wood (line 25), that energy is “created” (line 34), rather than a
manifestation of energy transformation.

Furthermore, J M] points to the idea that some hidden mechanism, which she
refers to as “chemistry,” (lines 34 — 35) happens to result into the energy of burning. I
interpret this, implicitly, as constituting the Claim about energy. Again, this Claim is of a
different nature from ANW’s in the sense that it presents hidden mechanism in a

mysterious way (line 34, green highlight). In addition to specifically probing J MJ about

28

matter and energy, the interviewer also seeks to understand how J M] thinks the data
(inputs and outputs) link to the claim made.

Warrants: When J MI is asked by the interviewer about how air (lines 20 & 22),
wax, and wood (line 24) help the flame to burn, she reasserts her original Data (lines 21
& 23, Pink highlights). In fact, rather than provide a scientific bridge (Toulmin, 1958)
between these three needs and the Claim about matter and energy, J M] provides human
analogy that connects claims to data in an entirely different way (see pink italicized
texts). This is unlike ANW who points to the idea that wood has chemical energy (line 4,
pink highlight). JMJ uses a Warrant that is analogical in nature to link the identified Data
to the Claims—fuel for the fire is like food for a person. Arguably, using Warrants that
are analogical in nature, in contrast to ANW’s responses, is less sophisticated.
Step 2: Developing coding rubrics for Data and Warrants

These two examples just presented raise important questions that relate to my
research questions. For instance, what is the nature of all other individual students’
arguments? How do individual arguments relate to those of other students? These are
among the questions I used in both guiding further data analysis relating to Claims, Data,
Warrants, and/or Backing as well as identifying patterns that arose from all the 16
students. But before I followed examples 1 and 2 to code all of the 16 respondents’ data, I
made a decision about the unit of analysis.

I designated an individual argument as all responses by a student in a single
process. For example, if a student did both pre and post-interviews in relation to the
process of Tree Growing, then her/his responses will constitute two arguments (one for

pre, and the other for post). In this study, I decided to use a single argument, for instance,

29

pre-interview for a specific process such as Tree Growing, as my unit of analysis---post-
interview for the same process will constitute another argument. The reason for this
decision is that this study was not about case studies of the interviewees. Rather, it was
about the nature of their arguments about Carbon Transforming Processes. A
consequence of this choice is likely to be that I will not be able to specify why a students’
argument changed, if at all, in particular ways. The overall focus here therefore was to
identify characteristics associated with elements of argument (Claim, Data, Warrants,
and/or Backing) rather than how each student’s arguments changed. It is for this reason
that I used Mohan, et al.’s (2009) Levels of Achievement to classify the Claims in each
argument. Then, I developed rubrics for classifying Data and Warrants, as described
below.

Identifying levels of Claims: Mohan et al. define “Levels of Achievement as
patterns in learners’ knowledge and practice that [extend] across processes” (p 8). In this
study, I focused my description of levels of achievement on students’ knowledge and
therefore use of elements of arguments as described in Table 1 above to construct their
arguments regarding Carbon Transforming Processes: That is, for each level, I tried to
describe how each of the identified element is factored into the students’ arguments.

Mohan, et al. (2009) and other papers from the environmental literacy project
provide rubrics for sorting claims into levels of achievement. For example, Table 2
illustrates the rubric we are currently using to designate levels of explanations in

students’ claims (from J in, Zhan, & Anderson, in preparation).

30

Table 2: Levels at Claims

 

Level 4. Linking
processes with matter
and energy as

Linking carbon-transforming processes at atomic-
molecular, macroscopic, and global scales with
matter and energy as constraints

 

constraints

Level 3. Changes of Link macro-processes with change of molecules
Molecules and Energy and/or energy forms at atomic-molecular or global
Forms with scale, but cannot successfirlly conserve
Unsuccessful matter/energy.

Constraints

 

Level 2. Force-dynamic
accounts with hidden
mechanisms

Link macro-processes with unobservable
mechanisms or hidden actors (e. g., decomposer), but
the focus is on enablers, actors, abilities, and results
rather than transformation of matter and energy.

 

Level 1. Macroscopic
force-dynamic accounts

Describe macro-processes in terms of the action-
result chain: the actor use enablers to accomplish its
goals; the interactions between the actor and its
enablers are like macroscopic physical push-and-pull
that does not involve any change of matter/energy.

My study sought to find similar patterns in Data, Warrants and/or Backing. That

is, what is the nature of each of these elements at each level of Claim? In particular, I

report results based on analysis of arguments fiom all the 16 students. However, I should

note here that some students did not complete all the interviews---some completed pre

only, others post only, yet others partially completed pre-post interviews. I coded

arguments for the processes of Tree growing, Flame burning, and Car Running from pre-

and post-interviews for each student—a total of 83 arguments.

I was specifically interested in how students justified Claims about matter and/or

energy. I looked for patterns of association between the Levels of Achievement in

students’ accounts and the nature of the Data and Warrants (and sometimes Backing)

they used to support their claims.

Some of these characteristics are suggestive in the two examples (this Chapter):

31

For instance, ANW’S idea that “Wood has chemical energy... ” [Warrant] and
that this energy is transformed “...into heat and light energy...” implicitly
suggests an understanding of the law of conservation of energy-that energy is
neither created nor destroyed during flame burning [Backing]. This kind of
Warrant and Backing appeal to scientific principles regarding energy.

In addition, ANW’s idea that the flame needs “. . .wood ...” (input) to burn and
that “When wood burns it gives off... carbon dioxide and water...” (output)
[Data] demonstrates her understanding that matter inputs (e. g. wood) undergo
chemical processes resulting in different kinds of matter outputs. This kind of
Data, like the Warrant and Backing in bullet 1 above, appeal to scientific
principles regarding matter.

In contrast to ANW, JMJ ’s idea that “wax and wood are kind of like flames ’
food... ” [Warrant] and that, “without it, they ’11 just die off” points to the idea that
this student understands the process of Flame Burning in terms of actors and
enablers. This suggests a different kind of Warrant that is analogical in nature.
Moreover, J MJ ’s idea that when wood burns, “It will disappear” [Data]
demonstrates reasoning based on enablers and actors. This suggests a different
kind of Data that is readily noticeable in nature.

Second, after color coding responses from all the 16 students, I used Excel to sort

the transcript into Data and Warrants. I copied these into worksheets labeled All Data and

All Warrants respectfully. I have included examples of color coding for the two elements

of argument in Appendix C, Tables C1 and C2. This coding helped me to indentify, from

general codes, the specific codes present in an argument.

32

Rubrics for classifying types of Data in students’ arguments: Table 3a
represents a snapshot of the coding for specific Data that emerged from data for all the 16
students. It shows a number of codes and their labels (in parenthesis) as identified from
the data. I conceptually organized these codes based on how they relate to the principles
of mater and/or energy. This was in the order of codes with little or no reference to either
one of or both of these principles (least sophistication) to those that relate to one or both
of the same Principles (most sophistication). For specific examples and a more detailed
description of each of these codes, Tables 43 and 4b (Data) and 5 (Warrants), Chapter 3.

Before I get into the details regarding the kinds of codes I use to show patterns
that emerged, I wish to try to present the distinctions among the identified three
categories of the codes. Codes with least sophistication are compatible with force
dynamic reasoning: the needs and results that they identify are more like causes and
effects in a story than like matter and energy that are being transformed. They show little
commitment to conservation of matter and/or energy. Codes with intermediate
sophistication are characterized by some force dynamic reasoning and at the same time,
show some commitments to conservation of matter and/or energy---these codes suggest
unsuccessfirl conservation of the principles of matter and/or energy. Codes with most
sophistication are compatible with the principles of matter and/or energy: the needs and
results that they identify align with matter and/or energy transformation in the sense that
the inputs and outputs are consistently accounted for in a specified process. Here below, I
briefly discuss by offering examples of each category of these codes.

Codes with Least Sophistication: these include;

33

Needs Other (N O) which refers to the needs that suggest conditions-they do not relate
to either matter or energy
Results Energy Other (REO)-results that suggest specific forms of energy for other
processes and/or treats inputs as results and/or physical observations and/or matter
inputs as energy results
In this category too was the code Results Matter Other (RMO). These are Results that
suggest specific forms of matter for other processes and/or specific matter resource
for other processes and/or treats inputs as results and/or visible observations and/or
energy/matter inputs as matter results.

Codes with moderate sophistication: these include;
Needs Energy General (NEG)-suggest non-specific forms of energy/or energy source
Needs Matter General (NMG)-Needs that suggest non-specific forms of matter and/or
treats results as input
Results Energy General (REG)-suggest non-specific forms of energy for a process
Results Matter General (RMG). These results suggest non-specific forms of matter
and/or treats results as input.

Codes with most sophistication: Among these are;
Needs Energy Specific (N ES). These suggest specific forms of energy and/or location
of energy for a specific process
Needs Matter Specific (NMS) which point to specific forms of matter and/or specific
matter resource for that (specific) process

Results Energy Specific (RES) show specific forms of energy for a specific process

34

0 Results Matter Specific (RMS)-point to specific forms of matter and/or specific
matter resource for that (specific) process.

I should point out here that whereas Needs refer to inputs of Carbon Transforming
Processes, Results refer to outputs of the same processes-these are of different types as
indicated in Tables 4a and 4b, Chapter 3. For example, whereas gasoline and oxygen may
be inputs or needs for the process of Flame Burning, the Results or outputs would be
carbon dioxide, light energy and heat for the same process.

For each student’s argument, I used the code one (1) to indicate that a specific
Data type is present, and code zero (0) to indicate that a specific code is absent. For
example, from Table 3a, whereas both codes NMS and RMS are suggested (Blue
highlight) and therefore present (coded 1 each), all other codes are absent (all coded 0).
Table 3a: Example coding for specific Data in transcript

Data type present (1)/absent (0)

N
O

 

Transcript

cor-112
tag:
me:
032
crux
ca:
03511”
QM”
0:

 

Interviewer: What does the tree need in order
to grow?

RKC: The air, yes, all kinds of plants take in
carbon dioxide as animals take in oxygen,
plants take carbon dioxide and they exhale
oxygen for us to breathe.

Interviewer: I’m going to ask you for
different things...

How does water help the tree to grow? (Does 0 0 1 0 0 0 0 0 1 0 0
it change in any way? Where does it go?)
RKC: It helps create glucose for the food that,
not food, but sugar I guess that the tree grows
on and draws its energy from through
photosynthesis again. I use that word a lot. I
believe it’s 6HZO and CO; turn into C6leO6
with is glucose and then 602 which is oxygen.
So oxygen is like a byproduct of (inaudible)

 

35

Rubrics for classifying types of Warrants in students’ arguments: Like Table
3a, Table 3b represents a snapshot of coding for specific Warrants that emerged from
data for all the 16 students. It also shows a number of codes and their labels (in
parenthesis) regarding Warrants as identified from the data. These codes too fall into
three categories organized from the least to the most sophisticated.

Codes with least sophistication: these include;

0 Analogical (A) indicated by statements which point to inference based on
resemblance of enablers/processes that are otherwise dissimilar

0 Tautological (T) shown by statements that reassert the already mentioned data,
suggesting that these are sufficient to justify their Claims

0 Other properties of enablers Matter (OEM)-associate properties of matter
inputs/outputs to non-key matter enablers for a specific process and/or some
properties of one type of matter input/output to others

0 Other properties of enablers Energy (OEE) which associate properties of energy
inputs/outputs to non-key energy enablers for a specific process-«may also associate
some properties of one type of energy input/output to matter/others

0 Other properties of Actors (OA)-attributes properties of energy and/or matter
inputs/outputs to Actors for a specific process.

Codes with moderate sophistication: these are;

0 Citation of Evidence (CE)-responses associate needs with readily noticeable effects
0 Special Properties of Actors Matter (SPAM) where students, rather than view
enablers as inputs for specific processes, describe the actor instead-as characterized

by matter necessary for a specific process

36

Codes with most sophistication: These include;

0 Special Properties of Enablers Energy (SPEE)-present enabler as either characterized
by energy necessary for a specific process, as a reactant for a specific process or the
sun as the only source of energy

0 Special Properties of Enablers Matter (SPEM) which present the enabler as either
characterized by matter necessary for a specific process or point to the enabler as a
reactant for a specific process

I also included statements that suggest Backing under Warrants because these
were too few to be discussed separately. In addition, this element of argument seemed to
serve a similar purpose as Warrants (i.e. they link Data to the Claim made). These were
indicated by statements that suggested Principle of conservation of Matter (PCM) and/or

Principle of conservation of Energy (PCE). For each student’s argument, as in Data

(Table 3a, this Chapter), I used the code one (1) to indicate that a specific Warrant type is

present, and code zero (0) to indicate that a specific code is absent. As an illustration,

from Table 3b below, only SPEM (Pink highlight) was present (indicated by code 1)

from the transcript in the table. All other Warrant types were absent (all indicated by the

code 0).

37

Table 3b: Example coding (or specific Warrant present in trgrscqlt

Warrant type present (1)/absent (0)

 

 

Transcript ATCSSSSPPOOA
E P P P P C C E E O

E E A A M E M E
M E M E

Interviewer: I’m going to ask you for

different things How does water help

the tree to grow? (Does it change in any

way?)

RKC: It helps creme glucose forthe 0 0 0 l 0 0 0 0 0 0 0 0

food that, not food, but sugar ...dr'aws

its energy from through phtiitosynthesis
again. [believe it’s 61120 and (‘02 turn
into C6H1206 with is glucose

 

Step 3: Reliability Checks

To work toward reliability checks, besides worksheets labeled All Data and All
Warrants described in the previous subsection (see under Identifying Levels of Claims), I
developed an exemplar worksheet and a procedure for a colleague to help me with
reliability check coding. The colleague who helped me with the reliability checks was not
familiar with the nature of the coding I was doing all along. An example of the exemplar
worksheets that I developed for the reliability check is included in Appendix D, Tables
D1 (Tree Growing), D2 (Flame Burning) and D3 (Car Running).

Before starting on the reliability coding, I discussed all coding materials with my
colleague who was blind to the condition and student identity. For two students, he coded
pre-post for the process of Tree Growing, Flame Burning, and Car Running (a total of 12
arguments). My colleague and I agreed on the types of Data and Warrants present in the

transcripts for the two students 82% of the time, and when disagreements occurred, we

38

discussed and reached a consensus. Although for the most part the disagreements arose
from interpretations of codes and color coding, the discussions lead to some changes to
my codes that reflected his codes. Following this, my colleague coded other two students’
work that l similarly purposefully selected. This time, we agreed on types of Data and
Warrants present in the transcripts over 90% of the time.
Step 4: Finding patterns of association among Claims, Data, and Warrants
The final step in my data analysis related to Research Question 1 involved
looking for patterns of association among Claims, Data, and Warrants in students’
arguments. In particular, I looked for patterns that connected students’ Data and Warrants
with Levels of Achievement in Claims established in the project’s earlier research
(Mohan, et al. 2009). Details of the patterns that emerged from data analysis are included

in the next chapter.
Data Analysis for Research Question 2

I was also interested in the Claims that students made about transformations of
matter and energy during Carbon Transforming Processes after targeted instruction.
Therefore, I constructed and compared a paired pre-post Table each for Tree Growing,
Flame Burning, and Car Running (a total of 3 tables). This was based on Argument
Levels of Achievement from pre-post-interviews from the 16 08-09 students. This
comparison helped me to document how secondary school students’ arguments
developed after targeted instruction. Thus, in responding to this question, I followed a
similar procedure as in question 1 to generate comparative tables. These tables are

included in the next chapter.

39

Chapter 3: Findings

In the previous chapter, I outlined examples of the data analysis I did in an
attempt to respond to my research questions. This analysis suggests that students present
different kinds of Data, Warrants and/or Backing in support of the different kinds of
Claims they make in their oral arguments about matter and energy transformations in
Carbon Transforming Processes (CTPs). In this chapter, I discuss two key finding in
relation to Research Question 1: (a) characteristics associated with Data, Warrants, and/or
Backing and (b) associations of these elements to the already established achievement
levels for Claims (e. g. see Table 2, previous chapter) about the same processes. In
relation to Research Question 2, I compare these characteristics before and after more

targeted instruction.

Characteristics associated with Data, Warrants, and/or Backing-Research

Question I

My overall goal of this study was to use the already developed Levels of
Achievement Framework regarding Claims about CTPs (Mohan et al., 2009) as a
template to try to construct a Levels of Achievement Framework regarding Data, Warrants
and/or Backing---these were the elements that emerged from my data analysis. This
framework helped me to focus on argumentation as I attempted to make sense of
students’ texts about matter and energy transformation. What I wish to argue here from
my data analysis is that different students tend to use different types of Data, Warrants,
and Backing to support different Claims. But before I return to this point, I first present

the types of Data and Warrants that emerged from my data analysis.

40

Types of Data

Overall, I found that students presented Data that fell into at least 11 types (5
needs & 6 results) in defense of different types of Claims they made about Carbon
Transforming Processes. I present a summary of these types of Data including examples
of each in Tables 4a an4b, this Chapter). These are organized in the order from the least
to the most sophisticated: That is, low level (N O, RMO, REO), medium level (NEG,
NMG, REG, RMG) and high level (NBS, NMS, RES, RMS). For a brief review about the
distinctions among these categories of codes, see under “Rubrics for classifying types of
Warrants in students’ arguments,” preceding Chapter 2.

These different types of data show that students tend to treat inputs and outputs
for specific CTPs in different ways. On the one hand, some students treated needs as
inputs for chemical reactions involving matter and/or energy transformations. This is
evident from students’ responses in Tables 4a and 4b. To illustrate, I discuss an example
each from the processes of Tree Growing, Flame Burning & Car Running. This
discussion is organized around 3 categories into which these Data types fall: Low,
Medium, and High.

Students provided Data (inputs or outputs) to support their accounts in the form of
(a) “obvious facts,”---general observations which may or may not be empirically
verifiable, (b) specific observations which may be empirically verifiable or (c) a
combination of these two. For example, almost all students agreed with RKC’s account
of “what does a tree need in order to grow?” RKC responded, “it needs water for

nutrients or nutrients in the soil, sunshine for photosynthesis and a space to grow and

41

fresh air.” These Data were used, however, to support different kinds of Claims by

different students.

42

622v

 

 

 

 

 

 

She 2: Set $305 Aommoogmv 85 com “#:0on
...oEBmmw mecca cowxxo 2:on sense:95:39:38}, 08:82 838: omega SEEN been. 8322
: .cowxxo ”av—m was @003... UU<Z mecca : and mo mace.“ 058% “momma 35 382 €be
amzv
803:3 3803 2.38% ocmooam
...mvcon 355 e805 MES $55 a 8m 38% me 8:80— 5:28 $.88 Exam
2: mecca : ”93m ...300::. ”an Ew:::m.:mecoa = ”Mam .8 mace.“ 058% “momma 85 382 382
8sz
E550
gm. .. 8: its 808 39: mm 3.32 385 8:28 bag .«0 Egg
memo: : HU<Z a co. . .50 ”2):. 85 mecca...oo.s 05 :03 “am 8:8 oEocaméo: “momma 35 €002 €002
6%
3380
08:8 08:8 388 8:38 3.88 we 38.5
355 :< HU<Z xwuocm. .. um? 356 Kim meta omfioaméoe “momma 65 €002 €002
.: 03.6 3628 E3 365 .8 8:9: 556 9 828

9 352:8 :8 z. .. sonata a: 8 use mammémozih $50 67c
38:... ”M2 be ”07E. 2:2? ”32m 9 Smartesosmccoo 6033 85 $82 850 $32

...?—5.. ...in—:95: n03— 23.—w

3 too: .39 a G 3 can: use: a a 3 .36.... E 68: 8.: a 2
GEE:
3.2:. 25>? "AU MU Q .3 man .3. Oh. .5.— 832—8“ ..a—naunm egg—tomca— 2—5 3.5

 

wumec .3. :3 £38 3.3 «.33: e a.» h .6» Sack

43

 

 

 

 

 

 

.0802m 805:
N00. . .0203... ON: 0:0 05 0808 00:03 2.000%
00 00800: 20 .8 ~00 00:90:88-3 05 0:0 828% 0:0 3.0008 2.000% 0 8.: 800:: 8002
:8 00:80 :5 figm— 05...0.0 >22. ”am 00080 89:00 53% .8 08.8.: 2.000% “003% :05 $800M 288%
5:00: 0:0 Em: 5:00: m0 8 080: 05 $va
88: 0088.080: .8 Em: 05 000:0 m0 008% 8 20:2 2.000%
2 088.0% 000020: 2 $03083 0055 8:: 0080.080: 000008 080% 0 8.“ E008 ambfim
8 300:5 "IMO .8 300:6 ”MB 2 300:0 :80va ”Em 8 088 2.000% umcwmnm 00.0 288% 280.0%
...a 625
80020: 0 :00 20 82% 0008 80020: 300020: :05 :0 80:00
000032 000: 80 80% 088 0:0 00:30 05 0:0 £m0£§mo8£9 .. 89: 00 380.0: 3008 80:0 00:08 :0002
0:8 00w ”m3? 00w 0800 ”Saw 0000:: ”m2 .8 08:8 2.000%-8: “008% :05 388% 288%
808
80:00
3:08:88. 00:08 0020008 3.0008 0 8.0 $008 3:0:m
00:80... 0.: ”m2 0 0:0 000.: 0.: ”m2 $080206 .8 08:8 2.000%-8: “003% :05 288% 3800M
3800.: 00:08
00 0:88 0000800008 80:0 08003008

0080 w:_:0w 00—: 2 .0223: 80:0 3800: 00 08:5 800: AOEMV
>80 20m $0000 wEfi 05 85:00:03 00:08:08 80:0 00000008 00:8 8.: 008000.: 00:08 0050
05. .. 00880 00—5 05 8208 00:03 05 $00 380% 80:0 00000008 0008 :8 00:08 00002
8:: 0:8: 5 umhm 2 00:00... ”02,—. 8: £00: 0 0.: .0... ”DZ .8 2:80 2.000% 608% :05 3800M $88M

...?—5.. 0:03:25 @000— wBP—w 3

3 000: ..00 0 AU 3 000: 080: 0 a ..008 8 000: 00.: 0 2
"88 .23 "O :88
MD 0% .8 mm .2 QB 08 00u8%0~— 8.985— :etntumon— 093 80a—

.nwhfifi huh RQN EN?“ ‘33 aNsuhML Q ZN

 

44

Low level types of Data, examples of Needs Other, Results Energy/Matter
Other: Some students mentioned Data that described conditions or triggers for the event
rather than materials or forms of energy that are transformed during an event. I was
interested in this category because it might suggest force-dynamic reasoning, where
students focus more on causes that lead to the processes as effects than on how the
processes transform matter and/or energy. In other words, this is a case where students
view matter and/or energy as enabling the natural tendencies of organisms and objects to
change rather than being transformed and conserved during processes of chemical change
(Pinker, 2007).

Indeed, students in this category treated needs as enablers for actors to fulfill their
natural tendencies (J in & Anderson, in press). I describe these as Needs Other (N O). For
the most part, these students’ responses focused on visible processes, perceived or
otherwise, in terms of “action-result” with little recognition of matter or energy
involvement in any way. As an illustration and in contrast to BKD briefly discussed
under "High level types of Data ” below, STB’s responses to questions about Tree
Growing demonstrated a different kind of reasoning of the nature described by J in and
Anderson (in press).

When asked, “Okay. from those pictures it (the tree) starts as a tiny seedling
and then years later it’s a tree that weighs so much more than that. So where does that
extra weight come from over time?” STB responded, “I would have to say the like the
branches because in the first picture, the smaller tree, it’s like there is not very many
branches and then as it, it’s just small, and then when it gets bigger. Well the branches

have a lot of leaves on them and yeah.” Indeed, when probed further about whether “the

45

area around the tree would be changed at all beside the fact that the tree is growing” she
replied, “Well the tree is bigger so it needs like more air, it takes up more space and
yeah.” From these responses, STB seems to be focused on readily observable
comparisons such as branches rather than inputs for chemical changes leading up to
matter and energy transformation. Again, although she recognizes the tree’s need for air,
it is in a way that suggests that it (air) is physically rather than chemically needed to fill
up the space in the tree.

TNC’s responses to the same questions as it relates to Flame Burning were similar
to those of STB. Regarding the question about needs of a flame to burn, she says it needs
“any surface so it can burn ...” and in the process providing needs that suggest conditions
rather than inputs with little relationship or reference to either matter or energy.
Moreover, in relation to whether the needs change in any way, she points to the idea that
the “...candle is melting the waxes...then the thing is like getting tinier.” Thus, she
focuses on results that are visible in nature with little reference to transformation of the
suggested needs into the presented results. Other students’ responses resembled those of
STB and TNC by providing Other Needs (N O) that suggests conditions, characterized by
little relationship to either matter or energy.

Some students also used other results for either energy and/or matter
(REO/RMO). To Demonstrate, and in reference to Tree Growing, the interviewer raised
the question, “Does it (the tree) use all those things in the same way, the soil, the carbon
dioxide, water, sunlight? Do they all help the tree in the same way or do they work a little
differently? Do they change in any way?” STB reasoned, “The sun and the air help with

photosynthesis and then the water makes it so it doesn ’t get dehydrated.” Whereas, the

46

interviewer perhaps deliberately tries to guide and then probe for how the tree uses inputs
(C02, water, & sunlight), STB resort to reasoning that describes inputs in general (sun &

air or NEG/N MG) rather than specific ways as presented in the prompt.

Equally important is STB’s view of water as being used by the plant to prevent it
from dehydration. Although the resultant results (in italics, preceding paragraph) points
to a specific form of matter input (i.e. water), this is for another process: That is, plant
support which involves little transformation rather than growth which involves
transformation of water into other forms (results). I have included more similar examples
in tables 4 above, this Chapter.

Medium level types of Data, examples of Needs and Results Matter/Energy
General (N MG/N EG, RMG/REG): Other students provided needs that suggest non—
specific forms of energy and/or matter including their sources (N EG/N MG). My focus on
this category was based on the idea that it might reveal student reasoning that is neither
fully force-dynamic nor model-based. Said in other words, this category might suggest
student reasoning that is inconsistent with the principles of matter and/or energy.

Take the case of student SLP. When the interviewer asked, “What does the tree
need in order to grow?” She replied, Sun and water. And soil. Somewhere to be placed in
the ground.” This response suggests that, while SLP views one of the needs of Tree
Growing as being water (Matter), her reasoning is much more focused on general needs
(sun rather than sunlight/light energy or NEG) and conditions (place on the ground or
NO). This kind of reasoning was also revealed in other cases regarding needs for Flame
Burning as illustrated by TNC’s response saying, ‘if you’re making like a fire, it could

be paper or something that could keep the fire burning.” This too was similar to DRH’s

47

reasoning about Car Running: “It needs the gas, that’s the energy, and somebody
controlling the car.”

High level types of Data, examples of Needs and Results Matter/Energy
Specific (N MS/N ES, RMS/RES): A few students presented Data that pointed to
materials (N MS) or forms of energy (N ES) for transformations in chemical processes. I
was interested in this category because it might suggest principled reasoning, where
students focus more on how the processes of TG, FB, and CR transform matter and/or
energy. As an illustration, E] R, in his response to the question where the mass of the tree
comes from, treated these needs as raw materials for transformations in matter and
energy. He said, “The mass comes from the food that the tree is producing during
photosynthesis, which is mostly carbon and hydrogen pieces bonded together and that is
then stored away and eventually enough of it is stored away so that it starts to grow and
continues growing.”

EJ R’s reasoning is similar to that of BKD in the sense that it presents matter as a
reactant during the process of Car Running. From Table 4a above, when asked, “What
does a car need in order to run?” he replies, “Oxygen. It needs gasoline...” This response
suggests BKD’s treatment of and therefore reasoning about inputs to car running in terms
of specific forms of matter (e. g. gasoline). In addition (see Table 4b, this Chapter) when

asked, “Where do (those) needs you mentioned go? Do they change in any way or do
they remain the same? He replies, “it comes out of the tailpipe as water ...COz. . .” Thus,

BKD’S reasoning suggests his understanding of needs of car running as inputs for, in this

case, matter transformation. Accordingly, he recognizes that the matter of gasoline is

48

transformed into the water and C02 that is eventually released into the atmosphere as

byproducts of combustion.

To sum up, relatively few students provided “Data” in the scientific sense of
empirically verifiable observations. This may be in part a result of the questions we
asked; we rarely challenged students to justify their Claims in detail. For fiiture studies,
we will include requests for more detailed justifications of their Claims.

Types of Warrants

Similar to Data, I found that students used Warrants and/or Backing that fell into
at least 10 types to bridge Data to different types of Claims they made about Carbon
Transforming Processes. As in types of Data, I have organized these types of Warrants in
the order of the least to the most sophisticated as follows: Low level (A, T, OEM, OEE &
0A); Medium level (CE & SPAM); high level (SPEM, SPEE, PCM & PCB). For a
summary of the various types of Warrants and an example of each, see Table 5, this
Chapter.

These different types of Warrants seemingly support the notion that students tend
to bridge the Data they present to the Claims made for specific Carbon Transforming
Processes in different ways. On the one hand, some students used Warrants that, besides
being analogical in nature (e. g. see Example 2, previous Chapter), were, as an illustration,
tautological in nature. This is evident fiom students’ responses in Table 5 below (this
chapter). To demonstrate, I discuss an example each from the three processes-Tree
growing, Flame Burning and Car Running. This relates to my second finding discussed

after Table 5.

49

100.:

 

 

 

 

 

 

 

 

 

 

 

 

2.3.1.1 .311000 .111 0.111.818 0811.110: .8 ...00100.10 3.105..— .8 501108800

0100.01..1.1100 110.»... ”DEM 1111.1 001008 011 10111 1.11110.» $11.10 1.18.... ”mm—>2. 001 1.110 0 10111..» »1u1.10 1111.1 OM4— .8 0310:100—

120.11

..:1 11111...111111111110 011100 ...111.10.1.1 0111 :1 000111.10 ..0010>.00.1 . 0.811.110: 8 1:0 8:02 .8 :011080500

0111.111111000101 0.11 ”mm—>00. 11111111110 011100 0111 0.1.0.110 0.0-8111... HU<Z 0180 1.1100011 1031.11.11.80... 12):. .8 01:10:15

1001101

.5111000011 0111 111 ...w10110 1&8:m 08305—

111.111 A0:11000wv.11.v/1<1m ..»...11.10110 0111 1111111 0011100 330:3 11 “am 0111011100 1111111111115 ”in .8 00180195 101025

11 :1 11011100 AEm—mmv

0011 01111111181910 11 :1 11011100 1111111111111 1111 1,100.11 01 110.111.».110 0111 00011 .. .0111010 . 0111001111: 0111111180 110 8:02 0.1030:m

0011 858003... ”Em 11 001100011...110.111>.110 01100: 11010... ”DVD“ 0111 :1 01111111111 1101.100 011. .Mmm .8 00180051010015

...01110: .:.10101110 1111101110 111.1111 . 3213119 8:02 0810<

$500106 0111 1111 0.1101: 10111 01111111.. 0111... ”mm? 0011 00.11 0111 00 1.111111 0/. ”2):. .8 0018005 10100qm

0.10111 1.1101183 2.01:0: 0 0110111. . 010110101: 01811111.».0111111111:1.11111 1.1101180 AmUV

.:.0 10 0111111011011 13111111 11 .mqm 11101-80 ...1111100 0001.11 111.11 HU<Z 11 .. 111011111110 0111 11111111111... ”Em 00:001>m.8 :0100110

.:::3: »..1110110 ~02 201111100 0111 110.110 :81 01111001011110 0:.»11 01:00 A<Ov 0810<

0011 0:10:01 11 110.1» 102.1. 6 111 1.111.111.1111 0:10: 011 .1. DE Am .111 0110111...A0=0Uv >0: .1. ”mam 8 0011:0001: 101110

1.1101 1011110000111 .11 0.111010 30..» 1811.11

00.11 0111 0:0 3.81881 0000 1110.11 110181 1.1 0111 111:1. .110101: 0111 0.111.110 0010 1110 >180 AmmOv 3:0:m 0.10508

1.:01 :0wzxo 05... .02 ”am 110..» 11011.11 11111.11 001111.10. ....»..111.10:0.: “mkm :00 00111110111110....110 ”mam .8 0011890.:— .1050
00.11

0000.11. 1 0010.103 20100111111310 .11 0:0 0111 0:: 1110.11 10111 00111111 0111.8 AEm—OV 8:02 0:05:80

11 01100.1 003 11 “mFm 011111111 12.01.8110 11.11111 0011 A0031 11 .QBM 1111 11101.1 00111110 $001.15 11 ”07:. .8 00118110.:— 850

...0.»0E 0: 01:10.1“

00.111 0.110011 80 05 0x14 HOE 0.1: 0:10: 0111 1100.1 0.1111 111 110011 0.: ”M2 3-10.11.11.11 11011.1 11 0010111 11 ”0.0.2 A15 100%08809
.100 ...0101110 A00:

0111 10.1 810.1.» 0.1111 0.11... ”mHm 0.1.111 00-11100. 000.1 0 0000: 0501.1 ”E13,?“ 8.: 000.1 0111 0.1111 01 810.111.. ”mu—m A<v1001w010:<
.::.— 3 12:30... .»05 00 .18: . 5&0.— .m05 00 30mm
0055 0005 000: .100 0 AU . 0001.15 0005 000: 0:10: 0 5 10001115 0005 000: 00.5 $1.

0 0000 .33 "AU HO 0% .3 m..— A< Oh 8.. 000.8%?— ..0_1_E0xm— 2.5115 093 800.1015

 

50

My second finding therefore suggests that there are differences regarding
Warrants students provided to link the data to the claims they made. As in types of Data,
I have organized the discussion of Warrants that emerged in three categories, each with a
specific example-low, medium, and high.

Similar to Data categories, my interest in these categories was based on my
hypothesis that they might indicate either force-dynamic reasoning in which case students
view Data as: fulfilling natural tendencies of actors (e. g. Pinker, 2007)--low category;
characterized by special properties that align with scientific principles of matter and
energy--high category; or both of these twoumedium category. Again, these categories
are products of my conceptual organization for purposes of discussion. Thus others may
view them differently.

Low level types of Warrants, examples of A, T, OEM, OEE, 0A: Some lower
level students assumed that the Data themselves (tautological) were sufficient to justify
their Claims. To a large extent, students who used Warrants that were tautological in
nature tended to repeat, with minor modifications, statements as used in the interviewer’s
prompt. Suggestively, students who use these kinds of warrants limit their focus to
enablers, actors, and results with little reference to hidden mechanisms that drive matter
and energy transformations. To illustrate, when RKC, in his pre-interview, was asked
“How does a tree use water to make food for itself?” he said, “Water is very important
and it has to be clean water too because if it’s polluted then the tree could not survive.”
RKC’s response indicates an understanding that water is important for the tree to grow.
Yet when asked for a Warrant that connects the tree’s need for water with a Claim about

matter and energy, he basically reasserts his original Data. Thus, he falls short of making

51

the actual connection between water and photosynthesis. I describe this view as being
Tautological (T) in the sense that, rather than point to the characteristics of inputs that
drive chemical changes for processes such as TG, it presents inputs as physically being
enough to drive processes.

A] K’s reasoning about Flame Burning resembles that of RKC. This is in the sense
that he reasserts data, suggested in the interviewer’s prompt, as a justification for the
claim he makes. From Table 5 above, when asked, “Why does a flame need those things?
How do they help it to keep burning?” he replies, “It’s used to like keep the flame like
going...” This response implies AJ K’s view of data in flame burning, rather than as
characterized by matter and energy thus necessary for chemical changes, as being
naturally necessary.

Moreover, in reference to Car Running, TNC also uses tautological statements to
link the suggested data to the Claim he makes. In response to similar questions about CR,
he says, “Like the car needs gas to move...” Consequently, in their responses, RKC,

A] K, and TNC seem to focus on visible, as opposed to hidden, processes in their attempts
to bridge suggested Data to Claims made.

Other students attempted to bridge Data to Claims using Other properties of
Enablers Matter (OEM). I describe this type of warrant as either reasoning that
associates; properties of matter inputs/outputs to non-key matter enablers for a specific
process or some properties of one type of matter input/output to others. For instance,

when RWD was asked “So how does oxygen help the flame to keep burning?” He
replied, “It (02) has high energy bonds and it’s flammable.” In this case, RWD attributes

special characteristics of matter inputs such as gasoline (has high energy bonds) to

52

Oxygen, a non-energy reactant. Similarly, we asked BKD “Are oxygen and gasoline used
for energy?” Compared to RWD, BKD presents a reversed view of inputs (02 &

gasoline). In his response, he thinks that “the oxygen isn’t used for energy and the gas
(gasoline) isn ’t.”

Medium level types of Warrants, examples of CE & SPAM: Some students
used Warrants that focus on readily noticeable effects in a specified process as opposed to
effects as demonstrating hidden mechanisms. I describe this as Citation of Evidence
(CE). An example of this type of Warrant is provided by SJF. In response to the question,
“Do you think sunlight could play any role in that process? How does that work?” SJ F
reasons, “Yes, because without that — without the sunlight — it couldn’t fully produce
photosynthesis to help it grow.” This response tends to associate sunlight to
photosynthesis but with little focus on how that happens: That is, rather than view
sunlight as a form of energy that is involved in the chemical process of photosynthesis,
SJF views it as somehow related to this process.

Another medium level Warrant that emerged from data analysis is what I refer to
as Special Properties of Actors Matter (SPAM). I describe this as students’ reasoning
that, rather than focus on enablers as inputs for chemical transformations in a specified
process, focus on the actor instead as characterized by matter necessary for a specific
process. This Warrant is different from SPEM, illustrated under High level types of
Warrants below, in the sense that whereas SPEM describe inputs that actors (e. g. tree,
flame, and car) obtain from their surroundings for chemical changes, SPAM describe
actors themselves as miraculously constituting those characteristics. Here is an example

from data analysis. In the interview, we asked TNC, “You said that a tree has carbon. So,

53

where does the carbon come from? She said, “It comes from the tree.” Consequently, this
response suggests that TNC is unlikely to view the carbon of the tree as a chemical
component of carbon dioxide that was once in the air around the tree.

High level types of Warrants, examples of SPEM, SPEE, PCM, &PCE: On
the other hand, some students used Warrants that were characterized by Special
Properties of Enablers for Matter and/or Energy (SPEM/SPEE) to bridge the presented
Data to the Claims made. In some cases, some students used, implicitly or otherwise,
either the Principle of Conservation of Energy (PCE) or Principle of Conservation of
Matter to link Data to the Claims made. Compared to less sophisticated responses, more
sophisticated responses were likely to be characterized by these types of Warrants in the
constructed arguments. This is illustrated by an example each for the three Carbon
Transforming Processes of Tree Growing (TG), Flame Burning (F B), and Car Running
(CR).

With regard to questions about TG, RKC in his post-interview provides a more
acceptable Warrant about water and an explicit claim about matter saying, “It helps create
glucose for the food that, not food, but sugar that the tree grows on I believe it’s
6H20 and C02 turn into C6H1206 with is glucose and then 602 which is oxygen. So
oxygen is like a byproduct ....” Although he incorrectly uses the term “create”, his
response is much improved because it correctly links and therefore shows his
understanding that water is a reactant in the chemical process of photosynthesis. In fact,
his Warrant makes the general principles underlying this process explicit through

conserving matter by identifying its reactants and products.

54

Indeed, later on in the interview, when asked, “Do you think water is used up” in
photosynthesis? He said, “No. It’s not used. It’s not like you use it and it’s gone. It’s just
exists in another form and its’ in the leaves.” Thus, he shows use, and therefore
understanding of the idea that the matter of water is neither created nor destroyed.
Consequently, he provides Backing for his Warrant about why water is needed for the
process of Tree Growing.

Regardless of whether students were talking about matter (e. g. RKC above) or
energy (e. g. SAM, discussed next), more sophisticated responses focused on special
properties of needs. Accordingly, students’ reasoning seemed to suggest the idea that
these properties enabled, through hidden mechanisms, transformations of matter and
energy which were then manifested in results, either as part of the increase in mass, or as
products of transformations of matter and/or energy. As an illustration, SAM’s response
to questions about CR reveals reasoning of enablers as constituting special properties of
either matter or energy. When asked, “Why does a car need those things to run? How
does that happen?” He says, “It (gasoline) gives it energy...” Again, like RKC, SAM
correctly associates the (chemical potential) energy of gasoline with Car Running. BKD
too, in relation to questions about FB (see Table 5, this Chapter) presents similar
reasoning, “It (energy) comes from the energy in the wood...”

Furthermore, a few students’ use of Backing, besides Warrants, seemed to
function as well to either connect Data to the Claims they made or to justify the Warrants
they used. Notably, their use of Backing to reason about Carbon Transforming Processes

was either explicit or implicit. RKC from the preceding paragraph exemplifies an implicit

55

use of Backing in his response to the question about whether matter is used up during
TG.

In brief, relatively few students provided Warrants and/or Backing that aligned
with scientific principles to link Data to Claims made. As earlier noted about Data, this
may be partly due to the nature of the questions we asked, and future studies will be
enriched with the inclusion of requests for more detailed justifications of their Claims. I
view such probing as another opportunity to possibly identify further challenges that
students face in their accounts about Carbon Transforming Processes. This in turn may

lead to designing kinds of instruction that attempt to respond to the identified challenges.
Associations of Data and Warrants to Claims- Research Question 1

In this section I (a) describe patterns of association between the types of Data and
Warrants that students used and the levels of achievement of their Claims and (b) suggest
possible Levels of Argument based on common patterns of association among Data,
Warrants, and Claims.

Patterns of association among Data, Warrants, and Claims

I found patterns in students’ use of elements of argument-Data, Warrants and/or
Backing as well as their Claims. I present a summary of these patterns in Table 6, as well
as a summary of the nature of the differences in elements after the same Table below.

After identifying characteristics associated with Data, Warrants, and Backing, I
used the established levels of achievement for Claims (see Table 2 in the previous
chapter) to identify the kinds of Data, Warrants and/or Backing, for all the 16
participants, that align with those Claims. Some alignments are suggestive from the

examples provided including the two examples in the data analysis section, Chapter 2:

56

0 For example, the kinds of Data, Warrant, and/or Backing (both appeal to
scientific principles) provided by EJ R and RKC (see also ANW, Example 1 in
data analysis section) seem to align with Level 4 Claims described as “Linking
processes with matter and energy as constraints” (see Table 2, chapter 2).

0 By contrast, the readily noticeable Data and analogical Warrant provided by
J AH and TNC (see also JMJ, Example 2 in data analysis section, previous
chapter), seem to align with Level 2 Claims described as “Force-dynamic
accounts with hidden mechanisms” (see Table 2, chapter 2).

To move toward identifying patterns, and based on how matter and/or energy are
involved in Carbon Transforming Processes (characteristics of elements of argument), I
conceptually arranged and tabulated the types of Data (columns) against Levels of
achievement (rows) that emerged for all the 16 students in Table 63 below. I repeated the
same procedure for Warrants (see Table 6b below). This was on a scale from low to high,
low being characteristics least associated with the principles of matter and/or energy
(Blue highlight), and high being characteristics most associated with these principles
(Green highlight). Medium level types of elements (Brown highlight) are characterized
by, for example, reference to these principles in general ways, use of easily noticeable
effects of processes, and a focus on the actors rather than on enablers. In particular, the
two tables below show the proportions of two variables, Data (6a) and Warrants (6b) at
each level of achievement.

These proportions are based on the overall total number of elements (Data &
Warrant) for each category: That is, I computed the overall proportions by summing up

percentage proportions of each type of Data (Table 6a) and Warrants (Table 6b) for each

57

category. The categories were, Low, medium, and High (3 categories for each level, a
total of 9 proportions for Data, and 9 for Warrants). The reason for focusing on overall
proportions rather than individual proportions was to try to reduce complexity in
presenting results. I have included details about proportions of individual types of
elements at each Level of Claim in Appendix E, Tables E1 (Data) and E2 (Warrants).
Associations between types of Data and levels of Claims. To illustrate
computation of the overall proportions, and based on Table E1 (Under Appendix E), I
added percentage proportions falling under the category low for Data types Needs Other
(0%), Results Energy Other (3.9%) and Results Matter Other (7.7%), a total proportion of
11.6% (Low level, Claim level 4). For each Claim level and category, I repeated the same
procedure. The result is tabulated in Table 6a below. I should note here that I repeated the
same computational procedure for overall Warrant proportions based on Table E2 (see

under Appendix E). I briefly discuss the resultant proportions for Data, next.

Table 6a: Overall Percentage (%) at {mes ot Data at each Level at Claim

Claim Low Medium High
level level level level Total %

4 % of Total (26) 11.6 (3) 7.6 (1) 80.8 (22) 100
3 % of Total (96) 20.8 (20) 15.6 (15) 63.6 (61) 100
2 % of Total (171) 35.1 (60) 27.5 (47) 37.4 (64) 100

 

Table 6a indicates that, overall, students tend to use Data that align with scientific
principles (High level column) to support their accounts of Carbon Transforming
Processes (CTPs). However, compared to less sophisticated responses (those at Level 2),

more sophisticated responses (those at level 4) are less likely to include low level types of

58

Data in accounts about CTPs of Tree Growing (TG), Flame Burning (FB), and Car
Running (CR). This is indicated by generally fewer types of Data I designated Low or
medium: That is, the proportions of low level types of Data (Low level column) are lower
than those at high level (High level column, Claim Level 4). Moreover, the proportion of
these Data types tend to generally increase from upper left part of the Table toward the
upper right part of the Table.

Compare to high level types of Data, the proportion of low level types of Data is
particularly high at achievement level 2. In contrast to level 4 Claims which tend be
characterized by a high category of Data types, level 2 Claims tend to be characterized by
both of these Data types. This is evident from a similar overall percentage proportions for
high (3 7.4 %) and low (35.1%). This indicates that, less sophisticated responses are
nearly equally likely to include high level and low level types of Data in accounts about
CTPs. This likely point to the idea that students' reasoning about inputs and outputs in
Carbon Transforming Processes tend to be disconnected: That is, perhaps they view
inputs as fitlfilling specified processes with little recognition of matter and/or energy
transformation and outputs as a natural consequence of processes.

Additionally, Table 6a shows that, generally, students at level 3 tend to use both
low and high level types of Data. This is indicated by the fact that, although the
pr0portions of Data at this level are higher (Level 3, High level), these overall, tend
toward average. It is important to note here that the analyzed data fi'om which patterns
emerged were from pre-post interviews, initially analyzed blind to either of these

conditions.

59

Table 6b: Overall Percentage (%) at types of Warrants at each Level at Claim

 

Claim level Low level Medium level High level Total %
4 % of Total (13) 7.7 (1) 0.0 (0) 92.3 (12) 100
3 % of Total (96) 42.7 (41) 9.3 (9) 48.0 (46) 100
2 % of Total (167) 66.6 (11 1) 9.0 (15) 24.4 (41) 100

Types of Warrants associated with levels of Claims. Like Table 6a above,
Table 6b indicates that, overall, students tend to use Warrants that align with scientific
principles (Level 4, High level column) to support their accounts of Carbon Transforming
Processes (CTPs). For example, compared to less sophisticated responses (those at Level
2), more sophisticated responses (those at level 4) are less likely to include low level
types of Warrants in accounts about Tree Growing, Flame Burning, and Car Running.
This is shown by generally fewer low and medium types of Warrants at level 4. Indeed,
out of a total of 13 types of Warrants (low Category), students used only 1 (or 7.7 %) of
these at this level. In contrast, to illustrate, students used 12 (or 92.3%) of the high
Warrant category at the same level. That is, the proportions of low level types of
Warrants (Low level column) are hardly present at level 4. Similar to Data, the
proportions of the types of Warrants included in Table 6b tend to increase from upper left
part of the Table toward the upper right part of the same table.

Table 6b also suggests that although less sophisticated responses include some
high level Warrants (i.e. 24.4 %, High level column), they are more likely to be
characterized by low level Warrants (i.e. 66.6%, low level column) in accounts about
Carbon Transforming Processes. This is overall indicated by a particularly higher

proportion of low level Warrants at achievement level 2 (66.6 %, Low level column)

60

compared to either levels 3 (42.7%) or 4 (7.7%). Additionally, in contrast to high level,
the proportions of these types of Warrants tend to, generally, decrease from lower left
part of the Table toward the lower right part of the same table.

Furthermore, with regard to level 3, Table 6b generally shows that students tend
to use both low and high level types of Warrants. Illustrative of this is the fact that,
although the proportions of high level Warrants at this level are higher (48.0%, High
level column) than those to the left side (42.7%, Low level column), all cells at this level
are nearly equally represented at each level of elements. Consequently, students’ use of
Warrants at this level, overall, tends toward average. On the whole, my analysis showed
patterns of association between these two elements and Claims. These patterns are

summarized in Table 7 below, and a discussion of each is presented after this Table.

61

08.5000 0.80:5
00.0.. 00 0.0800
8 30:0 .80
00:08 80.80.:

0.003000sw\m..0..5 :0 .0005 5808 30:0
55:0 .0502 .5 00.0.2.0...— .0.00.5 08.5 .
0.008 8.0.0 0... o. 0000 U.:..
8 0.3000. 0.50:0 9.0.505. .5 0050000.... ..050 .
0.008
8.0.0 0... 8 0.00 0.8. o. 00:035. .5 8500.0
.80 050800000 50.00.0500. a00.3.80. .

8.0000 300:0

..o\ .80 .0302 00.005 0. 5500.800

0.0... 00.. .0... 3.8.00 ..050 8.0 .5

0.530003300205 :0 00005 3808 05.00%
30:00 02.80 00502 5.00.5 0.85 .

0.000: 30:0

5:80 0:02 00.00% 0. 5:00.800

0.0... 0>0.. .0... 0.0007. ..050 00.0 .5

0.53 0000M>m.0..05 :o .0005 5808 5007.
300:0 005.0 ..0502 0500.5 08.5 .

30:0

55:0 00:08 00.00% 0. 0.0.00 .0: >08
.0 >08 .0... 0:050:05: 3.0505050 .

 

.0... 8.0.0 05 000.00.08.50 8:00:03 ”.5 00.800 .0... 0000 m
0.08 8.0.0 0... o. 0000 0.8. o. 00:03.»... 50000 02.80 00:02 0. 5:00.800
00 8500.0 0:0 680800000 .00.»:550... 0.0... 5.3 0.30% ..050 ”0:050:05:
00.00.05. .085 00.0 .5 5000050002 .5 .0050 55:0 050 00.0.0. ”00.0 :5
030. 005.08.... .0..0:0w 000: .0... “50.000 0.500055 . 00.300. 300:0 .:: ..:0 00502 0.0.00.5 .

5.0030800 5.3
8000.800 ..:... .0:
0. :5 50:0 .80
00:08 ..0 8.80000
00.5.0:. .0... 8.0.0

0008 8.0.0 0... o. 500 0.8. 0. 00.00000... 0050
000.0. 00.0 :5 30:0 .5 00.2000...— .0.00.5 .
0.00.: 8.0.0 0... o. 300 0.8. 0. 005.000.... .050
000.0. 00.0 :5 00502 .5 00.8000...— .0.00.5 .
05 0050008050 3:00:03

30:1”. 55:0 0002 0. 8500.800
0.8.. 5.3 0.0007. ..050 ”80.08050
80:00 55:0 ..050 000.0. ”00.0
.5 .5007. 300:0 0.50:0 ..0502 00.005 0
mo 00.800 .0... 0.00 m

 

83.000008 8 ...0
..0 8.. .00. 30:0
.80 00:08 00
02.080800 5......
8000.800 8.0.0

50:50.82 ..0

00.8088. .8000w 0m: .0... 050.000 «5.02.0.5 .
8:0 03803 0:80.080

:0 o. 0000 0.8. 0. 30:0 .5 00.0.2.0...— .0.00.5 .
8.0.0 050......0.» 5.005080

:0 o. 0000 0.8. o. 00:02 00 00.9.2.0...— .0.00.5 .
.5 30.00.0850 3:08.03

50.00.8050 3.0505050
.0820: ..:0 5.000.. 00.0 .5
3.5.0”. 300:0 .:: ..:0 00502 0....005 .
0:050:05:
.000:00\..050 5000.00 ..:0 5.000.. 00.0
.5 0.0007. 0800:”.— ..e\..:0 ..0502 0....005 .
mo 00.0.80 .0... 0000 v

 

 

8.00:3 05080005 .000

62

Argument Levels of Achievement

I used coded characteristics of elements from all 16 participants that align with
each Level of Claim (3 rows) to construct a 4—column Table consisting of Level, Data,
Warrant and/or Backing, and Claim. In this table, I designated Argument Levels of
Achievement in students’ work regarding transformations in matter and/or energy based
on the emergent types of elements discussed in the preceding subsection. But before
presenting this summary, I would like to illustrate how I got there from the types of
elements that emerged discussed under types of Data and types of Warrants in the
preceding subsection.

Students used different kinds of Data, Warrants and sometimes Backing to
support the Claims they made and these fell into recognizable patterns. Less sophisticated
responses (mostly by younger students) used elements that were mainly characterized by,
for instance, individual beliefs, readily noticeable observations and interpretations based
on personal experiences (see Table 7, this Chapter).

To illustrate, most students’ Data and Warrants were similar to those provided by
J MJ (Example 1, Chapter 2): That the matter of wood/wax will not only “disappear”
during Flame Burning but also that both of these “. .. are kind of like flames ’ food.”
Students who provided these kinds of Data (e. g. readily noticeable) and Warrants (e. g.
analogical) tended toward Level 2 reasoning. This is illustrative from Table 6a which
shows that students at this level are more likely to use Needs Other (N 0), Results Energy
Other (REO), and Results Matter Other (RMO) to defend the Claims they make about
how matter and/energy are involved in the Carbon Transforming Processes of Tree

Growing, Flame Burning, and Car Running. This is in addition to being more likely to

63

use Analogies (A), Tautological statements (T), Other properties of Enablers Matter
(OEM), Other properties of Enablers Energy (OEE), and Other properties of Actors (A0)
to defend the same Claims (see Table 6b).

By contrast, more sophisticated responses were more likely to include elements
that appeal to scientific principles. These responses’ elements were similar to those
provided by EJR: That light energy (input) is not only transformed into stored (chemical
potential) energy (output) but also stored in the bonds of molecules (Warrant). Students’
responses that provided these kinds of Data and Warrants (consistent with scientific
principles) tended toward Level 4 reasoning. This is illustrated in Table 6a which shows
that such responses are more likely to be characterized by Needs (that are) Energy
Specific (N ES), Needs (that are) Matter Specific (N MS), Results (that are) Energy
Specific (RES), and Results (that are) Matter Specific (RMS) to defend the Claims made
about Tree Growing, Flame Burning, and Car Running. Moreover, these responses are
more likely to be characterized by Warrants that align with scientific principles to
connect the presented data to the Claims made about these processes.

Examples of the identified types of Warrants (see Table 5, this Chapter) students
use at this level include those that focus on Special Properties of Enablers Energy (SPEE)
and Special Properties of Enablers Matter (SPEM). Again, responses that tend toward
level 4 sometimes factor in, implicitly or otherwise, either the principle of conservation
of matter or energy to link the presented Data to the Claims the made.

Other students’ responses showed elements that appeal to scientific principles and
at the same time, for instance, used other elements that have little relationship to mater

and/or energy in reasoning about Carbon Transforming Processes. That is, these

64

responses tended to provide elements that were overall inconsistent with model-based
reasoning in the constructed arguments. Such students’ responses are exemplified by
DRH’s. Here is how he responded to questions regarding needs of a car to run, “It needs
the gas, that’s the energy, and somebody controlling the car. ” In spite of the fact that he
recognizes that a car needs gas (gasoline) as an input (N MS) into the process of CR,
DRH alongside this reasons that “somebody controlling the car” [Non-matter (NO) input]
is part of the need for this process. This is in addition to his idea of gasoline “being
energy” (OEM). Responses that use a combination of these kinds of elements that are
inconsistent with scientific principles seems to align with Level 3 Claims described in
part as “. . .cannot successfully conserve matter/energy. ”

Thus, the kinds of Data, Warrants and/or Backing that students used in support of
their Claims, viewed together, suggest a type of proposed learning progression that
includes most of Toulmin’s elements of arguments-A rgument Learning Progression. To
back up a moment, it is probably helpful to make a note about Levels regarding this
proposed Argument Learning Progression.

The levels suggested here should not be viewed as levels into which students’
work precisely fit. Rather, that these are attempts to classify approximations of coherence
of students’ arguments about matter and energy transformations in Carbon Transforming
Processes (Wilson, 2005; in Alonzo & Steedle, 2008). In framing these stages of
development therefore, I do not suggest that students’ arguments fit neatly into particular
levels but that these lean more toward particular levels than they do others. 1 illustrate this

perspective in figure 2 below which represents a way of thinking of these “Levels as

65

stages in a general transition from force-dynamic to model-based reasoning” (C. W.

Anderson, personal communication, December, 2008).

 

 

 

Force Model-
\ dynamic based /
indicators indicators
Frequency of
indicators
Level: 1 2 3 4

Figure 2: Frequency of indicators against levels of achievement

With reference to the questions we asked about Tree Growing, Flame Burning,
and Car Running, for instance, students provided as indicators (e. g. phrases, words, and
statements) of any number that I eventually used to identify the emergent types of Data
(see Table 4a & 4b, this Chapter). These Data could fall anywhere between Level 2 (e.g.
Needs/Results Other) and Level 4 (e. g. Needs/Results Matter/Energy Specific). Students
similarly used, for example, phrases and ideas that signaled different types of the
identified Warrants (see Table 5 above, this Chapter) to link the Data they identified to
the Claims they made about Carbon Transforming Processes. These Warrants in a
comparable way tend to fall anywhere between Level 2 (e. g. Analogical & Tautological)
and Level 4 such as Special Properties of Enablers Matter/Energy; Principle of
Conservation of Matter and/or Energy.

Moreover, although less sophisticated responses used more level 2 indicators,

some of these responses were also characterized by some higher level indicators.

66

However, this was in a way suggestive of either contradictions between needs and results
or guesswork indicated by a number of terms or phrases (e. g. I think, probably, I guess,

and I don’t know). To demonstrate, when TNC was asked about needs of a tree to grow,

she replied, “It takes in the air and it gives off C02” In this case, whereas TNC seems to

factor C02 into reasoning about Tree Growing, she presents it as a product rather than

input of the same process. Similarly, SLP’s response to whether the needs/inputs she
identified in Car Running changed in any way, she said “Carbon dioxide maybe, I think.”

In comparison with less sophisticated responses, more sophisticated responses
tended to have more Level 4 indicators. However, a few of these students’ responses also
used lower level indicators but often quickly included higher level indicators as the
interview progressed. Here is an exemplar response from a question about Flame
Burning:

I: The melting candle loses weight as (it) burns, how does this happen?

EJR: The wax of the candle will melt and then often it will pour over the side and

spread onto the table or whatever it’s sitting on, or else it will slowly evaporate

into the air.

I: You said it slowly evaporates into the air, what form is that?

EJR: I guess it would be wax vapor or something like that, and i1 (is) basically the

molecules ofthe wax spread apart and far enough. ..

From this interaction, E] R, like SLP, expresses an element of doubt when he uses
the phrase “I guess” in his response to how the candle loses weight and in what form. A
possible interpretation of El R’s response unlike SLP, based the idea that he quickly

resorts to using more of model-based reasoning (Pink highlight), is that this suggests the

67

level of difficult of the subject matter regarding Flame Burning. Thus, whereas EJ R’s
reasoning overall tends toward Level 4, he rarely draws on a few indicators and therefore
elements of argument that characterize lower levels. Accordingly, this is suggestive of
another level, in this case Level 3 (see Table 7 above) where students tend to
concurrently use high and low level indicators.

I conceptualize Level 3 as being characterized by student responses with
indicators (constituting either Data or Warrants and/or Backing) that seem, overall,
inconsistent with scientific principles. An example of L3 Data is suggestive in JAH’s
response to the question about needs for Car Running. He responds that a car “needs
oxygen for the combustion, gasoline, it needs a person to drive it and I guess that would
be all ...” Although J AH recognizes inputs/needs for this process to proceed in the form
of gasoline and oxygen (mater), he, like DRH discussed earlier, mentions other needs that
do not relate to either matter or energy (in italics) as being necessary for this process.

In another example regarding where the heat generated during Car Running
comes from, another student (RWD) said, “Some of it’s from the energy from the
gasoline. And, the other part of it is fiom the oxygen” Thus, RWD not only draws on the
notion that gasoline is characterized by (chemical potential) energy necessary for Car
Running, but also draws on a more force-dynamic idea that oxygen has energy (in

italics).1 In particular, with reference to matter, RWD uses both special properties of

 

I While chemists might regard this response as better than one that locates the energy only in the
gasoline, I think that RWD’s response indicates that he is thinking of oxygen and gasoline as

separate energy sources, rather than of oxygen as an element that reacts chemically with gasoline.

68

enablers (SPEM) and other properties of enablers (i.e. OEM, Table 5, this Chapter) in
their accounts.

A general pattern that implicitly emerges fi'om the preceding responses, as well as
other responses from all the 16 students therefore, is the notion that students seem to
accumulate ideas over time, each (student) presenting an image of an ongoing
construction of a toolkit of some kind (Anderson et. al, personal communication, May 14,
2010). It would seem that each student draws fiom their kinds of toolkits any number of
tools so to speak (e. g. ideas, words, phrases, & statements) to use in reasoning about
Carbon Transforming Processes (CTPs). Compared to less sophisticated arguments, more
sophisticated arguments draw more on indicators that align with scientific principles.
Accordingly, the resultant Argument Learning Progression regarding CTPs I propose
here, like ecological progression, is one of complexity with different dimensions to it
such as time, instruction, and instructional tools. In this progression, for instance,
students might, perhaps with more focused instruction as suggested in the pre-post
development I discuss next, to move over time fi'om less sophisticated reasoning (force-
dynamic) toward more sophisticated reasoning (principled or model-based reasoning)

about CTP in a way that represent a continuum (Wilson, 2005).

Comparison of characteristics of elements at two different points in tim e-

Research Question 2

A note about what this study does and does not do. The results I present
regarding Research Question 2 should be viewed more as would results from a pilot
study. This study does not assess the effectiveness of the instructional intervention we

used. It is focused on how students’ arguments developed rather than what caused this

69

development. The reason for this is that although teachers of the participating students
were expected to use designed instructional tools/Tools for reasoning (see Appendix B;
see also J in & Anderson, in preparation), we do not have reliable data about how teachers
used them.

Certainly, an inclusion of the details of the instructional intervention would enrich
a similar future study particularly focused on the effects of using these tools themselves.
So, my hope is that the findings regarding Research Question 2 will potentially inform
future investigations. For example, what actually helps students to move, if at all, from
one level to another? How does that movement look like in terms of gender and socio-
economic status? How do these relate to teachers’ years of teaching experiences, if at all?
Responding to such questions would likely point to other factors that potentially support
students in moving toward a scientific way of reasoning about Carbon Transforming
Processes. This way, a possibility exists for designing instruction that would be
responsive to the goal of helping all learners in pursuit of scientifically literate citizenry.

Even without evidence about the instructional mechanisms of change, pre-post
comparisons can be valuable for developing learning progressions. When we see changes
in patterns of Data and Warrants for arguments about the same processes fi'om the same
group of students at two points in time, that potentially provide evidence that the changes
involve learning rather than differences among individual students (C. W. Anderson,
personal communication, July 7, 2010). Moreover, findings regarding Research Question
2 can point to problem areas for students and accordingly suggest targeted interventions.

I now return to the idea of comparison of students’ argument Levels of

Achievement from pre-interviews with those from their post-interviews. Whereas the

70

preceding discussion, overall, show the specific types of elements of argument that
emerged (see Tables 4a, 4b & 5, this Chapter), it does not show how these elements
changed during pre—post-interviews.

To examine what and how the identified types of elements shifted, I compared the
total number of each of these elements in the pre-post-interviews by tabulating them in
Tables 8a (Data) and 8b (Warrants). Together, these tables summarize the patterns that
emerged which I briefly discuss after each table. This likely provided a way of
illustrating the sort of change in students’ responses that emerged from data analysis. I
was interested in how, overall, elements changed during pre-post-interviews. This
change, in the total number of types of elements in each category (low, Blue; medium,
Brown & high, Green) is indicated by the numbers in the columns labeled Change
(column C) on Tables 83 and 8b below. For a full description of these categories, see
under “Patterns of association among Data, Warrants, and Claims,” this Chapter.

Before I proceed with the larger discussion about pre-post patterns, let me point
out here that rather than focus on patterns emerging from individual types of Data and
Warrants, I present overall patterns for these two variables. As in associations of Data
and Warrants to Claims (Table 6a & 6b, this Chapter), this was my attempt to reduce
complexity in emergent patterns. Thus, the overall emergent patterns are based on the
total number of types of elements in each category in each of these tables. This was
computed fiom shifts in each type of Data (Appendix F1) and Warrant (Appendix F2).
For example, for Table 8a, I added the number of each type of element under the category
Other (NO, REO & RMO, see Table F1) for Pre and then for Post. Next, I computed

change from the sums, which in this case was 1 or 2.4 % (blue highlight, Column C). I

71

repeated the same procedure for other Data categories (Table 8a) as well as for Warrant

 

 

 

 

categories (Table 8b).
Table 8a: Overall (%2 Pre/Post Data comgarison
Data category Pre (A) Post (B) Change
lB-Al (C)
Needs Results Su Need Result Su Sum B -
m s s m Sum A
(%)
Low level (other) 23 18 41 15 27 42 1 (2.4%)
Medium level 26 9 35 22 11 33 -2 (5.9%)
(General)
High level 41. 1 8 59 41 46 87 28
(Specific) (38.4%)

Table 8a provides overall comparisons between the numbers of types of Data in
the pre-post-interviews. This includes data from all the 16 students irrespective of
whether they did all or part of the interviews. The upper part of the Table (Low level
row) shows what I consider to be low category (or Other), the middle part (Medium level
or General row), and the lower part of the same Table (High level row) shows high level
category (Or Specific).

The numbers in the cells in column C are similarly identified and these indicate
how each category of Data changed, with percentage change in parenthesis. For instance,
results suggest that overall, there was virtually no change in students’ response about
Data in the low level category (Other). This is indicated by an overall increase by 1(or
2.4%) ---I expected an overall decrease though. Similarly, there was overall no change in
students’ response about Data in the medium category (General). This is based on the fact
that there was an overall marginal decrease, as expected, by 2 (or 5.9%) in the differences

72

in sums for pre-post. Viewed together, results from these categories perhaps suggest that
students’ ideas tend to persist over time with a likely persistent force dynamic view of
phenomena including Tree Growing, Flame Burning, and Car Running.

The most expected shift occurred in the high category (High level row): That is,
there was a resultant increase by 28 (or 38.4%) in the difference in Sum for the Pre-Post
in column C on the lower part of Table 8a. Another pattern that is noticeable fiom Table
8a is that overall and compared to Needs, students talked more about Results in their post
than pre responses. For details about shifts in specific types of Data, see Appendix F
(Table Fl). For example, there were a total of 18 total types of Results fiom all of the 16
students’ pre-interview responses. By contrast, there were a total of 46 types of Results
from their post-interviews, a total increase of 28. The total types of Needs in these
students’ responses remained at 41 (no increase). Together, these patterns suggest that
maybe students were beginning to think more about the consequences of the processes as
well as their causes and less about transformations in matter and/or energy.

Another interesting general pattern from Table Fl (see under Appendix F) is that
students talked more about Results in their post-interviews. The overall number of Needs
statements at all levels tended to remain unchanged, while the number of Results
statements increased by 28. This suggests that students may be showing a greater
tendency to think about processes as involving elements that can be traced through time.

Overall, and given that I expected the number of each type of Data to first
decrease from the upper part of the Table (lower level Data) and then increase toward the
lower part of the same Table (higher level Data), these results suggest a pre-post change

that was marginal. Among other reasons, this could be due to the nature of the

73

comparison that I did---I included all of the 16 students’ data whether or not they
completed their pre-post-interviews. I will return to this point after a brief interpretation

of Table 8b below.

Table 8b: Overall (%) Pre/Post Warrant comparison

Warrant category Pre (A) Post (B) Change [B-A] (C)

 

Low level 86 67 -19 (24.8%)
Medium level 13 11 -2 (16.7%)
High level 48 59 11 (20.6%)

Table 8b shows comparisons between the total number of types of Warrants in the
pre-post-interviews from all the 16 students irrespective of whether they did all or part of
the interview. The upper part of the Table shows what I designated low level category of
Warrants (For specific examples of types of Warrants, see Appendix F2). In the middle
part of the Table, I designated as medium level category of Warrants. Finally, on the
lower part of the Table I include high level category of Warrants.

As in Data, 1 first expected a decrease in the number of the low level category of
Warrants (column C) and then an increase in the high level category toward the lower
part of the table on the same column in the pre-post-interviews. Unlike Data, all
categories of Warrants showed the expected trend: That is, a general increase from force-
dynamic types of Warrants at the top part to more scientifically verifiable ones at the
lower part of Table 8b. For shifts in specific types of Warrants (e. g. Analogies, Other

properties of Actors, & Special Properties of Enablers Energy) see Appendix F2.

74

Based on Table 8b, there was overall a small shift in the direction of using
Warrants that tend to align with scientific principles. This suggests that, to a limited
extent, students were beginning to lean more, in their responses about Carbon
Transforming Processes, toward scientific reasoning in their attempts to link Data to
Claims made. However, this shift is not attributable to any intervention in any way.
Rather, this could be attributed to, as earlier noted, reasons beyond this study. For
example, although participating teachers were expected to use the units we developed
(e. g. Systems and Scale, Plants, Animals and Decay) into which tools for reasoning were
built to teach, we know little about how they used them (see also under Comparison of
characteristics of elements at two different points in time-Research Question 2).

One factor that I wish to check here is whether the number of pre-post interviews
students did had anything to do with the results in Tables 8a and 8b. To do this, I
identified and excluded all data (pre-post) from students with partial participation (either
did pre or post)--a total of 7 students’ data. The remaining 9 students’ overall data is
presented in Tables 9a (pre-post Data), and 9b (pre-post Warrants). For details about pre-
post shifts in individual Data and Warrants, see Appendix F3 and F4 respectively. A

commentary is included after each of these tables.

Table 9a: Overall (W Pre/Post Data comparison with partial data excluded

 

 

 

 

lData category Pre (A) Post (B) Change
lB-Al (CL
Needs Results Sum Needs Results Sum Sum B -
Sum A
(%)

Low level 19 15 34 12 22 34 O (0.0%)
Medium level 21 4 25 19 11 30 5 (18.2%)
High level 32 15 47 41 44 85 38 (57.6%)

75

After excluding all data from students who had not completed pre-post interviews,
the trend in change (Table 9a, column C) in the overall number of types of Data basically
mirrors that of 8a. To illustrate, even after excluding partial data, Pre-post low level
category of Data (Other) showed no change. With regard to medium level category,
exclusion of partial data yielded unexpected overall increase by 5 (or 18.2%). Similar to
table 8a, these two categories of Data suggest that students’ responses at low levels of
achievement tend to include data that refer to, for instance, conditions and specific forms
of matter for other processes and that this tend to persist over time. This is with a likely
consequence of persistent force dynamic view of phenomena, in this case Tree Growing,
Flame Burning, and Car Running.

As in Table 8a, the most expected change occurred in the high level category, an
increase by 38 (or 57.6 %, column C). This possibly point to the idea that, over time,
students’ responses at high level category are characterized by reasoning that tends
toward scientific ways of thinking. Despite the expected change at the high level
category, the patterns from Table 9a remain fairly unclear. This is in the sense that
against this expected increase (column C), was an overall unclear expected decrease in
those categories designated lower level (for shifts in the number of various high level
types of Data see Appendix F, Table F3). This could be, among other reasons, due to the
way we asked the questions. I should note that at the time of conducting interviews, we
were engaged in an iterative development of the interview protocol. Thus, a possible
future study might include an examination into whether, compared to the old protocol,

the revised protocol generates same or different kinds of elements of argument, if at all.

76

Table 9b: Overall (%) Pre/Post Warrant comparison with partial data excluded

 

 

 

Warrant category Pre (A) Post (B) Change [B-A] (C)
Low level 63 59 -4 (6.6%)
Medium level 11 ll 0 (0.0%)
High level 34 56 22 (48.9%)

Unlike Data, excluding all data from students who had not completed pre-post
interviews showed a mixed trend in change (Table 9b) of the total types of Warrants they
used to link Data to their Claims about Tree Growing, Flame Burning, and Car Running.
That is, whereas I expected a decrease in the overall number of low level category of
Warrants in the pre-posts, this was by what I consider a little (i.e. by -4 or 6.6%, column
C). Additionally, there was overall, no change in the medium level category. While there
was an overall expected net change in the high level Warrants category of 22 (or 48.9%),
the overall little change of low level category of warrants in students’ responses about

Carbon Transforming Processes likely point to mixed results.
Summary

In brief, students seemed to provide different types of Data and Warrants, ranging
from low to high level, to support their accounts about the processes of Tree Growing,
F lame Burning, and Car Running. Whereas most less sophisticated students’ responses
utilized Data and Warrants that were more in the form of “obvious facts,” which may or
may not be empirically verifiable, a few more sophisticated students’ responses tended to
employ the types of elements that were more aligned with scientific principles in the

sense that these were increasingly elaborated during the interviews, sometimes with

77

Backing. Furthermore, other students’ arguments tended to utilize both lower and higher
level elements at the same time in their accounts of Carbon Transforming Processes.

Although a few students’ accounts included elements that align more with
scientific principles (e. g. EJ R) in their post-interviews, others’ accounts such as that of
TNC included similar elements (those they used in their pre- interviews) in their post-
interviews. Furthermore, compared to Data, students’ responses tend to include more of
low level Warrants at level 2 than they do at level 4.

On the one hand, this likely indicates that more sophisticated responses tend
toward what Toulmin (1958) calls Basic argument. This is in the sense that these
responses seem to follow more of the rules of scientific reasoning by including all the
elements of a Basic argument (Claim, Data, & Warrant). On the other hand, less
sophisticated responses tend to focus less on, from Toulmin’s viewpoint, elements of
Basic arguments. Illustrative of this is the idea that most of these responses are more
often than not characterized by Data and Warrants that hardly link to each other and/or to
Claims made. That is, although it may be true that some of these students’ responses may
include Data that are scientifically verifiable, these are often linked to Claims that may or
may not align with scientific principles of matter and/or energy. This is in addition to
using Warrants that neither relate to the Claims made nor the Data used. These patterns of
association among Data, Warrants, and Claims were used as the basis for the suggested
Argument Levels of Achievement (Table 7), which summarizes the kinds of Data and
Warrants associated with each level of achievement for Claims.

There were also patterns of change between students pre— and post-interviews.

Students generally increased the overall number of high-level category statements for

78

both Data and Warrants, while numbers of low level category statements showed mixed
patterns. Students’ responses also showed an increase in the number of times that Results
of processes were cited at all levels. Overall, this study’s findings suggest implications to

teaching and learning science discussed in the next chapter (Chapter 4).

79

Chapter 4: Discussion and Conclusion

In this Chapter, I (a) briefly discuss limitations of this study in relation to data
collection process in terms of intentions and likely unintended questions raised that are
beyond the purpose of this study, (b) present assumptions that I brought into this study
and (0) suggest contributions of this study to the field of science education in terms of
implications for research and practice and teaching and learning of science as well as

suggest future directions in research.
Limitations of the study

The data used for this study have important limitations. In the process of
interviewing respondents (see Appendix 1), we tried to develop an environment that
elicited students’ ideas about Carbon Transforming Processes. For example, we explained
to respondents that we were seeking their ideas about these processes to help us design
classroom tools/materials for use in teaching and learning science. At the same time, we
tried to establish a working relationship with respondents by stating that they were free to
ask questions at any time during the interview. However, there were limitations in the
data collection process including the following:

0 The only data source for this study was clinical interviews. Thus the study lacks
triangulation that would allow me to associate characteristics of students’
interview performances with deeper characteristics of their reasoning about
carbon-transforming processes.

0 Although I was interested in the analysis of pre-post interviews, I was not able to
specifically say how instruction affected interviews: That is, we knew little about

the specific instruction students received. As a consequence of this limitation for

80

Research Question 2, I only focused on documenting how students’ interview
responses changed at two points in time and not what changed them.

I used the Practices of Responsible Citizenship framework as part of the rationale
for the study, but I could not focus on how students’ arguments were connected to
other practices of environmental literacy---students were not really trying to
persuade the interviewer about issues regarding, for example, environmental
choices. Indeed, our protocol was not designed to elicit students’ ideas about
environmental choices, explicitly or otherwise. Thus, students may not have made
complete arguments that include environmental choices with the result that I
could not document their cultural practices in context.

Additionally, given that I was interested in the nature of arguments rather than
what caused them, I included only demographic information that I believe was
sufficient to respond to the research questions for this dissertation study. Thus, I
cannot address a question about patterns by demographic information, for
example gender and socioeconomic status, of participants.

These limitations notwithstanding, findings from this work can provide important

contributions to the field of science education some of which I briefly discuss next. As a

way of setting the stage for this discussion, I begin by pointing out some of the

assumptions that I brought into this study.

Assumptions

I focused this dissertation study on how students constructed arguments to support

claims made regarding transformations of matter and/or energy in three carbon-

transforrning processes: Tree growing, Flame Burning, and Car Running. I bound my

81

study within our large environmental literacy project and literature about argumentation
as inquiry. These contexts provided me with the opportunity to focus on developing a
Learning Progression perspective for argumentation. Based on this view, this dissertation
study was driven by a number of implicit assumptions associated with the field of science
education.

First was the idea that information from data analysis relating to argumentation
should be informative about the challenges students face in reasoning about phenomena
and therefore identify the likely supports they need as they work toward achieving
science literacy (e. g. NRC, 1996). In this study, the informative information comes from
data analysis about socio-ecological events for matter and energy transformation. This
assumption relates to the mastery of the principles of matter and energy as an educational
end-goal.

Second was the notion that working toward providing for students’ science
learning support relating to argumentation would better position them to utilize inquiry
practices in their science learning roles such as critiquing sources of information (e. g.
Collins et al., 1989; Duschl, et al., 2007) in classrooms.

A third assumption, closely related to the second, was that we need to understand
how students approach data-driven arguments about complex socio-ecological issues
such as global climate change. Embedded in this assumption was the hope that students
who use the principles of matter and energy in analyzing information or responding to
questions about current socio-ecological issues, such as global warming, would develop

into environmentally responsible citizens (e. g. Covitt et al., 2009).

82

Implications for Research and Practice

Based on the assumptions in the preceding subsection and borrowing from

Toulmin’s (1958) model of argumentation, I note that in spite of continued work on

Learning Progressions and therefore supports in student science learning, learners still

struggle to use the big ideas of matter and energy in ways that align with scientific ways

of thinking. To illustrate and with respect to Research Question 1, different students tend

to use different types of Data, Warrants, and sometimes Backing to support different

kinds of Claims. Data analysis shows students in three groups:

On the one hand, most students (e.g., see Tables 6a & 6b, Chapter 3) continue
to use the kinds of Data and Warrants that are in the low level of
sophistication category even after a period of introducing them to the concepts
of matter and energy transformation in Carbon Transforming Process. The
likely reason for this is that these are abstract concepts and therefore not easy
to grasp. This is especially evident at the low Claim level where students’
responses indicate that perhaps they view socio-ecological events in force-
dynamic ways (see e. g. Pinker, 2007) where actors (e. g. Trees) constantly
need enablers (e. g. air) to fulfill their natural tendencies (e. g. growing).

On the other hand, data analysis suggests that a few students (e. g. see Tables
6a & 6b) tend to use the kinds of Data and Warrants that are in the high level
category of sophistication alter a period of introducing them to the concept of
matter and energy transformation in Carbon Transforming Process. In
addition, one of the patterns fi'om data analysis shows that students talk more

about Results in their post-interviews than in their pre-interviews. Suggestive

83

here is the idea that students may be beginning to factor into their reasoning
about Carbon Transforming Processes elements of argument that involve
tracing matter and energy over time.

0 Some students used elements from both force-dynamic and model-based
reasoning. These students tend to use types of Data and types of Warrants
from medium level category of sophistication (for details, see c. g. discuss
under types of Data and types of Warrants).

At this point, I suggest implications for these findings about students’ arguments

related to science teaching and learning, as well as possible future directions for research.
Implications for science teaching and learning

A major goal of science education has long been to prepare all students to achieve
science literacy (e. g. NRC, 2007; National Science Education Standards, 1996). Findings
from this study suggest that achieving such goal remains challenging. This is on the basis
that large proportions of students’ arguments showed low levels of sophistication. This
finding is similar to findings in other literature on Learning Progressions (LP) which
suggest that most students still grapple with the problem of using reasoning that aligns
with scientific ways of thinking to explain socio-ecological events (see 6. g. Alonzo &
Steedle, 2008; Covitt et al., 2009; Jin & Anderson, 2008).

This study develops an empirically-based Argument Learning Progression
framework in terms of Levels of Achievement (see Table 7, Chapter 3). This work thus
complements other empirical studies on Learning Progressions (see e. g. Mohan, Chen, &
Anderson, 2008). Like other studies about Learning Progressions, this study is focused on

describing students’ development regarding specific practices. Knowledge from this

84

description can support designing and selecting both curricula and instructional tools that
better align with identified needs.

Other research has shown that most students struggle with defending claims they
make with evidence (see e. g. Sadler, 2004) in the scientific sense. This could be because
these students “are seldom supported in this scientific practice” (Krajcik & McNeil,
2009). Lack of student support perhaps accrues from teachers’ own struggles with
balancing promising data-driven instructional practices that emphasize a systematic way
of understanding phenomena with other demands of their workplace, such as well-
meaning but misguided standardized testing (see e. g. Huber & Moore, 2000) that shift
away from these practices. This study suggests ways to help students construct more
effective arguments and by extension, better assessment.

Like other recent studies about learning progressions (e. g. Gunckel, Covitt, &
Anderson; 2009; Krajcik & McNeil, 2009; Mohan, Chen, & Anderson, 2009), my study
is consistent with what Popharn (2007) calls less-is-more model. This is in the sense that
it can empirically “show how students build understanding of important ideas as they
move through the curriculum. . .or experience instruction...” (Stevens, Delgado &
Krajcik, 2010, p.708). Using argumentation as an instructional tool in classrooms can
help support students in developing skills (e. g. Kuhn, 1991) for constructing inquiry-
based arguments (Berland, & McNeil, 2009; Covitt et al., 2009; Gotwals et al., 2009) that
are both based on empirically verifiable data and connected to the claims made in line
with scientific principles. My proposed Argument Learning Progression therefore,
viewed as a tool, can complement our efforts in trying to provide resources for both

instruction and assessment of students’ work

85

This study can also support preservice teacher preparation and teacher
professional development that emphasizes important foundational ideas, in this case
matter and energy, and argumentation. Together with intended instruction, this can ensure
curriculum coherence (Stevens, Delgado & Krajcik, 2010) that is generally lacking across

time and disciplines (Kedisou & Roseman, 2002; in Stevens, Delgado & Krajcik, 2010).
Future directions in research

My study is my attempt to contribute directly to one other broad research goals
for our larger multiyear Environmental Science Literacy research program. We are
working to develop and validate a learning progression framework in which we describe
how learners can transition from intuitive or force-dynamic reasoning by elementary
students to a level of environmental science literacy needed by informed citizens.
Whereas other studies in the project describe our progress in a framework for accounts of
carbon-transforming processes (e.g., Mohan et al., 2009), my study extends this
framework to include students’ arguments. Thus, this study attempts to make connections
to our overall goal of continued promotion of our ideas about environmentally
responsible citizenship (for details about Practices of Responsible Citizenship, see e. g.,
Covitt et al., 2009).

This study can also contribute to our goal of developing assessment resources tied
to our common Learning Progression framework, including on-line and paper-and-pencil
written assessments and clinical interviews. These assessments provide an empirical basis
for our continued refinement of our framework as well as rich descriptions of the

knowledge and practice of diverse learners as they reason about phenomena.

86

This study raises important questions that were beyond its design. For example,
do students who support their accounts of individual processes with true arguments from
empirical evidence also support positions on environmental issues? If they cite well-
defined observational data and warrants in support of their accounts of tree growing, do
they also favor arguments about environmental issues that are supported with well-
defined observational data and warrants? When provided with alternative ideas that
challenge the Claims they make about Carbon Transforming Processes, do they also
appeal, if at all, to evidence to support those Claims? What is the nature of that evidence?

These are questions for a possible firture study.

87

Appendix A: Interview Protocol

Environmental Literacy Carbon Interview
FORM A

Please start by briefly introducing yourselves-«include the idea that you are a
member of Environmental Science Literacy Research Project from MSU. Then, briefly
explain the purpose of the interview: In our work, we seek students’ ideas about such
processes as tree growing, girl growth, girl running, dead tree decaying, flame burning,
car running, lamp lighting, and cross processes. Our goal is to use these ideas to design
classroom tools/materials for use in teaching and learning science. The purpose of this
interview, therefore, is to seek your help in terms of your ideas about some of these
processes. Please feel free to ask questions at any time during the interview.

Next, please write down the student’s names, grade (and age) here below---you
may ask the student to help you spell his/her names. At this point, you may proceed to the
interview items (Next Page).

Name Grade Age

 

 

The italicized questions are for higher level students.

It is possible that you may not be able to finish all the interview questions.

88

PLANT GROWTH

Tree Growing

 

 

 

A small tree was planted in a After 20 years it has grown into a big tree, weighing
meadow 500 lb more than when it was planted.

 

Actor: tree
Enablers: sunlight, water, soil, and air
Figure A. I : Oak tree
1. What does the tree need in order to grow?
2. You said that the tree needs [sunlight, water, soil, air] in order to grow. Follow up
probes about each enabler:

a. How does [the enabler] help the tree to grow?

b. What happens to [the enabler] inside the tree?

0. Is [the enabler] used up to help the tree to grow? Does it change into other
things inside the tree’s body? Or, do you think it will not change inside the
tree’s body?

d. Does the tree use [the enabler] for energy? How does that work?

3. F ollow-up probes on enablers not mentioned

a. Some other students have mentioned [other enabler]. Do you think [the other
enabler] is necessary for the tree growth?

b. [If yes, same probes as for other enablers]

c. [If no] Why not?

89

4. Scale

5. Matter

Do you think that the tree is made of cells? Why?
Do you also think that the tree is made of molecules? Why?
You said that the tree is made of both cells and molecules. How are the cells

and molecules related? What’s the connection?

Does the growing tree change the air? How does that happen?

The tree gets heavier as it grows. How does that happen?

Where do the increased materials come from?

Do you think the tree’s body can naturally create more and more materials?
Why?

Do you think the increased materials of the tree’s body are changed from
things outside of the tree? [If yes], how do these things change into the tree’s
body structure

If the student mentions glucose/starch/cellulous/carbohydrates, ask: Do you
think it contains carbon atoms? [If yes], where does the carbon atoms come
from?

[If the student talks about C 02—02 exchange, ask]: You said that the tree
needs Carbon dioxide and breath out oxygen. Where does the carbon atom

of C02 go?

6. Energy

a. Does the process of tree growing involve energy? [If yes], where does the

energy come from ?

90

b. Why do you think the things you mentioned have energy?

c. [If the student associates energy with sunlight, ask]: Where does the energy
of sunlight go? Is it used up? Does it change into other materials? Or, is it
still energy? Where is it?

d. Do you think the tree stores energy inside its body? If yes, where does the
tree store energy? In cells? In molecules? Where does that energy come
from ?

e. If students do not mention photosynthesis, ask: Is there any connections

between the things you mentioned and photosynthesis?

91

 

FLAME BURNING

Burnin' Candle

 

Actor: flame

Enablers: fuels (wax, wick, wood), air

Figure A.2: March and Flame

1. What does the flame need in order to keep burning?

2. You said that the flame needs [wax, wick, air, wood ...]. Follow up probes about

each enabler.

d.

How does [the enabler] help the flame to burn?

What happens to [the enabler] inside the flame?

Is [the enabler] used up? Does it change into other things? Or, do you
think it does not change?

Does the flame use [the enabler] for energy? How does that work?

3. Follow-up probes on enablers not mentioned

a.

b.

C.

4. Scale

Some other students have mentioned [other enabler]. Do you think [the
other enabler] is necessary for the flame to burn?

[If yes, same probes as for other enablers]

[If no] Why not?

92

a.

b.

5. Matter

a.

b.

6. Energy
a.

b.

Do you think that the flame is made of materials?

If yes, do you think the flame is made of molecules and atoms? Please

explain.

What change will happen to the match?

Do you think the match will lose weight? [If yes], where does it go? Is it
used up? Does it change into other things? Why?

What change will happen to the wax of the candle?

Do you think the candle will lose weight? [If yes], where does it go? Is it
used up? Does it change into other things? Why?

Does the event of flame burning change the air? How does that happen?
Do you think wax/Wood contain carbon atoms? [If yes], where do the

carbon atoms go when the flame is burning?

Does the process of flame running require energy?

If yes, where does the energy come fiom?

Why do you think the things you mentioned have energy?

[If the student associates energy with wood or wax, ask]: Where does the
energy of wood/wax go? Is it used up? Does it change into materials?
Or, is it still energy? Where is it?

Why do you feel warmth when the flame is burning? Do you think heat is

released from burning?

93

f [If yes], how is heat released? Do you think heat is created in
combustion, or do you think it is changed from other forms of energy in
combustion? Please explain.

g. If students do not mention combustion, ask: Is there any connections

between the things you mentioned and combustion?

94

CAR RUNNING

Car Runnin

 

Tom’s family wentto Chicago on vacation. When they came back,
Tom’s dad found that their car consumed 50 gallons of gasoline
for the trip.
Actor: Car
Enablers: gasoline, air
Figure A.3: Car running
1. What does the car need in order to carry the family to Chicago?
2. Why do people use gasoline instead of water to run their cars?
3. You said that the car needs [gasoline, air]. Follow up probes about each enabler:
a. How does gasoline/air help the car to run?
b. What happens to the gasoline/air inside the car when the car runs?
0. Does the car use gasoline/air for energy? How does that work?
(1. Is gasoline/air always necessary for car running? Why or why not?
4. Follow-up probes on enablers not mentioned
3. Some other students have mentioned gasoline/air. Do you think it is
necessary for car running?

b. [If yes, same probes as for other enablers]

c. [If no] Why not?

95

a. When your family arrives at Chicago, the gas tank is almost empty? Where
does the gasoline go?

b. Do you think the gasoline is used up? Or, does it change into other things?

c. Does the event of car running change the air? How does that happen?

d. Do you think gasoline contains carbon atoms? If yes, where do the carbon
atoms go when the gasoline is used by the car?

6. Energy

a. Does the process of car running require energy? If yes, where does the energy
come from?

b. Why do you think the things you mentioned have energy?

c. [If the student associates energy with gasoline, ask]: When the car stops,
where does the energy of gasoline go? Is it used up? Does it change into
materials? Or, is it still energy? Where is it?

(1. After the car runs for a while, the front part of the car will become very hot.
Why?

e. [If the student mentions heat, ask]: how is heat released?

fl You said that the gasoline is burning inside the car. Do you think heat is
created in burning, or do you think it is changed from something else?
Please explain.

g. If students do not mention combustion, ask: Is there any connections

between the things you mentioned and combustion?

96

Appendix B: Tools for Reasoning

Introduction

These tools were designed for teaching experiments at elementary, middle, and
high school levels focusing on carbon-transforming processes in socio-ecological systems
at multiple scales, including cellular and organismal metabolism, ecosystem energetics
and carbon cycling, carbon sequestration, and combustion of fossil fuels. These
processes: (a) create organic carbon (photosynthesis), (b) transform organic carbon
(biosynthesis, digestion, food webs, carbon sequestration), and (c) oxidize organic carbon
(cellular respiration, combustion). The primary cause of global climate change is the
current worldwide imbalance among these processes.

Our teaching goal is to support students as they move through a learning
progression leading to environmental science literacy—the capacity to understand and
participate in evidence-based discussions of socio-ecological systems and to make
informed decisions about appropriate actions and policies. As discussed in more detail in
the Appendix, our research shows that in order to develop environmental science literacy,
students must master three key principles: scale, matter, and energy. The tools for
reasoning are designed to embody those principles. As an example of the kinds of the
kinds of tools for reasoning we used, I have included the molecular models tool for
reasoning in the illustration below.

1. The Powers of I 0 T 001 embodies the principle of scale. Students can use this tool
to connect representations of systems and processes at multiple scales: atomic-
molecular, microscopic, macroscopic, and landscape scales.

2. The Process T 001 embodies the principles of conservation of energy and matter:

97

a. The top row of the tool embodies conservation and degradation of energy:
Students can use this part of the tool to describe how any process
involving chemical and/or physical change transforms energy without
changing the total amount of energy, including some energy that is
converted to waste heat.

b. The second row of the tool embodies conservation of matter. Younger
students can use this tool to identify solids, liquids, and gases as reactants
and products. More advanced students can use the tool more rigorously to
balance mass of reactants and products and to trace atoms through

processes.

3. Molecular models embody conservation of matter. Students can use this tool to

model how all carbon-transforming processes rearrange atoms into new molecules
without creating or destroying atoms.

We have designed these Tools for Reasoning to be flexible enough to use in every

lesson, including both lessons in our teaching modules and other lessons involving

carbon-transforming processes. The teaching experiments include the following modules:

1.

Systems and scale: Introductory module, designed to be used before any of the
other modules

Plants: Plant growth and metabolism

. Animals: Animal growth and metabolism

Decay and decomposers
Carbon cycling: Ecosystem scale

Human energy systems

98

An illustration of tools for reasoning

Tools for reasoning: Matter and Energy

Molecular Model Kits

   

02 molecule

l-IzO molecule

 

Figure B] : Water and Oxygen molecule

C02

1
molecule

  

Figure 3.2: Carbon dioxide molecule

99

 

Figure 8.3: Butane molecule

Butane (C4H10)

100

H H
'a
C: Cs
H H

Appendix C: Examples of Color coding

Table C1: Example Coding for Data

 

Transcript Procas Dat Commen Genera Specific Level_Explainin Pre/pos

 

 

 

 

 

 

s a t I Code Code g t
(D)
RWD: It l---Needs
needs Energy
water, TG D (NE);
sunlight Needs
and carbon Matter NE; NMS;NE
dioxide. (NM) NM S 2 Pre
SLP: Sun
and water. TG D 1 NM NEG; NMS 1.5
Pre
SLP: And Other
soil. TG D N0 NMO 1.5
needs
Pre
SLP:
Somewher
e to be TG D Condition NO 1.5
placed in
the ground. Pre
TNC: It
needs TG D 1 NM NMS 1
water. Pre
TNC: And
"k". TG D I NM NMS l
nutrients.
[0:02:05.6] Pre

101

Table C2: Example Coding (or Warrants

 

 

 

 

 

 

 

 

"/4"; 1111.”. r".

.. .. ,,
l/ll "'t

Transcript Pr \\ :ll'l':l ('omment (Iener Speril‘ie l,(‘\ el Pre/p
oc nt (\\) al (‘oile laxplainin ost
CS c\" "or code g:
5 Bar Lin
1: (lit
RWD: Itiust FB ll' 1 (see line lautol l‘ l pre
helps it hum 255) ogieal
heeause it it‘s (1)
\\ et oi‘ \\ et
material that
\\ multlviust
Slut) lllt‘ llt‘C.
RWD: lt-s TG \\' Rea‘tetzlnt.‘ llltltlL‘ Sl’lflX'l 3 pOSt
joined with lNeept for n
the etn'bon the term .\leeh
tlimitle to ”acute". anism
create sugar aligns with Matter
\Vhieh the. seieiitilie (MM)
plant uses for principles
loud.
RWD: TG \\' (ilueose is R \‘l Sl’lr\l 3 post
Glucose. The food
plants l'ootl.
RWD: TG \\' Reaettnits; .\l,‘\l Sl’l‘M 3 post
(’ai‘hon lixeept lttl'
tlio\itle (llltl the It‘l‘lll
sunlight and "et‘eate".
the water are aligns with
all eonihinetl seientitie
to create the pi‘iiieiples
glucose.
RWD: \\’he11 TG \\ lttzplietlu- \ll\l Sl’li\l 3 post
it burns the glueose
glucose to has enerm
Hui/iv ,i.f\_/rH‘/t/.
RWD: In it‘s TG \\ .\liun> .\ll\l Sl’lzkl 3' post
hotttlx in lll\C with
carbon to seientitie
t‘zit‘hon :tntl pi'ineiples
eni'lion
ll_\tll‘ttgtftl
l‘t‘!ltl\.
RWD: 1'// Me TG ll 1’ tim/i \ lift] (il l: 3 post
hum/s u/ the Nail (t )3
(mutt/w it l/Jtt
(lint/tile (llll/ film

 

102

Appendix D: Exemplar Worksheets for Reliability Checks

Given limited space (worksheet too large to fit allowed page specifications), I

split the large exemplar Table (on the exemplar worksheet) into three parts one each for

Tree Growing, Flame Burning, and Car Running.

Table D1: Exemplar color coding Tree Growing

Elements: Meanings & Examples

 

 

Pro Claim: Statement(s) Data: Visible Warrant:
ces students make about observation(s) about Universal premises
s how matter and/or CTPs, regarding a claim students make that
energy are involved in that students may make: link either one type
CTPs: Relate to hidden May include verbal of data and/or
mechanisms. observations- typically different types of
statements about needs of data to the claim
organisms or conditions regarding specified
for processes to occur CTPs.
and statements about
Examples include: visible results of
processes. Examples include:
Examples include:
TG I: ...So where does light I: what does a tree need I: So how does a

energy (water and C02)
go?

1. EJR: lt (energy) is
converted to a stored
energy) AND/OR

2. RKC: ...there would
be a chemical reaction...
light energy...it provides
heat and things.
AND/OR

3. AJB: ...it goes
through the tree...

in order to grow?

1. EJR:..water...,
nutrients... carbon
dioxide...sunlight
AND/OR

2. JMJ: ...sunlight, water
good soil AND/OR

3. NAC: Water,
sun...Oxygen AND/OR

tree use air (and
water)?

1. EJR: ...carhon
dioxide contains
nnilccules. atoms
2. SAM:

water. . .llltll\C it do
photoss nthesis
...carhott
tlioxitle_..gets used
as energy AND/OR
3. SLP: 'lhe water
is: like lltt.‘ l‘tltltl (for
the tree)

 

103

Table D2: Exemplar color coding (or Flame Burning

 

FB I: ...do you think the I: What does a flame need I: What is in wood

chemical energy still
exists somewhere or
changing...other types
of energy, or just burn
up?

1. ANW: It changes
into heat and light
energy

AND/OR

I: ...So where does the
lost material go?

2. RWD: lt’s
combined with the
burning oxygen and
creates C02. And,
anything lc/t turns into
a liquid

‘ND/OR

1: Ok. ...Why does the
flame need wax and
wood? What happens
to them?

3. ...wax and wood are
kind of like flarnes’
food... without it,
they’ll just die off

in order to keep burning?
1. ANW: It needs oxygen,
wood. .. AND/OR

I: ...So how about wax and
a wood... What happens to
them?

2. JAH: I think part ofthe
actual wick gets used up
and then the wax just kind
of melts and then reforms
later.... AND/OR

3. JMJ: It will disappear...

that makes it burn?
l. ANW: Wood has
chemical energy

AND/ OR

I: So how does
oxygen help the
flame to keep
burning?

2. RWD: It has high
energy bonds and its
flammable. ..

AND/ OR

I: ...So how about
wax and a wood...
What happens to
them?

3. JMJ: wax and
wood are kind of like
flames” food.

 

104

Table D3: Exemplar color codigg for Car Running

 

 

CR 1: ...So do you think a

car moving needs air
and gasoline? (What
happens to those?)
How does that work?
1. EJR: ...gasoline is
burned within the
engine...it is
converted to water
vapor and carbon
dioxide... AND/OR

I: So where does the
energy initially in the
gasoline go?

2. RWD: It runs
through the engine
and then is converted
to carbon dioxide

AND/ OR

3. DRH: Yeah... the
spark is used to start
the car ...makes the
gasoline and brings
the air and the spark
and the gas/rout
behind, the molecules
like makes an
explosion aml it
makes the car the
more

I: ...what does the car
need in order to carry the
family to Chicago?

1. EJR: It needs oxygen. It
needs a source of fuel,
which in the case of the
ear is going to be
gasoline... AND/OR

2. JAH: It needs oxygen
for the combustion,
gasoline, it needs a person
to drive it and I guess that
would be all AND/OR

3. DRH: ...the gas and
somebody controlling the
car.

I: ...what does the
car need in order to
carry the family to
Chicago? How does
that work?

1. EJR:
...gasoline...is
...comhination of
carbon and h_\*drt’igcn
molecules. [1 uses
the oxygen in the air
in the process? of
burning the

gasoline... AND/OR

I: So where does the
heat energy come
fi'om?

2. RWD: Some oi"
it‘s from the energy
from the gasoline.
And. the other part
ot'it is from the
o\_\gcn AND/OR
3. DRH: ...the gas,

that‘s the energy.

 

Note: All the parts should be viewed as one long Table with column headings Process,
Claim, Data, and Warrant (see headings for the first part, this Appendix).
In addition to the coding worksheet labeled AllData (see tables 3a and 3b,

Methods section) and the just presented three exemplars, I included the following

105

procedure for coding, first, a sample of two purposefully selected students’ responses
(one more sophisticated than the other).

0 On the worksheet labeled "AllData" (this workbook), and for each process
(Pre/Post) please follow this procedure to do reliability coding for these two
students: AH and J M] :

0 Begin by reading and identifying these elements of argument and accordingly
highlight them (see exemplar workbook for definitions & examples): Claim
(green); Data (blue); Warrant (pink); Backing (orange

0 Next, identify the specific types of Data you have highlighted (e. g. NES, &
RMS)---all possible types are provided under Data: If present, indicate by typing
in "1"; If absent, indicate by typing in "O"

0 Then identify the specific types of Warrant you have highlighted (e. g. A, SPEM
& SPEE)---all possible types are provided under Warrant: If present, indicate by
typing in "1"; If absent, indicate by typing in "0"

0 Finally, identify the types of principles, either matter (PCM) or energy (PCE)
conservation you have highlighted: If present, indicate by typing in "1 "; If absent,

indicate by typing in "0"

106

Appendix E: Proportions of Data and Warrant types

Lest Tables El and E2 seem like duplicates of Tables 6a and 6b respectively, I

wish to note that they supplement each other-the latter present overall category patterns

(low, medium, and high) for Data and Warrants, while the former present individual Data

and Warrant types. Thus, Tables 6a and 6b are developed from Tables E1 and E2.

Table E1: Percentage (%] ot gypes 01 Data at each Level of Claim

 

 

 

 

 

 

Cl” NO RE RM .\1-: \\1 to; 1m NF. .\'.\1 RE R.“
m 0 0 r; r; r; (1 s s s s
level

cot“ 0 1 2 n 1 t1 1 5 5 5 6
4 %of

Total 0.0 3.9 7.7 on s on 3.8 19.2 10.2 19.2 23.2

(26)

cot“ 6 3 11 4 u 1 1 9 21 11 20
3 %of

Total 6.3 3.0 11.5 4.: {>4 1.11 1.0 94 21.9 11.5 20.8

(96)

cot“ 32 5 23 11 M 5 s 6 36 8 14
2 %of 18

Total 7‘ 3.0 13.4 64 14 3.11 4.7 3.5 21.1 4.7 8.1

(171)

LOW Level \lctlium l e\ cl High Level

 

Note: Compared to levels 2, 3, and 4, most students used elements that were, overall, at level 2,

with only a few at level 4. Students used fewer elements at level 3 than they did at level 2. Color

coding is for purposes of identifying patterns.

From Table El above, there were a total of 26 different citations of Data that

students used to support Level 4 Claims about Tree Growing, Flame Burning, and Car

Running. Out of these, students used 5 (or 19.2 %) of Data type NES compared to 0 of

107

Data type NO. In contrast, out of a total of 171 different citations of Data to support

Level 2 Claims, students used 32 (or 18.7 %) NO compared to 6 (or 3.5%) NES.

Table E2: Percentage (%l of gypes of Warrants at each Level of Claim

 

 

 

 

 

 

Clai A T OE OE O (‘ SP.\ SPF. SPE PC PC
m M E A 1:: M E M E M
level
4 Coun 0 0 0 1 0 u u 4 6 l 1
t
%of 0.0 0.0 0.0 7.7 0.0 (H) 11.11 30.8 46.1 7.7 7.7
Total
(139
3 Coun 8 7 12 9 5 4 5 18 22 l 5
t
%of 8.3 7.3 12.5 9.4 5.2 4.1 5.2 18.8 23.0 1.0 5.2
Total
(96)
2 Coun 29 27 28 16 11 13 2 17 24 t) 0
t
%of 17. 16. 16.8 9.6 6.6 7s 1.: 10.1 14.3 0.0 0.0
Total 4 2
(167)
Low Level Medium 1 excl High Level

Like Table El above, Table E2 indicates that, overall, students tend to use
Warrants that align with scientific principles (Green highlight, level 4) to support their
accounts of Carbon Transforming Processes (CTPs). This is with the exception of
Backing (PCE & PCM) which levels 3 and 4 responses only tend to include in these
accounts. However, compared to less sophisticated responses (those at Level 2), more
sophisticated responses (those at level 4) are less likely to include low level types of
Warrants in accounts about Tree Growing, Flame Burning, and Car Running. This is
shown by generally fewer low and medium types of Warrants at level 4. Indeed, out of a
total of 13 types of Warrants, students used only 1 (or 7 .7 %) of the Warrant type Other

Enablers Energy (OEE) at this level. In contrast, to illustrate, students used 6 (or 46.2%)

108

of the Warrant type SPEM at the same level. That is, the proportions of low level types of
Warrants (Blue highlight) are hardly present at level 4. Similar to Data, the proportions of
the types of Warrants included in Table E2 tend to increase from upper left part of the
Table toward the upper right part of the same table.

Table E2 also suggests that although less sophisticated responses include some
high level Warrants (i.e. SPEE, & SPEM, Green highlight), they are more likely to be
characterized by low level Warrants (e. g. A & T, Blue highlight) in the accounts about

Carbon Transforming Processes (CTPs). This is overall indicated by a particularly higher

 

proportion of low level Warrants at achievement level 2 (Blue highlight) compared to
either levels 3 or 4. Additionally, in contrast to high level, the proportions of these types
of Warrants tend to, generally, decrease from lower left part of the Table toward the
lower right part of the same table.

Furthermore, with regard to level 3, Table E2 generally shows that students tend
to use both low and high level Warrants. Illustrative of this is the fact that, although the
proportions of high level Warrants at this level are higher (Green highlight) than those to
the left side (Blue highlight), all cells at this level are nearly equally represented at each
level of elements. Consequently, students’ use of Warrants at this level, overall, tends

toward average.

 

109

Appendix F: Pre-Post Comparisons

Table Flprovides comparisons between the numbers of types of Data in the pre-
post-interviews. This includes data from all the 16 students irrespective of whether they
did all or part of the interviews. The upper part of the Table (Blue highlight) shows what
I consider to be low level Data (NO, REO, & RMO), the middle part (Brown highlight)
medium level (NEG, NMG, REG, & RMG), and the lower part of the same Table (Green
highlight) shows higher level types of Data (N ES, NMS, RES, &RMS).

Table F1: Pre/Post Data comparison

Data type Pre Post Change [B-A]

 

(A) (B) (C)
Needs Other (NO) 23 15 -8
Results Energy Other (REO) 1 8 7
Results Matter Other
(RM0) 17 19 2
\t't;1l\ hilt!!!) (.ttlt'tttl 7 ,
t\l (.1 i 8 l
\t'r-rls \l.tlltt'(.t‘lttit.rl 4 -
(M111) 1‘; 14 -5
Results I mum (vt‘ltt'tnl - _
s 5 11
Results “21211! (urinal 7
1!; \l( .) 4 h _
Needs Energy Specific (N ES) 8 12 4
Needs Matter Specific
3 _
(NMS) ‘ 3 29 4
Results linerov Specific
h. g
(RES) - 19 14
Results Matter Spccrfic 13 2.7 14

(RMS)

The numbers in the cells in column C are similarly highlighted and these indicate

how each specific type of Data changed. For instance, except for NO discussed earlier,

110

results suggest that the number of lower level types of Data increased by 7 for RED, and
by 2 for RMO, which I view as substantial and marginal respectively. I expected a
decrease in the number of these types of Data. In addition to this, except for NMG in
which, as expected, there was a decrease by 5, there also was a marginal unexpected
increase in the number of Data types NEG by l and RMG by 2, both of which I consider
marginal. REG did not change.

The most expected change occurred in the Results Energy Specific (RES) and
Results Matter Specific (RMS) by 14 each. Given that I expected an increase of these
types of Data in column C on the lower part of 8a, there was an unexpected decrease by 4
for the Data type Needs Matter Specific (N MS). However, there also was an increase, as
expected, by the same number (4) in Needs Energy Specific (N ES). Another pattern that
is noticeable from Table 8a is that compared to Needs, students talked more about
Results at all levels. It suggests that maybe they were thinking more about the
consequences of the processes as well as their causes.

An interesting general pattern is that students talked more about Results in their
post-interviews. The number of Needs statements at all levels declined by 12 from pre to
post-interviews, while the number of Results statements increased by 41. This suggests
that students may be showing a greater tendency to think about processes as involving
elements of argument that can be traced over time.

Table F2 (below) shows comparisons between the numbers of types of Warrants
in the pre-post-interviews from all the 16 students irrespective of whether they did all or
part of the interview. The upper part of the Table (Blue highlight) shows what I

designated low level Warrants which include Analogies (A), Tautologies (T), Other

111

properties of Enablers Matter (OEM), Other properties of Enablers Energy (OEE), and
Other properties of Actors (OA). In the middle part of the Table (Brown highlight), l
designated as medium level these Warrants: Citation of Evidence (CE) and Special
Properties of Actors Matter (SPAM). Finally, on the lower part of the Table (Green
highlight) I include what 1 view as higher level types of Warrants. These are Special
Properties of Enablers Energy (SPEE), Special Properties of Enablers Matter (SPEM),
Principle of Conservation of Energy (PCB), and Principle of Conservation of Matter
(PCM).

Lable F 2: Prg/Post WrLrant comparison

 

Warrant type Pre Post Change [B-A]
(A) (B) (C)
Analogical (A) 20 17 -3
Tautological (T) 22 12 -10
Other properties of enablers Matter 18 22 2
(OEM)
Other properties of enablers Energy 18 8 -10
(OEE)
Other properties of Actors (OA) 8 8 0
(ilatintlol l \iilittt‘t'H I.) IN ‘7 '3
\jtet'ial l’r‘ojn 1111's ()1 \1‘ltil‘s \laltt'l‘ 3 i l
t‘~ lJ \ \l '1
Special Properties of Enablers Energy 29 29 0
(SPEE)
Special Properties of Enablers Matter 18 21 3
(SPEM)
Principle of conservation of Energy (1 2 2
(P(‘E)
Principle of conservation of Matter 1 7 6
(PCM)

As in Data, 1 first expected a decrease in the number of the lower level Warrants
(Blue highlight, column C) and then an increase in higher level toward the lower part of

112

the Table (Green highlight) in the pre-post-interviews. Unlike Data, all types of Warrants,
with the exception OEM that showed an unexpected small increase by 2, followed the
expected trend-a general increase from force-dynamic types of Warrants at the top part to
more scientifically verifiable ones at the lower part of Table F2.

Although the most expected decrease occurred in Tautologies and Other
properties of Enablers Energy, by 10 each, a small decrease of -3 each happened in
Analogies (A) and Citation of Evidence (CE). Two types of Warrants did not show any
change. These were Other properties of Actors (A0) and Special Properties of Enablers
Energy (SPEE). Furthermore, although I have included Principle of Conservation of
Energy (PCB) and Principle of Conservation of Matter (PCM) as types of Warrants, and
given that only a few of the students used them, this is for purposes of discussion. These
principles’ increased use to defend Claims made in the post-interview was limited to a
few students. The same thing also applied to CE, and SPAM. Based on Table F2, there
was overall a small shift in the direction of using Warrants that tend to align with

principled reasoning.

113

TabIeF 3 : Pre/Post Data comparison with gartial data excluded

 

Data Type Pre (A) Post (B) Change [B-A] (C)
Needs Other (NO) 19 12 -7
Results Energy Other (REO) 1 6 5
Results Matter Other (RM0) 14 16 2
Needs Energy (siencral (\E(.') 4 7 3
Needs .\1atter (ieneral (NMG) I7 12 -5
Results Energy (.‘eneral (REG) I 5 4
Results \latter General (RMG) 3 (w 3
Needs Energy Specific (\ ES) 6 12 6
Needs Matter Specific (NMS) 26 29 3
Results Energy Specific (RES) 5 18 13
Results Matter Specific (RMS) 10 26 16

Table F3 shows that, after excluding all data from students who had not
completed pre- and post interviews, the trend in change (column C) in the overall number
of types of Data basically mirrors that of F 1. However, this is with the exception of two
types of Data: RMG which changed fi'om O to 4, and NMS which showed more expected
change from -4 to 3. Thus, even with these two changes, the patterns from Table F3
remain fairly unclear. This is in the sense that despite an expected increase in the number
of various high level types of Data (Green highlight, column C), the expected decrease in

those designated lower level overall was not clear.

114

T‘able F4 : Pre/Post Warrant comgarison with partial data excluded

 

Warrant type Pre Post Change [B-A]
(A) (B) (C)
Analogical (A) l4 l4 0
Tautological (T) 15 12 -3
Other properties of enablers Matter 15 19 4
(OEM)
Other properties of enablers Energy 13 7 -6
(OEE)
Other properties of Actors (OA) 6 7 l
(imlinunl'l‘xiiltme(t If) 3 7 1
Spatial Properties olflh‘lms ,‘llLilltl' 3 l l
(s l' \ \ 1)
Special Properties of Enablers I3 20 7
Energy (SPEE)
Special Properties of Enablers 2(1 27 7
Matter (SPEM)
Principle of conservation of Energy (1 2 2
(PCE)
Principle of conservation of Matter 1 7 O
(PCM)

Table F4 shows that, although I expected a decrease in the number of lower level
Warrants in the pre-post-interviews, some such as Other properties of Enablers Matter
increased instead by 4 (column C). This is besides a small increase in Other properties of
Actors (OA), and no change in Analogies. While there was an expected net change in all
the high level Warrants such as Special Properties of Enablers Energy (by 7), the use of

lower level warrants by students in their arguments likely point to mixed results.

115

References

Alonzo, C. A., & Steedle, T. J. (2008). Developing and assessing a force and motion
learning progression. Science Education, 93 (3) p 389-421

American Association for the Advancement of Science. (1993). Benchmarks for science
literacy. New York: Oxford University Press.

Anderson, W. C., Onyancha, M. K., & Jin, H. (2009). Learning progressions for
environmental science literacy [PowerPoint slides]. Retrieved from Michigan
State University Web site: http://te955.wiki.educ.msu.edu/September+2l

Barton, C. A., Tan, E, & Rivet, A. (2008). Creating hybrid spaces for engaging school
science among urban middle school girls. American Educational Research
Journal, 45(1), 68-103.

Bell, P., Linn, M. (2000). Scientific arguments as artifacts: Designing for learning from
the web with KIE. International Journal of Science Education, 22(8) 797-817.

Berland, K. L. & Reiser J. B. (2009). Making sense of argumentation and explanation.
Science Education, 93(1) 26—55

Carlsen, W. S. (1997). Never ask a question if you don’t know the answer: The tension in
teaching between modeling scientific argument and maintaining law and order.
Journal of classroom interaction, 32, 14-23

Chi, T. H. M., (1997). Quantifying qualitative analysis of verbal data: a practical guide.
Journal of the Learning Sciences, 6(3) 271 -3 1 5

Clark, B. D. & Sampson, D. V. (2007). Personally seeded discussions to scaffold online
argumentation. International Science Education, 29(3) 253-277

Collins, A., Brown, S. J ., & Newman, E., S. (1989). Cognitive apprenticeship: Teaching
the crafts of reading, writing, and mathematics. Lawrence Erlbaum Associates:
Hillsdale, New Jersey

Covitt, A. B., Tan, E., Tsurusaki, K. B., & Anderson, W. C. (2009). Students’ Use of
Scientific knowledge and practices when making decisions in citizens’ roles.
Paper presented at the Annual Conference of the National Association for
Research in Science Teaching, Garden Grove, April 17 — 21, 2009

Driver, R., Newton, P., & Osborne, J. (2000). Establishing the norms of scientific
argumentation in science classrooms. Science Education, 84(3), 287 - 312

Duschl, R., Ellenbogan, K., & Erduran, S. (1999). Promoting argumentation in middle
school science classrooms: A project SEPIA evaluation. Paper presented at the
Annual Meeting of the National Association for Research in Science Teaching,
Boston.

Duschl, R., Schweingruber, H., & Shouse, A. (2007). Taking science to school: Learning
and teaching science in grades K-8. Washing ton, DC: National Academy Press.

116

Erduran, 8., Simon, S., & Osborne, J. (2004). TAPing into argumentation: developments
in the application of Toulmin’s argument pattern for studying science discourse.
Science Education, 88(6) 915 -933

Feest, U. (2005). Operationism in psychology: what the debate is about, what the debate
should be about. Journal of the History of the Behavioral Sciences, 41(2) 131-149

Gerbens-Leenes, W., Hoekstra, A. Y., & Van der Meer, H. T. (2009). The water footprint
of bioenergy. Proceedings of the National Academy of Sciences, 106 (25) 10219-
1 0223

Gotwals, W. A., Songer, B. N., & Bullard, L. (2009). Assessing students ’ progressing
abilities to construct scientific explanations. Paper presented at the Learning
Progressions in Science (LeaPS) Conference, June 2009, Iowa City, IA.

Gunckel, L. K., Covitt, A. B., & Anderson, W. C. (2009). Learning a secondary
discourse: Shifts from force-dynamic model-based reasoning in understanding

water in socio-ecological systems. Paper presented at the Learning Progressions
in Science (LeaPS) Conference, June 2009, Iowa City, IA.

Huber, R. A. & Moore, C. J. (2000). Educational reform through high stakes testing-don’t
go there. Science Educator, 9, 7-13.

J in, H., & Anderson, C. W. (2008). A Longitudinal Learning Progression for Energy in
Socio-ecological Systems. Paper presented at the 2008 National Association for
Research in Science Teaching.

J in, H., Zhan, L., & Anderson, W. C. (2009). Development and validation of assessments
for a learning progression on carbon transforming processes in socio-ecological
systems. Unpublished Manuscript, Michigan State University, East Lansing, MI.

Kilboum, B., (2006). The qualitative doctoral dissertation proposal. Teachers College
Record, 108(4) 529-576

Krajcik, J. & McNeill, L. K. (2009). Designing instructional materials to support
students in writing scientific explanations: Using evidence and reasoning across
the middle school years. Paper presented at the Annual Conference of the
National Association for Research in Science Teaching, Garden Grove, April 17 —
21, 2009

Kuhn, D. (1991). The skills of argument. New York: Cambridge University Press.

Latour, B. (1987). Science in action: How to follow scientists and engineers through
society. Cambridge, MA: Harvard University Press

Lee, H. S. & Songer, N. B. (2003). Making authentic science accessible to students.
International Journal of Science Education, 25 (1) p 1 - 26

Lemke, J. L. (1990). Talking Science: Language, learning and values. Norwood, NJ:
Ablex.

McNeil], L. K. (2009). Teachers use of curriculum to support students in writing
scientific arguments to explain phenomena. Science Education, 93(2) 233-268

117

Mercer, N., Dawes, L., Wegerif, R., & Sams, C. (2004). Reasoning as a scientist: Ways
of helping children to use language to learn science. British Educational Research
Journal, 30(3) 359 - 377.

Mistler-Jackson, M., & Songer, N. b. (2000). Student motivation and intemet technolohy:
Are students empowered to learn science? Journal of Research in Science
Teaching, 37(5) 459 - 479

Mohan, L., Chen, H., & Anderson, W. C. (2009). Developing a multi-year learning
progression for carbon cycling in socio-ecological systems. Journal of Research
in Science Teaching, 46(6) 675-698

National Research Council (1996). The national science education standards.
Washington, DC: National Academies Press.

Newton, P., Driver, R., & Osborne, J. (1999). The place of argumentation in the
pedagogy of school science. International Journal of Science Education, 21(5),
553-576.

Pinker, S. (2007). The stufl of thought: Language as a window into human nature. New
York: Viking Press

Popham, J. W. (2007). The lowdown on learning progression. Educational Leadership,
64, (7) 83-84.

Sadler, T. D. (2004). Informal reasoning regarding socioscientific issues: A Critical
review of research. Journal of Research in Science Teaching, 41(5) 513-536.

Sandoval, A. W. & Millwood, A. K. (2005). The quality of students’ use of evidence in
written scientific explanations. Cognition and Instruction, 23(1) 23-35

Smith, C. L, Wiser, M., Anderson, C. W., & Krajcik, J. (2006). Focus article:
Implications of Research on Children’s Learning for Standards and Assessment:
A proposed Learning Progression for Matter and the Atomic-Molecular Theory.
Measurement: Interdisciplinary Research and Perspectives, 4 (1&2), 1-98

Songer, N. B. Lee, H. S., & McDonald, S. (2003). Research towards an expanded
understanding of inquiry science beyond one idealized standard. Science
Education, 87 (4)490 — 516

Stevens, Y. S., Delgado, C. & Krajcik, S. J. (2010). Developing a hypothetical learning
progression for the nature of matter. Journal of Research in Science Teaching, 47
(6) 687-715

Tippett, C. (2009). Argumentation: The Language of Science. Journal of Elementary
Science Education, 21 (1) 17-25

Toulmin, S. (1958). The uses of argument. Cambridge, MA: Harvard University Press.

Wadhams, N. (2009). How a biofiiel 'Miracle' ruined Kenyan farmers. Retrieved October
5, 2009, from TIME: Website:
http://www.time.com/time/world/article/0,8599, 1 92753 8,00.html

White, B., & Frederiksen, J. R. (1998). Inquiry, modeling, and metacognition: Making
science accessible to all students. Cognition and Instruction, 6 (1) 3 — 118

118

Wilson, M. (2005). Constructing measures: An item response modeling approach.
Mahwah, NJ: Erlbaum.

Yore, L., & Treagust, D. (2006). Current realities and future possibilities: Language and
science literacy-empowering research and informing research. International
Journal of Science Education, 28(2-3), 291 -3 14.

Zembal-Saul, C. (2009). Learning to teach elementary school science as argument.
Science Education, 93 (4) 687 - 719

119

MICHIGAN STATE UNIVERSITY L18 I

"III III III III W
12 0

93 3220 8567

3