Browsing by Author "Kennedy, Paul, advisor"
Now showing 1 - 5 of 5
Results Per Page
Sort Options
Item Open Access A mixed methods explanatory study of the failure/drop rate for freshman STEM calculus students(Colorado State University. Libraries, 2013) Worthley, Mary, author; Gloeckner, Gene, advisor; Kennedy, Paul, advisor; Banning, James, committee member; Siller, Thomas, committee memberIn a national context of high failure rates in freshman calculus courses, the purpose of this study was to understand who is struggling, and why. High failure rates are especially alarming given a local environment where students have access to a variety of academic, and personal, assistance. The sample consists of students at Colorado State University (CSU) who attended a course in freshman calculus from Fall 2007 to Fall 2012. An explanatory sequential mixed methods approach was used in this study. Using data from CSU's Registrar's Office and Mathematics department, descriptive statistics highlighted several student attributes worth pursuing. Fall and spring cohorts have a different make up and different outcomes. Hence this study concentrated on the fall cohort, which comprises mainly of freshmen. The combination of attributes that produced the strongest prediction of student's final result in calculus were Colorado Commission on Higher Education index scores, CSU Mathematics department placement test scores, and calculus repeat status (R2 =.30, n=1325). For Fall 2012, these attributes were combined with student motivation and student strategies constructs, measured using the MSLQ instrument. The combination giving the strongest prediction of student's first mid-term examination results (R2 =.34, n=124) included CSU Mathematics department placement test scores, along with MSLQ constructs test anxiety, and self-efficacy for learning and performance. However, using logistic regression only 38.7% of the students who failed were correctly predicted to fail. Former students of CSU's calculus course aimed at freshmen STEM students were interviewed or surveyed, in an attempt to probe how students experience this course. Several common elements emerged. Students were dedicating vast amounts of time to this course. There was a common belief this course could be passed if the student worked hard enough. The difference between those who succeeded and those who did not appeared to relate to how this study time was spent. Those who floundered often struggled to locate appropriate help, although they were quite aware they needed assistance. Many of those interviewed also avoided working with other students. Reasons cited ranged from claims of being individual learners, to frustration at finding a group who had the same study goals. Some non-traditional students were also alienated by the prospect of working with 'teenagers'. Two other results from the analysis of student interviews suggested reanalyzing the quantitative data and including student's prior history with mathematics, as well as if the student was non-traditional. The combination of attributes that gave the strongest relationship (R2 =.40, n=101) were CSU Mathematics department placement test results, combined with MSLQ constructs test anxiety, self-efficacy for learning and performance, organization, as well as the student's own appraisal of the quality of mathematics teaching they received in high school. However, the ability to accurately predict if a student will fail was minimal. Focusing on students who do fail, three groups of students of interest were isolated: those who have yet to declare their major, 'non-traditional' students, particularly those enrolled in the eight a.m. class, and, curiously, those students who choose to enroll in the ten a.m. class.Item Open Access A qualitative study of college students' conceptions of rivers(Colorado State University. Libraries, 2008) Sexton, Julie Marie, author; Cobb, R. Brian, advisor; Kennedy, Paul, advisorThis study explored two research questions: (1) What are college students' conceptions of river topics and (2) What are the emergent patterns between students' conceptions and their gender and ethnicity? A basic interpretive qualitative research design was used. Purposeful sampling was used to recruit 24 college students from an introductory geology class. In-depth interviews were conducted and a demographic questionnaire administered. Interview responses were analyzed using a modified version of constant comparative analysis. Between-gender patterns and between-ethnic group patterns of the type of conceptions held were investigated by calculating relative likelihood statistics. Four findings emerged. Finding 1: students held mostly scientific and incomplete scientific conceptions of the nine river topics covered in the study. Finding 2: students' conceptions were complex. In addition to scientific and alternative conceptions, students also held incomplete scientific, incomplete scientific-alternative, and scientific-alternative conceptions. Individual students held a range of conceptions across the river topics. For each topic, there was a range of conceptions held by students. Finding 3: students had more alternative conceptions for processes, causes, and difficult-to-observe features. Finding 4: patterns were observed between students' conceptions and their gender and ethnicity. Men held scientific, incomplete scientific, and scientific-alternative conceptions more frequently than did women. Women held incomplete scientific-alternative and alternative conceptions more frequently than did men. White students held scientific, incomplete scientific, and scientific-alternative conceptions more frequently than did Hispanic students. Hispanic students held incomplete scientific-alternative and alternative conceptions more frequently than did White students.Item Open Access Cooperative learning instruction and science achievement for secondary and early post-secondary students: a systematic review(Colorado State University. Libraries, 2009) Romero, Christopher C., author; Cobb, R. Brian, advisor; Kennedy, Paul, advisorA systematic review of 2,506 published and unpublished citations identified in a literature search on science outcomes associated with cooperative learning in secondary and early post-secondary science classrooms between 1995 and 2007 was conducted. The goal of this review was to determine what impact cooperative learning had on science achievement of students compared to traditional instruction. A tri-level screening and coding process was implemented and identified 30 original, empirical studies that met the inclusionary criteria while yielding an overall effect size estimate. The minimum methodological criteria for inclusion were as follows: (a) the study utilized a treatment/control design, (b) cooperative learning was the intervention, and the control group experienced traditional instruction, (c) the subjects in included studies were secondary or early-post-secondary students, (d) the study was performed in a science classroom, and (e) student achievement was the outcome measure. This meta-analysis describes the main effect of cooperative learning; additionally, a variety of moderator analyses were conducted in order to determine if particular study and participant characteristics influenced the effect of the intervention. The results of this review indicate that cooperative learning improves student achievement in science. The overall mean effect size was .308, a medium effect (Cohen, 1988). Moderator analyses on study participant characteristics gender and ability level were inconclusive based on the small number of studies in which data on these characteristics were disaggregated. If the intervention was structured in a particular fashion, the effect on student achievement was greater than that for an unstructured intervention. The intervention showed a greater effect on student achievement in biology classes than in other science disciplines. Studies performed using cluster randomized or quasi-experimental without subject matching methodologies showed a greater effect on student achievement in science than studies that used the quasi-experimental with subject matching methodology. Implications for teacher education policy and recommendations for improvements in methodological practices and reporting are given.Item Open Access The flipped classroom model for college algebra: effects on student achievement(Colorado State University. Libraries, 2014) Overmyer, Gerald Robert, author; Gloeckner, Gene, advisor; Kennedy, Paul, advisor; Rambo-Hernandez, Karen, committee member; Penttila, Tim, committee memberIn the past few years there has been a substantial rise in the use and interest in a teaching and learning paradigm most commonly known as the flipped classroom. The flipped classroom model encompasses any use of using Internet technology to leverage the learning in a classroom, so that a teacher can spend more time interacting with students instead of lecturing. This is most commonly done by using teacher created videos that students view outside of class time. It is called the flipped class model because the whole classroom/homework paradigm is "flipped". In its simplest terms, what used to be classwork (the lecture) is done at home via teacher-created videos and what used to be homework (assigned problems) is now done in class. Five sections of college algebra where taught using the flipped classroom model. Six sections of college algebra where taught using the traditional method of lecture and homework. This quasi-experimental quantitative research compares sections of college algebra using the flipped classroom methods and the traditional lecture/homework structure and its effect on student achievement as measured through common assessments. In the traditional sections, students spent class time receiving lecture and reviewing homework and exams. Outside class time was spent on traditional homework. In the flipped sections, students viewed short video lectures and submitted basic homework solutions online outside of class time. Students then completed their homework assignments in class with the instructor. Some flipped section instructors also used collaborative group work, inquiry-based learning, and active whole-class discussions. All sections took common assessments for their final exam and completed a pre/post algebra readiness exam. The exam data from the sections were analyzed and compared using regression and ANOVA methods with instructional method, gender, and ACT mathematics scores as independent variables. Final exam scores and pre/post algebra readiness exam scores were the dependent variables. The findings of this research show that there was not a statistically significant difference in the scores of student in the two groups, however students in the flipped sections did score slightly better than student in the traditional sections. Instructors of flipped sections who had previous classroom experience with inquiry-based and cooperative learning methods had sections with statistically significant higher common final exam scores. The results are followed by implications for teaching and recommendations for practice and further research.Item Open Access Using a variation of the cohort control design to evaluate large-scale, long-term, complex professional development programs(Colorado State University. Libraries, 2009) Sample McMeeking, Laura, author; Cobb, R. Brian, advisor; Kennedy, Paul, advisorThe purpose of this study was to introduce a variation on the posttest-only cohort control design and answer questions concerning both methodological credibility and practical utility of employing the variation design in evaluations of large-scale, complex professional development programs. The original design and variation design, which adds a pretest measure for prior student performance, were compared theoretically and practically using data from the RM-MSMSP program to compare the advantages and disadvantages of the two evaluation designs. Two separate 2 x 2 ANOVA analyses were used to compare the designs. Findings indicated that, as expected, there were differences in the outcomes using the two designs. While the outcomes were not consistently different, they could plausibly be explained. Because the findings of the variation design were supported by similar findings in the literature, credence was given to the variation design. Given the added control of the variation design, evaluations using the variation design could control for selection bias where those using the original design could not. Realistically, however, the choice of evaluation design is one of trade-offs, because the addition of controls through gain scores, as is the case for the variation design, also comes with some disadvantages. If certain conditions for the data are met and the outcome measure is calibrated, the variation design would be a good choice for a professional development evaluation. If not, the original design would also be perfectly reasonable. From a practical standpoint, the variation design is no less practical to employ than the original cohort control design if certain program conditions about data collection and availability are met. In addition, the outcome measure in the form of a gain score is similar to value-added evaluation designs that are politically popular due to a focus on student growth. This coupled with the methodological advantages of the variation design make it a useful evaluation design for large-scale, long-term, complex professional development programs wishing to investigate the effects of professional development on student achievement.