Student Performance in Curricula Centered on Simulation-Based Inference

Autor/inn/en	Chance, Beth; Tintle, Nathan; Reynolds, Shea; Patel, Ajay; Chan, Katherine; Leader, Sean
Titel	Student Performance in Curricula Centered on Simulation-Based Inference
Quelle	In: Statistics Education Research Journal, 21 (2022) 3, Artikel 4 PDF als Volltext Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; online; Zeitschriftenaufsatz
ISSN	1570-1824
Schlagwörter	Mathematics Curriculum; Simulation; Student Characteristics; Prior Learning; Prediction; Textbooks; Statistics; Mathematical Concepts; Outcomes of Education; College Mathematics; College Students; Achievement Gains; Student Attitudes; College Entrance Examinations; Statistical Inference; SAT (College Admission Test); ACT Assessment + Suchen Sie Ihr Suchwort? Simulation program; Simulationsprogramm; Vorkenntnisse; Vorhersage; Textbook; Text book; Schulbuch; Lehrbuch; Statistik; Lernleistung; Schulerfolg; Collegestudent; Achievement gain; Leistungssteigerung; Schülerverhalten; Aufnahmeprüfung; Inferential statistics; Schließende Statistik; Assessment; Eignungsprüfung; Eignungstest; Hochschulzulassung
Abstract	Using simulation-based inference (SBI), such as randomization tests, as the primary vehicle for introducing students to the logic and scope of statistical inference has been advocated with the potential of improving student understanding of statistical inference and the statistical investigative process. Moving beyond the individual class activity, entirely revised introductory statistics curricula centering on these ideas have been developed and tested. Preliminary assessment data have been mostly positive. In this paper, we discuss three years of cross-institutional tertiary-level data from the United States comparing SBI-focused curricula and non-SBI curricula (86 distinct institutions). We examined several pre/post measures of conceptual understanding in the introductory algebra-based course using multi-level modelling to incorporate student-level, instructor-level, and institutional-level covariates. We found that pre-course student characteristics (e.g., prior knowledge) were the strongest predictors of student learning, but also that textbook choice can still have a meaningful impact on student understanding of key statistical concepts. In particular, textbook choice was the strongest "modifiable" predictor of student outcomes of those examined, with simulation-based inference texts yielding the largest improvements in student learning outcomes. Further research is needed to elucidate the aspects of SBI curricula that contribute to observed student learning gains. (As Provided).
Anmerkungen	International Association for Statistical Education and the International Statistical Institute. PO Box 24070, 2490 AB The Hague, The Netherlands. Tel: +31-70-3375737; Fax: +31-70-3860025; e-mail: isi@cbs.nl; Web site: https://iase-web.org/ojs/SERJ
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2024/1/01