No description available.

No description available.

No description available.

No description available.

This dissertation presents the results of an experiment that measured the learning outcomes associated with three different pedagogical approaches to introductory physics labs. These three pedagogical approaches presented students with the same apparatus and covered the same physics content, but used different lab manuals to guide students through distinct cognitive processes in conducting their laboratory investigations. We administered post-tests containing multiple-choice conceptual questions and free-response quantitative problems one week after students completed these laboratory investigations. In addition, we collected data from the laboratory practical exam taken by students at the end of the semester. Using these data sets, we compared the learning outcomes for the three curricula in three dimensions of ability: conceptual understanding, quantitative problem-solving skill, and laboratory skills. Our three pedagogical approaches are as follows. Guided labs lead students through their investigations via a combination of Socratic-style questioning and direct instruction, while students record their data and answers to written questions in the manual during the experiment. Traditional labs provide detailed written instructions, which students follow to complete the lab objectives. Open labs provide students with a set of apparatus and a question to be answered, and leave students to devise and execute an experiment to answer the question. In general, we find that students performing Guided labs perform better on some conceptual assessment items, and that students performing Open labs perform significantly better on experimental tasks. Combining a classical test theory analysis of post-test results with in-lab classroom observations allows us to identify individual components of the laboratory manuals and investigations that are likely to have influenced the observed differences in learning outcomes associated with the different pedagogical approaches. Due to the novel nature of this research and the large number of item-level results we produced, we recommend additional research to determine the reproducibility of our results. Analyzing the data with item response theory yields additional information about the performance of our students on both conceptual questions and quantitative problems. We find that performing lab activities on a topic does lead to better-than-expected performance on some conceptual questions regardless of pedagogical approach, but that this acquired conceptual understanding is strongly context-dependent. The results also suggest that a single “Newtonian reasoning ability" is inadequate to explain student response patterns to items from the Force Concept Inventory. We develop a framework for applying polytomous item response theory to the analysis of quantitative free-response problems and for analyzing how features of student solutions are influenced by problem-solving ability. Patterns in how students at different abilities approach our post-test problems are revealed, and we find hints as to how features of a free-response problem influence its item parameters. The item-response theory framework we develop provides a foundation for future development of quantitative free-response research instruments. Chapter 1 of the dissertation presents a brief history of physics education research and motivates the present study. Chapter 2 describes our experimental methodology and discusses the treatments applied to students and the instruments used to measure their learning. Chapter 3 provides an introduction to the statistical and analytical methods used in our data analysis. Chapter 4 presents the full data set, analyzed using both classical test theory and item response theory. Chapter 5 contains a discussion of the implications of our results and a data-driven analysis of our experimental methods. Chapter 6 describes the importance of this work to the field and discusses the relevance of our research to curriculum development and to future work in physics education research.

No description available.

This study investigated the relationship between experiences of childhood maltreatment, family functioning, problem-solving style, and criminal offending. Participants (N = 120) were all male and classified as violent (V), non-violent (NV), or non-offenders (NO) based on their criminal histories. The Childhood Trauma Questionnaire (CTQ), Family-of-Origin Scale (FOS), Social Problem-Solving Inventory-Revised (SPSI-R), Beck Depression Inventory (BDI), and Brief Symptom Inventory (BSI) were administered to all participants. Results suggest that V's reported significantly more maltreatment compared to NO's. V's came from the least healthy families while NO's came from the healthiest families. Finally, V's were more likely to use an impulsive approach and less likely to use a rational approach to solving problems than NO's. Experiences of physical neglect discriminated the groups. This study also discovered previously ignored correlations between these variables within each of the three groups and implications of these new findings are discussed.

Practical intelligence that accounts for people's performance on real-life problem solving is not related to intelligence in the traditional theories. The primary purpose of this research was to investigate the role of two competing cognitive models in explaining practical intelligence. The author extracted from the literature four cognitive processes and two types of knowledge that significantly accounted for performance on real-life problem solving. The cognitive processes model included (a) metacognition, (b) defining a problem, (c) flexibility of thinking, and (d) selecting a solution strategy. The types of knowledge model included (a) structural knowledge, and (b) tacit knowledge. The secondary purpose of this research was to determine the contribution of some non-cognitive factors to practical intelligence. These factors included (a) self-efficacy, and (b) motivation. These processes and constructs were derived from contemporary theories of intelligence including the Triarchic Theory of Sternberg (1985a), the Bioecological Treatise of Ceci (1996), and theories of expertise.The author developed a Practical Intelligence Instrument (PII) battery based on components of the cognitive processes model, the types of knowledge model, and non-cognitive factors. The PII battery consisted of several subscales to measure components mentioned above. The PII also included items to measure familiarity with problems. The PII was administered to 116 volunteer participants. The validity of the PII subscales was derived from three sources: content, face, and construct validity, including convergent and discriminant. The reliability of the subscales in the PII battery ranged from .63 to .93. The PII also included four scenarios that are real-life problems. Participants were asked to provide solutions for these problems. Three experts from the social science field evaluated participants' strategies based on four criteria. Several statistical procedures were used to analyze the data including a hierarchal multiple regression model, ANOVA, and the Pearson Product-Moment correlation.The results showed that around 54% of the variance in practical intelligence was explained by the cognitive processes model, the types of knowledge model, and self-efficacy and motivation. The cognitive model explained around 42%. The types of knowledge model explained around 15%. The non-cognitive factors explained around 20 % of the variance in practical intelligence.

While theory formation and the relation between theory and data has been investigated in many studies of scientific reasoning, researchers have focused less attention on reasoning about experimental design, even though the experimental design process makes up a large part of real-world scientists' reasoning. The goal of this thesis was to provide a cognitive account of the scientific experimental design process by analyzing experimental design as problem-solving behavior (Newell & Simon, 1972). Three specific issues were addressed: the effect of potential error on experimental design strategies, the role of prior knowledge in experimental design, and the effect of characteristics of the space of alternate hypotheses on alternate hypothesis testing. A two-pronged in vivo/in vitro research methodology was employed, in which transcripts of real-world scientific laboratory meetings were analyzed as well as undergraduate science and non-science majors' design of biology experiments in the psychology laboratory. It was found that scientists use a specific strategy to deal with the possibility of error in experimental findings: they include "known" control conditions in their experimental designs both to determine whether error is occurring and to identify sources of error. The known controls strategy had not been reported in earlier studies with science-like tasks, in which participants' responses to error had consisted of replicating experiments and discounting results. With respect to prior knowledge: scientists and undergraduate students drew on several types of knowledge when designing experiments, including theoretical knowledge, domain-specific knowledge of experimental techniques, and domain-general knowledge of experimental design strategies. Finally, undergraduate science students generated and tested alternates to their favored hypotheses when the space of alternate hypotheses was constrained and searchable. This result may help explain findings of confirmati

No description available.