Models as Epistemic Artifacts for Scientific Reasoning in Science Education Research

Round 1

Reviewer 1 Report

This paper develops a philosophical issue about the use of models as artifactual approach, in scientific reasoning and science education. In the latter the scientific reasoning can be viewed as a question oriented activity, like a question/hypothesis-based investigation. Modeling is useful for this, and explicit discussion of models as artifacts is helpful since it fosters an understanding of science   by keeping the focus on both the interrogative, uncertain character of scientific reasoning. In addition, artifactual approach views any representational relationships between models and real-world targets as contingent scientific achievements, and prompts researchers and teachers to reflect on the assumptions they make about systems under study.

 

The authors have developed a grounded proposition about the use of modeling in teaching science.

The issue addressed here in not new, however, the paper has some strong points, which are based on the well-written and presented material. The literature review is rich and complete, presenting in a lucid way the relative philosophical theses and show the contemporary discussions that are taking place on the issue in question, and informs the reader about the underlying connection between philosophy and science education. Moreover, very relevant empirical studies are presented, demonstrating this link between epistemological issues and their impact in teaching chemistry.

 

Conclusively, in my opinion, the paper has a complete form, developing the central thesis through an articulated review and it is useful to scholars involved in science education. It would like to see it published in the present form.

Author Response

We are, as a matter of course, happy about the kind and constructive conclusion to recommend publishing our article as is. We especially would like to thank the reviewer regarding the effort to explicitly stating the reasons for the decision!

Reviewer 2 Report

The paper addresses aspects of model-based scientific reasoning in science education.

Perspectives from science education and from philosophy of science are presented and  another approach called model as artifacts for scientific reasoning is contrasted.

 

Although it becomes clear that much literature work has been done and that the authors are well informed in the field of models and modeling I cannot recommend to publish the paper in the actual version. In the following I will show with reference to concrete text passages the attitude with which the authors wrote the text. As it is not a matter of quick changes to address the following critique I speak out for a reject. This is a pity because the authors obviously are motivated and able to foster the discussion in the theoretical field. At the end of my review I will make suggestions for a revised or rather new version of the text. 

 

The Abstract shows the problem very directly:

3. 3. The “tends to” sentence follows an introductory sentence, and is a subjective statement without any relation to theoretical aspects. I expect (even in an abstract) that relevant theoretical underpinnings are mentioned as such. Starting from that, I would expect the description of a research gap with regard to e. g. empirical data concerning the relevant theory or the goal of your work, which might be to conduct a literature review or to analyze selected studies with regard to the regard theory. The inconsistency you found may than be the result. Just stating an approach is inconsistent and suggesting another without any kind of argumentation falls short (even in an abstract).

Epistemic artifacts as well as scientific reasoning should also be shortly introduced in the theory sentence of the abstract. The way you come up with „question-oriented understanding of scientific reasoning”, the relation between models and reasoning and your work and reasoning do not become clear.

 

The introduction entails not referenced subjective statements which cannot be “used” by the reader because links to theory or even theory-based criteria are missing.

In this sense a clear focus on a main goal of this paper is missing at the end of the introduction. What becomes clear is the relevance of aspects you raised, but as they are not linked to a structured theoretical background, this does not lead to a clear aim or question for the work you want to present in this paper.

In turn raised aspects could be collected, justified and systematized theoretically as a focus for an analyses or systematic review.

 

1. 1, l. 25 “rarely” and in l. 26 “is often”: references for the studies you have in mind here are needed.

 

Footnote 1: What you personally “prefer” is not the main point. The main point is e. g. how authors define the terms they use, or for what purpose they use the term in their work and whether they consider what we already know about the underlying construct.  

 

1. 2, l. 46-48 di.

 

1. 2, l. 48-51: Concrete pieces of evidence in terms of examples are needed for what was not layed out properly.

 

2. 2., l. 54: As you work on research, the perspective on science teachers is not relevant here nor helpful here. To find a clearer focus would be more important.

 

1. 2, l. 69-70: Your rely your interpretations on internal proofs, but still, external overarching criteria are missing in order to conduct a reliable analyses of papers in the field. This calls for a systematic criteria-based literature review.

 

Chapter 2 and chapter 2.1: I would expect a definition of scientific reasoning before links to other constructs are described (p. 2, l. 77). Model-based reasoning neither is defined in chapter 2.2. The concrete focus of the chapter does not become clear. Under the headline “scientific reasoning in general” the most important point seems to be the question of domain-generality and domain-specificity. But still, this is not a definition.

 

1. 3, l. 142: How can “astonishingly” be referenced by references 44 and 45?

 

2. 4, chapter 2.3: From a headline I expect that it is related to content/theory. Your headline leads to examples, but for what do these examples stand for?

 

1. 4, l. 146-153: “… our main concerns” again is subjective without any “objective” foundation. Studies and the related publications were selected for your analyses when they were “careful in articulating how they understand the notion of model…. “. This is not understandable for me as a selection criteria. For me, the problem addressed in l. 151 firstly is a problem of validity. This has to be taken into account before thinking about authors’ possible biases.

 

5. 5., l. 217 “… we do contest” sounds like redemption for the criticism. This has to be rephrased to communicate scientifically.

 

2. 7, chapter 2.3.3.: “Further empirical studies” sounds random.

 

1. 9, l. 434, 436-437 “… appears to be” is rather weak, is there a tension (which one concrete?) or not? “… usually unexplained” is related to “many authors”, which authors exactly explain and which do not?, “tends to lead” also is a weak formulation.

 

1. 9, l. 438: “Artifactual account” needs to be defined. For the reader it would be easier to understand if you introduced theoretical constructs right at the beginning of the theoretical part (after a clear goal was formulated at the end of the introduction) next to each other. Then you can derive relevant theoretical criteria, select studies in a well-founded manner for your review, describe the method of analyses and present the results followed by a discussion.

 

1. 9, l. 455-458: You „suspect“ – is speculative.

 

3. 10, Chapter 3.1. deals with the theoretical aspect of representation, it comes up late with regard to the title and the interpretations already made. This fosters my hint to have a theoretical right at the beginning defining and describing all relevant constructs as well as according empirical findings.

 

1. 10, l. 484-486: I do not understand your thought concerning model construction. Model construction and model application are parts of the modeling process, which is not necessarily meant physically.

 

1. 11, l. 533, 534 “… a model is a representation if it is used as such”. A model is loaded with the incorporated knowledge. The model object is any kind of representation, whereas the model is a model in thought, in a person’s mind, loaded with knowledge.

 

1. 12, l. 564-570: „... the artifactual approach to models builds directly on the idea that models are human-made objects, whose construction and use in scientific practices is the key to their epistemic value.” Many of the cited authors probably would agree on this sentence, but not exclusively with regard to the artifactual approach. Stachowiak up to Gere stress the role of the modeler, the subject, within the modeling process. I don’ t see the objection.

 

Chapter 4: You define Artifacts as human-made to access empirical and theoretical questions. I see no objections to other work.

 

1. 12, l. 612-616: Volterra example: „… that are not inherently tied to any specific target system, but are rather hypothetical systems constructed to study general theoretical questions”. Systems could also be biological phenomena. It should be checked how authors define phenomena in their work.

 

13. 13. l. 616-619: You state that the Lotka-Volterra model was „applied across different disciplines“ which is related to model application.

 

1. 16, l. 806-808: The differentiation between - the aim of modeling to present what currently is accepted as being the case - refers to the model of perspective and - the aim of modeling is to think about and test the consequences of what if something were the case?” – refers to the model for perspective. Make the objection explicit.

 

69. 16, l. 816-818: You claim to consider the cognitive dimension of induction, deduction, and abduction as mandatory in empirical studies. See reference 69.

 

The chapter “Conclusion” refers to the question-based understanding already mentioned in the abstract and is paraphrased with “question-oriented activity, i.e., question/hypotheses-based investigation”.

Question and hypotheses are used as synonyms here, and hypotheses-based might be the hypothetico-deductive approach. As this hypothetico-deductive approach is shared by many of the authors you cited and as it stresses the subject-related idea of modeling I still do not see the objection.

 

The implementation of the artifactual approach is motivated by supposed weaknesses in other authors’ works which is not the scientific way to present innovative ideas.

 

I see two options for the authors to improve their argumentation:

 

1. Derive your approach out of theory and thus, lack in theoretical knowledge, and discuss them with regard to other work with regard to a priori-chosen criteria.
2. Conduct a systematic literature review with a well-founded selection of papers and applying overarching theoretical criteria.

 

The paper includes relevant literature and elaborates deeply on models and modeling.

My main critique is the subjective and generalizing handling of other work instead of presenting a theory-driven approach to analyze other work. In the whole paper a more proper differentiation between constructs and own reflections and interpretations is needed. Reflections and interpretations should be motivated by more general criteria. I wonder why the idea for the paper wasn’t realized in a literature review.

 

 

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Thank you for this interesting paper. Overall, I agree with the arguments outlined and find such an article long overdue. I have a few suggestions I recommend to consider for further refinement:

I found the discussion of models and representations reasonable. Although, you intensively referred to philosophical literature you might find the following source helpful as the author (unfortunately, rather short) discusses this topic as well: Hoyningen-Huene (2013). Systematicity. The nature of science. Oxford University Press.  (i.e., chapter 3.9).

I understand the critical evaluation of the empirical studies in science education. However, the selection of the studies seems to be a bit random. I strongly suggest outlining why you have chosen these studies to present in more detail. For example, how do you define "well-received empirical studies" (p. 2, line 68)? Also, to what extent are the presented studies exemplary for science education studies as a whole? What, if any, are the criteria used to determine this?

In chapter 2.1 you refer to "domain-generality" and "domain-specificity". I do not see the reason for this with regard to the following text. Please elaborate more on the domain-generality later on in your paper and connect it to chapter 2.1.

Please consider changing the structure of the article by presenting the philosophical references (chapter 3) first and science education studies second. In this way, the problems that arise in science education studies can be presented in a more differentiated way and with direct reference to the philosophical explanations.

You are referring to the Lotka-Volterra model but are inconsistent with the terminology used (p. 14). Although your paper is already quite long regarding the philosophical background, I suggest you are elaborating on the understanding of terms such as "law" and "principle". This is not trivial, especially when thinking about science education. To what extent can these terms be equated with each other?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

• I agree that papers written from a philosophical point of view can be inspiring and can enrich approaches from science education by broadening the theoretical background and by enabling reflections from other perspectives.
• Regardless of the perspective from which one is looking at models and modeling (philosophy, science education), I ask the authors to comment on all hints as part of our community’s convention and not only “to specific comments that we found valuable”. In fact the authors responded to 8 out of 27 hints from the review.
• The paper is written for a Special Issue called Scientific Reasoning in Science Education: From Global Measures to Fine-Grained Descriptions of Students’ Competencies”. This Issue is part of Education Sciences publishing Articles: Original research manuscripts and reviews(https://www.mdpi.com/journal/education/instructions)
• There is no “misunderstanding regarding the approach and scope of our article”, there is a need to fit the aim and scope of the journal and to fulfill the instructions for authors in order to discuss different scientific perspectives within this well-founded frame.
• In general, the paper is characterized by an abundance of references (which was appreciated in the first review), and yet there is a lack of citations in places at the same time (e. g. l. 40-44).
• As the structure of the paper is questionable for me (which was already commented in the first review), I ask for making transparent the chronology of the parts of the text and the line of argumentation right in the introduction in order to be transparent to the reader.
• What is “the representational approach to model-based reasoning” you refer to in l. 40?
• What is “the mainstream philosophical discussion of modeling” you talk about in line 41?
• I am interested in your approach and your ideas and you should communicate these by referencing such statements.
• Describe the artifactual approach, respectively the question-oriented character of modeling in depth and compare this with the ideas from science education studies related to concrete criteria (which was already commented in the first review).
• In which studies do you find “that the representational conception of modeling often features in them side-by-side with the notion of models as tools? The use of “often” without any reference is not helpful (l. 44).
• In l. 51 you point out that “it turns out to be highly problematic in practice”, give some explanation or evidence for this statement.
• The following paragraph (l. 52-65) gives explanations or even criteria as a ground for your analyses, elaborate this more in depth and enrich it with references.
• Following e.g. Mahr, the perspectives of model construction and model application as well as the perspective of model being are inherent in models from a theoretical perspective. When it comes to modeling processes conducted by a modeler, a subject, the purpose of the model determines the use of the model. And this might be representational e.g. to communicate the model’s content or tool-like e.g. to get new insight. Referring to the subject or to the purpose of models speaks for not regarding the representational conception side-by-side with models as tools. Is the purpose of models the question at hand in your perspective? For me it would be benefitial to get to know how the selected exemplary science education studies deal with these points (your chapter 4.1. and 4.3.).
• Mahr also calls for a differentiation between the model and the model object which you contest. This is an interesting point and indeed people from science education might get new insight when you elaborate this point more deeply.
• I expect explanations like the last four answers to my review within the paper as part of a valuable discussion.
• In l. 137 you state that the understanding of models in science … and science education … remains astonishlingly …. Having checked one of the references 43/44, I find: „... erstaunt es möglicherweise, dass allgemeingültige Klassifikationssysteme für Modelle in fachübergreifenden Diskursen nicht vorliegen.“ Thus, the astonishing point is related to the classification of models, not to the understanding of models.
• Furthermore I ask for clarification how the considered exemplary empirical and theoretical studies from the field of science education research were selected (l. 70) for arguing “for the artifactual approach to models”.
• In chapter 2.3.1. you reflect on Schwarz et al.’s study, in chapter 2.3.2. you give a second example from Reith & Nehring, and in chapter 2.3.3. you reflect about “further studies on modeling” which seems unstructured and very random as the choice is not explained and justified. Schwarz at al. and Reith & Nehring are discussed in detail, the others are discussed in single paragraphs each within one chapter.
• As the reader of your text I would expect an explanation for the choice of papers. Furthermore, I expect to get information about the way you analyzed the studies.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

Thank you very much for your explanations, which helped me to understand your approach from the philosophical perspective a bit more. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

My comments are in the attached file.

Comments for author File: Comments.pdf

Reviewer 2 Report

I think it is better to send this manuscript to another journal because it is not very interesting to the readers of this journal.

Reviewer 3 Report

This is a research paper with specific hypotheses and empirical data analysis. It is rather a position paper regarding the role of model in scientific reasoning. It uses insights from Philosophy of Science. It initiates dialogue between science education and philosophy of science.  In the article, by presenting studies from chemistry education research, it is argues and the identification of modeling with representation is wrought with theoretical and empirical challenges. The challenge concerns also science education actually, where treating models as tools is suggested rather than as uniquely representations of real-world systems is proposed. Some illustrative examples from science education explicates the thesis of the paper.

 

Overall the paper is a well-written philosophically orientated essay focusing on models that science uses to represent reality and on the scientific reasoning based on them. The content is well informed by the relevant and huge literature. While the core idea of implementing models in developing scientific reasoning in science education is not new, the approach of this paper brings a different and deeper explication about this relationship. In my opinion the paper brings something interest aspects of the links between science education and philosophy of science as far as the models-the representations of the reality and their function in reasoning.

 

There are some concepts and expressions such as “erotetic” which I think do not fit in such analysis. An anthropomorphic approach, I think, does not help to understanding what the issues are about. However, OK , it is the author(s) choice.

 

 

 

I am positive in seeing it published in this form, however, I would suggest some changes in the Abstract. It should be improved by summarizing the main points and positions and conclusions of the paper.  

Reviewer 4 Report

You find all comments in the attached word-document.

Comments for author File: Comments.pdf