Please join us on Friday at 1 PM in 103A Walker Hall to hear Kristen Roland speak on...
Confidence Doesn't Mean Probability: The Jamie's Colleague Task
Dr. Kristen E. Roland's research interests include conceptualizations of confidence intervals, post-secondary statistics education, and pre-service secondary mathematics teachers’ development in statistics. She is an AMTE STaR Fellow, cohort 2022 and has recently received an honorable mention for Best [International Conference on Teaching Statistics] Paper by Early Career Researchers. The colloquium presentation will discuss the findings presented in the conference proceeding.
Abstract: The use of data to communicate current events and issues (on-going pandemic, climate change, sports analysis, etc.) has increased over the years, meaning world citizens need to understand the uncertainty that exists within statistical results, now more than ever. During this same time, professional associations (American Statistical Association and American Psychological Association) have called for increased discussion of the uncertainty associated with statistical analysis within research, including the use of confidence intervals in reporting inferential results. Little is known, however, about the knowledge structures necessary for a robust understanding of communicating this uncertainty. This talk will focus on one way to discuss confidence intervals. I will present a case study that demonstrates an important connection within a formal concept image for the concept of confidence intervals. Understanding the interpretations of confidence intervals and confidence levels requires understanding the probabilistic difference in a confidence interval estimator and a confidence interval estimate. Several tasks were developed to explore participants’ demonstrated knowledge of this difference and its connection to the interpretations of confidence intervals and confidence levels. This case study provides evidence of 1) the necessity of a hypothesized curricular concept cluster including the concepts of confidence interval estimator/ estimate and 2) the importance of its connection to both the interpretation of confidence intervals and confidence levels.
There will also be zoom availability for this talk. To get the zoom link, please contact Nadun Dissanayake (nadunkul).