A study of the effects on statistical reasoning, skill and conceptual development, and attitude of students in an elementary statistics class which is taught in a contextually and technologically rich environmen

This study was an effort to identify and compare the effects of two different teaching environments on statistical reasoning, conceptual and skill development, and attitudes and beliefs of elementary statistics students. The control class was taught using a formal, rule-driven approach to learning. The experimental class was taught in a contextually and technologically rich environment and was designed as a statistics working group with the instructor as group leader and the students as apprentices learning data analysis. This class used a single large database throughout the semester.;Data from this study suggest that it is possible to identify and categorize statistical reasoning used by elementary statistics students. Reliability of categorization by a panel of trained raters was 0.766. There was no significant difference between the two classes in the frequency of use of the Categories of Naive/Statistical, Developing/Statistical, and Functional/Statistical reasoning. In their interview transcripts the experimental group used significantly more episodes of reasoning for the Categories of Non-determinable and Arithmetical/Statistical reasoning, and the control group had a significantly higher frequency of use of the Category of Procedural/Statistical reasoning. There was no significant difference between the two classes on the Garfield/Konold Reasoning Assessment either before or after the treatment.;The control class scored significantly higher on 5 of the 14 skills questions and on 1 of the 13 concept questions. The experimental scored better on 2 of the 13 concept questions. A series of four interviews suggested that students from both classes built understanding around contextual experiences and that the nature of their learning was different. The evidence suggested that the experimental class interviewees reasoned more empirically; whereas, the control class interviewees exhibited a greater tendency to reason using formulas.;Statistical beliefs and attitudes were identified and classified from the interview transcripts. No differences between the two classes was apparent.;The students in the experimental class demonstrated a holistic understanding of the components of statistical analysis. This was evidenced in the written reports of the results of individually-designed statistical analyses which were done as a final project by each member of the class.