Students’ Understanding of Definitional and Relational Characteristics of Confidence Intervals: Initial Results

• Activity Number: 125 - SPEED: Modernization of What, How, and Where We Teach Statistics Part 1
• Type: Contributed
• Date/Time: Monday, July 29, 2019 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and Data Science Education
• Author(s): Kristen E. Roland* and Jennifer J. Kaplan
• Companies: University of Georgia and University of Georgia
• Keywords: statistics education; confidence intervals; undergraduate education; introductory statistics; statistical inference

Abstract

Statistical inference is known to be a difficult idea to teach and learn. Recently, position papers by the APA and ASA question the use of p-values in published research, recommending decreased emphasis on the use of the critical value of alpha=0.05 and increased emphasis on reporting confidence intervals. Concerning these positions, there are studies on statistician and scientists’ understanding of confidence intervals, but there is little published research on the conceptions novices hold about definitional and relational characteristics of confidence intervals. This paper will address the gap in the literature, presenting results from responses to questions about confidence intervals posed to Undergraduate Introductory statistics students. The students were asked to demonstrate, through open-ended assessments, their understanding of the confidence level, the range of a confidence interval, the relationship between confidence level and interval width, and the relationship between sample size and interval width. The initial findings from the responses will illustrate typical students’ understanding of a confidence interval.