Assessing Student Writing in Statistics

• Activity Number: 544
• Type: Invited
• Date/Time: Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Education
• Author(s): Jennifer Kaplan*+ and Neal Rogness and Diane Fisher
• Companies: University of Georgia and Grand Valley State University and University of Louisiana at Lafayette
• Keywords: lexical ambiguity ; language ; writing ; assessment ; random ; instruction

Abstract

The overarching goals of the lexical ambiguity project are to: 1) highlight words that exhibit ambiguity; 2) make suggestions for instructors and authors to mitigate the effects of the ambiguities; and 3) assess the effects of the mitigation efforts on student learning outcomes. The data collected for this project consist of several thousand student-generated sentences and definitions for the target words. A sample of the responses was hand coded to create coding categories. While the inter-rater reliability for the hand scoring is quite high, hand scoring is both time consuming and only provides a count of responses that fit each category In this presentation we will discuss the efficiency in analysis provided by the use of lexical analysis software and the ability of the software to uncover connections between student responses and the statistical concepts underlying the target words. In particular, the analysis of students' writing about the word random will be presented and differences between students who underwent a mitigation experience and those who did not will be highlighted, but this method of analysis of may be used more broadly for assessment and research purposes.