By Susan Lloyd (Bucknell University)
Information
Research suggests that students’ ways of thinking about key statistical concepts are highly interrelated. Considering confidence intervals in particular, several related ideas—referred to here as attributes—are commonly taught in an introductory statistics course. We identify five main attributes for this research: confidence level, error, sample size, p-value, and method. These attributes guide the development of an educational assessment designed to measure statistical literacy surrounding confidence intervals in a way that allows for the study of interaction effects. To qualitatively study the meanings students have about these attributes, we conducted six think-aloud interviews in which students worked with select items from the assessment. We then used grounded theory to model students’ conceptualizations of confidence level and error, focusing on preliminary evidence and patterns of interaction among these students’ ways of thinking. Our findings suggest that students coordinate their thinking about related attributes of confidence intervals and highlight the prevalence of lexical ambiguity in statistical terms, both of which warrant further exploration. By better understanding how students coordinate their thinking about related attributes of key statistical concepts, educators can foster more connected understanding across concepts and strengthen classroom interactions. When students are able to see how concepts are interrelated and use their thinking about one concept to enhance—rather than hinder—their understanding of others, they are more likely to experience the joy and discovery that comes with learning statistics. This work contributes to conversations around qualitative research, educational assessment, and innovative approaches that provide preliminary evidence of interaction effects in students’ thinking about confidence intervals. This study was conducted at an R1 institution with undergraduate students who had recently completed an introductory statistics course. Each student participated in an individual think-aloud interview conducted via Zoom.