Doctor of Philosophy, The Ohio State University, 2023, Philosophy

This dissertation is about what is called a conception of scientific explanation. A conception of scientific explanation concerns some ontological questions about scientific explanation. Two related but distinct questions have been investigated: (1) What type of entity is an explanans? (2) In virtue of what, is a thing of that type explanatory? There are two competing conceptions of scientific explanation. One is the ontic conception and the other is the non-ontic conception. A conception is ontic if it says that (a) a type of thing that explains is primarily a thing in the world that is responsible for a target of explanation (ontic explanation) and (b) a type of entity explains if and only if it is either ontic explanation or accurately represents the relevant ontic explanation (ontic determination). A conception is non-ontic if it denies either (a) or (b) or both in one way or the other. The aim of this thesis is to propose and defend a new non-ontic conception. The three chapters are designed to proceed toward the said aim. In Chapter 1, I challenge claim (a) of the ontic conception, i.e., the existence of an ontic explanation. I examine the existing reasons to believe in ontic explanations and argue against them. The conclusion of this chapter is tentative, for all it shows is that no good reason has yet been proposed to support our commitment to an ontic explanation. Still, the conclusion is strong enough to lend support to the non-ontic conception denying ontic explanations in that it is better not to posit something if not necessary for the ontological parsimony. In Chapter 2, I deal with claim (b) of the ontic conception, i.e., the ontic determination. To this end, I examine Angela Potochnik's non-ontic view, according to which whether something is explanatory is determined not only by what the relevant part of the world is like but also by some cognitive factor of those seeking an explanation. I agree with Potochnik in that some cognitive facto (open full item for complete abstract)

Committee: Christopher Pincock (Advisor); Richard Samuels (Committee Member); Stewart Shapiro (Committee Member) Subjects: Philosophy of Science

MA, University of Cincinnati, 2002, Arts and Sciences : Philosophy

This thesis is devoted to examining Leda Cosmides and John Tooby's use of evolutionary psychology as a heuristic framework for explaining human social behavior. Cosmides and Tooby are among the most vocal advocates of a now popular version of evolutionary psychology. They argue that the functional complexity of human reasoning can be best explained within the framework of adaptationism and that knowledge of the evolutionary environment of adaptiveness is essential to a scientifically satisfying explanation for why humans behave as they do. I first discuss the design logic of evolutionary psychology, and the methodology Cosmides and Tooby use. I then discuss the consequences of not adhering to standard scientific practice and whether Cosmides and Tooby's adhere to standard scientific practice in developing and testing their models.

Committee: Robert Richardson (Advisor) Subjects: Religion, Philosophy of

Doctor of Philosophy, The Ohio State University, 2006, Philosophy

The dissertation examines two putative explanations from statistical mechanics with the aim of understanding the nature and role of idealizations in those accounts, namely, the Yang-Lee account of phase transitions and the Boltzmannian account of irreversible behavior. Like most explanations in physics, these accounts involve idealizations to some extent. Many idealized explanations hold out the hope that the idealizations can be removed or eliminated with further work. However, the idealizations that occur in the accounts of phase transitions and irreversibility are ineliminable. The only way (in principle) to obtain a description – let alone an explanation – of these phenomena is to invoke various idealizing assumptions. Ineliminably idealized explanations are not well-understood from a philosophical point of view. Indeed, most philosophers of science would probably hold that no idealizations are ineliminable. The dissertation argues that this view is mistaken, showing where and why extant accounts of idealization miss this fact by distinguishing the widely-accepted understanding of idealizations as falsehoods from a novel understanding of idealizations as abstractions. As abstractions, idealizations are devices for ignoring certain details about the real world. The dissertation argues that ineliminable idealizations cannot be falsehoods, and that they should be understood as abstractions. The dissertation also examines the confirmation of idealized hypotheses and their role as guides to what the world is like. At least some idealized hypotheses have some degree of confirmation; and less idealized hypotheses tend to be better confirmed than their more idealized counterparts. If idealizations are falsehoods, Bayesian confirmation theory seems unable to obtain these results, because it lacks a way of defining the prior probabilities of idealized hypotheses. If idealizations are abstractions, however, idealized hypotheses about a system are incomplete claims that omi (open full item for complete abstract)

Committee: Neil Tennant (Advisor) Subjects: Philosophy; Physics, General

PhD, University of Cincinnati, 2021, Arts and Sciences: Philosophy

Understanding a normative concept like oppression requires attention to not only its harms but also the causes of those harms. In other words, a complete understanding of such a concept requires a proper causal explanation. This causal explanation can also inform and constrain our moral response to such harms. Therefore, the conceptual explanatory framework that we use to inform our moral diagnosis and our moral response becomes significant. The first goal of this dissertation is to propose complexity theory as the proper framework for not only explaining a social phenomenon like oppression but also understanding the proper sites for social change. The second goal of this dissertation is to answer three interrelated questions about how we should respond, morally, to a chronic and complex social problem like racial or gender inequality: (1) Why do the current interventions to address these problems fail? (2) Do social movements play any unique role in addressing these problems? (3) What is our individual responsibility to participate in social movements? In response, I argue that the explanatory frameworks that we choose to understand the cause(s) of social problems can be the source of the inadequacy of our intervention. I argue that a proper social and moral intervention needs to capture the complex and dynamic nature of the social world. I also show that changing the explanatory framework allows us to see the unique role social movements play in making effective and sustainable social change possible. Finally, I conclude supporting such movements is a moral imperative.

Committee: Vanessa Carbonell Ph.D. (Committee Chair); Anthony Chemero Ph.D. (Committee Member); Amy Lind Ph.D. (Committee Member); Angela Potochnik Ph.D. (Committee Member) Subjects: Philosophy

Doctor of Philosophy, The Ohio State University, 1970, Graduate School

Committee: Not Provided (Other) Subjects: Philosophy

PHD, Kent State University, 2016, College of Education, Health and Human Services / School of Lifespan Development and Educational Sciences

The purpose of this quantitative study examined the effects a classroom blog had on student performance in the area of conceptual and procedural understandings of fractions. Specifically, the study examined the effects of self-explaining with a peer (explain, justify, and argumentation) to the solving of traditional paper pencil mathematical tasks alone (solving on your own). The experimental groups (i.e. face-to-face and blog groups) solved identical mathematical tasks to the traditional alone group by explaining their solution through justification with evidence from the task by self-explaining with peers. Both experimental groups engaged in mathematical discourse by explaining and justifying their understandings, as well as critiquing and arguing the thinking of other student responses through self-explaining with peers; however, one group used a multimedia tool. This quasi-experimental design study further explored how interactive and constructive mathematical discourse (i.e. explanation, justification, and argumentation) through a classroom blog supported student performance of fifth-grade students on conceptual and procedural fraction knowledge and the retention of this knowledge over time. To measure the change in student performance, a pretest- posttest, and delayed posttest was administered to measure the conceptual and procedural knowledge of fractions. Participants included 134 fifth grade students, ages 9-11 years old. Data collection was analyzed using repeated measures ANOVA with one between –subjects factor.

Committee: Bradley Morris PHD (Committee Co-Chair); Richard Ferdig PHD (Committee Co-Chair); Karl Kosko PHD (Committee Member) Subjects: Educational Psychology; Educational Technology; Elementary Education; Mathematics Education; Middle School Education

Doctor of Philosophy, The Ohio State University, 2015, Philosophy

Philosophers of mind and cognitive science have recently gravitated towards a new mechanistic approach to constitutive explanation, and an interventionist approach to causation and causal explanation. In this dissertation I discuss the implications of these new approaches for four issues in the philosophy of mind and cognitive science: Mental causation, the nature of explanation, the extended mind hypothesis, and the dynamicist approach to explanations in cognitive science. I argue that the interventionist account of causation can be used to solve the problem of causal exclusion for non-reductive physicalist approaches to mental causation. However, in order to do so I propose a new improved version of interventionism which rules out the possibility of a certain kind of overdetermination, thereby showing that at most only one of the two different interventionist solutions to the exclusion problem is indeed a viable solution on behalf of non-reductive physicalism- namely, the interventionist proportionality argument. Next, I argue that the ontic and epistemic construals of mechanistic explanations should be seen as reflecting different (though related) legitimate senses of explanation, and there is no reason to think that the different senses of explanation are in conflict with one another. Next, I provide a novel argument against the extended mind hypothesis, employing only premises that proponents of the extended mind hypothesis (including the New Mechanists) have independent reasons to accept. I conclude that while the mutual manipulability account of constitutive relevance- embraced by the New Mechanists- may be useful for demarcating the boundaries of mechanisms studied in many of the sciences, it cannot be used to support the extended mind hypothesis. Finally, I show that two arguments commonly made in the debate between Dynamicists and the New Mechanists are mistaken, and I conclude that there is- or need be- far less theoretical disagreement between Dynamicist (open full item for complete abstract)

Committee: Richard Samuels (Advisor) Subjects: Philosophy

PhD, University of Cincinnati, 2015, Arts and Sciences: Philosophy

Mechanistic frameworks of investigation and explanation dominate the cognitive, neural, and psychological sciences. In this dissertation, I argue that mechanistic frameworks cannot, in principle, explain some kinds of cognition. In its place, I argue that complexity science has methods and theories more appropriate for investigating and explaining some cognitive phenomena. I begin with an examination of the term `cognition.' I defend the idea that “cognition” has been a moving target of investigation in the relevant sciences. As such it is not historically true that there has been a thoroughly entrenched and agreed upon conception of “cognition.” Next, I take up mechanistic frameworks. Although `mechanism' is an umbrella term for a set of loosely related characteristics, there are common features: linearity, localization, and component dominance. I then describe complexity science, with emphasis on its utilization of dynamical systems modeling. Next, I discuss two phenomena that typically fall under the purview of complexity science: nonlinearity and interaction dominance. A complexity science framework guided by the theory of self-organized criticality and utilizing the methods of dynamical systems modeling can surmount a number of challenges that face mechanistic frameworks when investigating some kinds of cognition. The first challenge is epistemic and concerns the inadequacy of mechanistic frameworks to facilitate the comprehensibility of massive amounts of data across various scales and areas of inquiry. I argue that complexity science is more appropriate for making big data comprehensible when investigating cognition, particularly across disciplines. I demonstrate this via an approach called nested dynamical modeling (NDM). NDM can facilitate comprehensibility of large amounts of data obtained from various scales of investigation by eliminating irrelevant degrees of freedom of that system as relates to the target of investigation. The second shortcomi (open full item for complete abstract)

Committee: Anthony Chemero Ph.D. (Committee Chair); Rick Dale Ph.D. (Committee Member); Valerie Hardcastle Ph.D. (Committee Member); Robert Richardson Ph.D. (Committee Member) Subjects: Philosophy

PhD, University of Cincinnati, 2015, Arts and Sciences: Philosophy

I argue that philosophical accounts of scientific explanation appear to agree that identification of constraints is a significant feature of scientific explanation. Moreover, scientific explanations are evaluated according to how well they facilitate human prediction, manipulation and understanding of a given phenomena. The constraints identified in an explanation may be due to the physical states and structures as they are observed, as in the speed of light or the mass of the coffee mug on the table. These are physical constraints. Constraints also depend on the choices and perspectives of the individuals or communities producing and consuming the explanation. These latter constraints I refer to as framing constraints. Framing constraints include the choice to observe a biological organism within a particular eco-system as well as the choice to explain in the context of the observable universe. Ultimately, both physical constraints and framing constraints are not distinctive categories but extremes on a continuum. Complex dynamical systems theory provides a framework for characterizing and understanding increases in system order in the context of certain constraints. Increases in system order entail increases in the observed correlations of spatial, temporal, or energetic features as represented by variations in a system's degrees of freedom. I argue that these increases in correlation length provide a basis for identifying characteristic scales of a system of interest that are larger than the scale defined in terms of the system's smallest components. In the context of scientific explanation, increases in order also result in the elimination of smallest scale degrees of freedom and their corresponding constraints. When the smallest scale degrees of freedom are eliminated from an explanation, the explanation is non-reductive. Given the proposed account of explanation in terms of constraints and insights regarding scale in terms of complex dynamical systems theo (open full item for complete abstract)

Committee: Anthony Chemero Ph.D. (Committee Chair); Jessica Wilson Ph.D. (Committee Member); Thomas Polger Ph.D. (Committee Member); Angela Potochinik Ph.D. (Committee Member) Subjects: Philosophy of Science

Master of Arts, Miami University, 2013, Psychology

The goal of this study was to utilize an Intelligent Tutoring System (ITS), AutoTutor Lite, in comparing different methods of learning for an ill-structured domain via interaction, through self-explanation or argumentation. One-hundred twenty participants were assigned to one of four conditions: 1) didactic-explanation, 2) didactic-argument, 3) didactic-no interaction, or 4) no didactic, no interaction. All 3 didactic groups performed significantly better than the control on the multiple-choice test. The didactic-explanation group outperformed the control group, but not other didactic groups on the photography advice measure. Significant differences were found for written-type scenarios for photography advice measures, with the didactic-explanation group and the didactic-argument group outperforming the control group, but not the didactic only group, nor each other. For the photography judgment task, the didactic-no interaction group outperformed the control on the composition dimension and outperformed the didactic-explanation, which outperformed the control on the shutter speed dimension.

Committee: Christopher Wolfe (Advisor); Yvette Harris (Committee Member); Doris Bergen (Committee Member) Subjects: Psychology

Doctor of Philosophy, The Ohio State University, 2013, EDU Teaching and Learning

Assessment is a key element in the process of science education teaching and research. Understanding sources of performance bias in science assessment is a major challenge for science education reforms. Prior research has documented several limitations of instrument types on the measurement of students' scientific knowledge (Liu et al., 2011; Messick, 1995; Popham, 2010). Furthermore, a large body of work has been devoted to reducing assessment biases that distort inferences about students' science understanding, particularly in multiple-choice [MC] instruments. Despite the above documented biases, much has yet to be determined for constructed response [CR] assessments in biology and their use for evaluating students' conceptual understanding of scientific practices (such as explanation). Understanding differences in science achievement provides important insights into whether science curricula and/or assessments are valid representations of student abilities. Using the integrative framework put forth by the National Research Council (2012), this dissertation aimed to explore whether assessment biases occur for assessment practices intended to measure students' conceptual understanding and proficiency in scientific practices. Using a large corpus of undergraduate biology students' explanations, three studies were conducted to examine whether known biases of MC instruments were also apparent in a CR instrument designed to assess students' explanatory practice and understanding of evolutionary change (ACORNS: Assessment of COntextual Reasoning about Natural Selection). The first study investigated the challenge of interpreting and scoring lexically ambiguous language in CR answers. The incorporation of `multivalent' terms into scientific discourse practices often results in statements or explanations that are difficult to interpret and can produce faulty inferences about student knowledge. The results of this study indicate that many undergraduate biology majo (open full item for complete abstract)

Committee: Ross Nehm (Advisor); David Haury (Committee Member); Lin Ding (Committee Member) Subjects: Science Education

PhD, University of Cincinnati, 2007, Arts and Sciences : Philosophy

The new mechanistic philosophy, which is comprised of conceptions of mechanism due to Glennan, Bechtel and the team of Machamer, Darden, and Craver, promises to account for a wide variety of natural mechanisms. I examine three cases of mechanistic, ecological explanation that represent three different classes of ecological situations: biodiversity mechanisms responsible for ecosystem functioning; the theory of island biogeography; and the experimental work of Gause on the Lotka-Volterra equations of population dynamics. I argue that the new mechanistic philosophy fails to characterize the componential, the causal, and the organizational aspects of ecological mechanisms that are the focus of these mechanistic explanations. In light of Woodward's manipulability account of causation and Levins' loop analysis, I argue that ecological mechanisms are insensitive networks of invariant causal relationships. The networks comprise organisms and causally relevant environmental factors, while organization lacks the central role it plays in the new mechanistic philosophy. Depending on the phenomenon under scrutiny, the network can comprise more or fewer components and interactions. Thinking in terms of invariant causal relationships captures both the individual-level and the population-level causal relationships. Graphs represent the insensitive networks of causal relationships at both levels. I show how my proposal characterizes the biodiversity mechanisms of ecosystem functioning, the mechanisms responsible for the species-area curves of the island biogeography, and the mechanisms that underlie the dynamics of competition and of predation.

Committee: Dr. Robert Richardson (Advisor) Subjects: Philosophy

PhD, University of Cincinnati, 2007, Arts and Sciences : Philosophy

In this dissertation I offer an account of the relationship between psychology and neurobiology. I do this in terms of two types of levels, levels of organization and levels of explanation. A hierarchy of levels of organization orders the entities and activities that are found in nature. Alternatively, the different ways of describing those things that we find in nature are placed at levels of explanation. These two types of levels need to be used together in order to understand the relationship between psychology and neurobiology. Neurobiological entities are located at the appropriate levels of organization. The descriptions offered in cognitive psychology of the capacities that humans have are located at a level of explanation above the neurobiological levels of organization. Selecting the correct levels of organization entails identifying the types of entities and the types of activities that are able to carry out psychological capacities. Based upon this requirement the appropriate levels of organization are the level where neurons and their activities occur and the level where macromolecules and their activities are found. The activities at these two levels of organization carry out the psychological capacities that are described by cognitive psychology at a higher level of explanation. In the first part of the dissertation I develop and defend a hierarchy of levels of organization that is based upon Wimsatt's account of levels of organization. In the second part of the dissertation I use Marr's account of levels of explanation as the basis for my analysis of levels of explanation. I argue that the type of description that is offered in cognitive psychology is the type that belongs at Marr's highest level of explanation. In the final part of the dissertation I combine these two different hierarchies, levels of organization and levels of explanation, into a two-dimensional framework. This entails locating the lowest level of explanation at one, or in this case (open full item for complete abstract)

Committee: Dr. Thomas Polger (Advisor) Subjects: Philosophy

PhD, University of Cincinnati, 2006, Arts and Sciences : Philosophy

The principal goal of this dissertation is to develop and defend a criterion of intelligibility adequate to the task of explaining the variety of historical appeals to this notion in philosophy and, more importantly, science. The thesis I defend holds that any account capable of explaining the historical use of intelligibility in criticizing scientific explanations must begin with the fact that charges generally encapsulated concerns with theoretical terms and their associated representational contents. For individuals like Leibniz and Boyle as well as latter atomists and other nineteenth century scientists like Maxwell the pertinent question was whether the putative referents of the theoretical terms of explanations could be conceived of. “Conceiving of” for early modern thinkers had principally to do with associating terms with ideas that are either rationally grounded or derived from experience. Later theorists like Maxwell, while remaining faithful to the traditional use of intelligibility, implicitly extend the idea by associating theoretical abstractions with models that are accessible to the imagination. After establishing this core thesis I proceed to develop an alternative account of intelligibility that can accommodate the historical worries while at the same time addressing some potential philosophical challenges to the use of this notion in the assessment of scientific theories. The thesis defended in this regard is that terms are intelligible if and only if they can be associated with observationally based representations that figure in a broader network of such representations whose relations are suitably constrained by the theory.

Committee: Dr. Robert Richardson (Advisor) Subjects: Philosophy

Doctor of Philosophy, The Ohio State University, 2006, Philosophy

I articulate and defend a new version of the coherence theory of epistemic justification. It is new, in part, because, unlike traditional varieties of coherentism, it is externalist—viz., it implies that justification supervenes, in part, on things that are neither mental nor supervenient on the mental. The theory, overall, has three distinguishing components. First, there is an explanationist component, which says that an inductive inferential relation is cogent only if it is explanatorily virtuous. In this respect, my account is in the spirit of both William Lycan's brand of coherentism, and the brand oft attributed to Gilbert Harman and Wilfrid Sellars. Second, there is a meta-perspectivalist component, saying (for starters) that S has good reason for thinking that p obtains only if from S's perspective it is likely that he is connected to p. This, when fully spelled out, has the result that S's belief system is coherent only if S has a view as to how he is connected to the outside world, and according to which the mechanisms involved (e.g., vision) are reliable. And third, there is a veridicality component, which requires that S's reasons be true, and that S's reasons for his reasons be true, and so on. This, together with the meta-perspectivalist component, requires that S be correct as to how he is reliably connected to the outside world. This is what makes my position externalist. I develop and argue positively for each of these three distinguishing components, thereby refuting my theory's chief rivals in the coherentist camp. I also argue against the three standard objections to coherentism: the Alternative-Systems Objection, the Isolation Objection, and the Experience Objection. Take the Experience Objection, for instance. It charges that foundationalist theories are superior to coherentist theories because, unlike coherentist theories, they allow a role for experiences (e.g., visual experiences) in justification. I argue that, initial appearances notwithst (open full item for complete abstract)

Committee: George Pappas (Advisor) Subjects: Philosophy

Doctor of Philosophy, The Ohio State University, 2002, Political Science

Members of the United States Senate have choices about how to convey their Washington activities to their constituents. This study examines one of those choices: whether or not to explain a significant roll call vote. This study goes to the archival record to determine how members explain these votes in press releases, in newspaper coverage, in mass mailings, and on the floor of the Senate. Through analysis of three bills before the 106th Congress, the study shows that there are key factors which affect the propensity for members to explain. This study uses content analysis to look at senators' explanations of their votes on the 1999 Juvenile Justice Act, the 1999 Bankruptcy Reform Act, and the U.S.-China Relations Act of 2000. It examines both the explanations that members give for their own votes as well as the statements that they issue regarding the behavior of the Senate as a whole. Further, the study argues that members give explanations with two goals in mind: claiming credit and managing blame. The importance of the bill to the member's constituency, the member's electoral concerns, and other characteristics such as the member's position in the chamber are shown to affect the propensity for senators to offer these explanations. The study discussed here makes three main contributions to the existing literature. First, it refocuses the examinations of political accounts toward the antecedents of these explanations. It demonstrates that not only how members vote, but how they explain these votes are important parts of the representation process. Second, the study also refocuses attention to explanations given both for positive and negative behavior. Finally, the study demonstrates a gulf between how senators wish to convey their representation (through their press releases) and how this representation is actually conveyed to the public (through news coverage). This work sets the stage for other studies of explanations in a political context.

Committee: Kathleen McGraw (Advisor) Subjects: Political Science, General

Master of Arts (MA), Ohio University, 2010, Philosophy (Arts and Sciences)

In Representation Reconsidered, William Ramsey claims that the physical structures described in representational terms by newer cognitive theories, e.g. connectionism and dynamic systems theory, aren't actually functioning as representations. As it turns out, connectionist systems should not be given representational explanations because representational theories applied to these systems do not meet Ramsey's job description challenge. In this paper I respond to his claims that teleological theories, such as those proposed by Fred Dretske and Ruth Millikan, do not provide adequate conceptual motivation for treating certain states as representations. I argue that Ramsey's analysis of representation in connectionist systems is incomplete and ultimately overlooks the strengths of Millikan's theory and its importance to this debate. These shortcomings stem from 1) Ramsey's reliance on a weak analogy between on Dretske and Millikan's theories and 2) his cursory rebuttal of Millikan's theory. I show this by revisiting Ramsey's objections to Dretske's theory as applied to connectionism and explaining how these objections are not applicable to Millikan's theory as applied to connectionism.

Committee: Robert Briscoe PhD (Committee Chair); Wendy Parker PhD (Committee Member); Scott Carson PhD (Committee Member) Subjects: Philosophy

Master of Arts, Case Western Reserve University, 2011, Cognitive Linguistics

Building on Aristotle's analysis, this thesis starts with a definition of change as the result of a network of influences operating to shape a particular outcome in manifesting a potential. It argues that change can be modeled as an interaction among flows, and that arrow diagrams are effective tools for such modeling. The thesis demonstrates how the phenomenology, ontology, and dynamics of change are prompted in the interpretation of arrow diagrams. A typology is presented that emerged from a corpus analysis of arrow diagrams sampled from medical and biomedical journals. In Chapter 2, the typology is applied to the analyses of arrow diagrams from this corpus. In Chapter 3, I propose that a systematic treatment of ways in which arrow diagramming are used to model change can become the basis for a new field that I dub “fluxemics.” Moving forward, this idea will be developed mathematically and tested empirically.

Committee: Per Aage Brandt PhD (Advisor); Peter Whitehouse MD/PhD (Committee Member); Mark Turner PhD (Committee Member) Subjects: Scientific Imaging; Systems Science

Doctor of Philosophy, The Ohio State University, 2013, EDU Teaching and Learning

Although multiple-choice assessment formats are commonly utilized throughout the educational hierarchy, they are only capable of measuring a small subset of important disciplinary competencies and practices. Consequently, science educators require open-response format assessments that can validly measure more advanced skills and performances (e.g., producing written scientific explanations). However, open-response format assessments are not practical in many educational contexts because of the high cost of scoring, the delayed feedback to test-takers, and the lack of scoring consistency among human graders. In this study, the efficacy of automated computer scoring (ACS) of written explanations is examined relative to human scoring. This study aims to build ACS models using machine-learning methods in order to detect a suite of scientific and naive ideas in written scientific explanations, and to explore approaches for optimizing these ACS models. This study develops and evaluates nine machine-learning models to detect six scientific concepts and three naive ideas of natural selection. In addition, it examines the effects of three machine-learning parameters (i.e., n-gram selection, stop words, and misclassified data) on the performance of the ACS models. In order to test the efficacy of the ACS models, a corpus of 10,270 written evolutionary explanations--in response to a variety of items differing in surface features was gathered. The corpus was scored by expert human raters and by the ACS models, and four correspondence measures were calculated: kappa, raw agreement, precision, and recall. Methodologically, the ACS models were built using the SMO (Sequential Minimal Optimization) algorithm in the LightSIDE software. Repeated-measures ANOVAs, Pearson correlations, and logarithmic regressions were used to examine the effects of the three machine learning parameters on human-computer correspondence measures, and to examine the effects of sample size on model performa (open full item for complete abstract)

Committee: Ross H. Nehm PhD (Advisor); David L. Haury PhD (Committee Member); Lin Ding PhD (Committee Member) Subjects: Educational Evaluation; Educational Technology; Science Education