Teacher’s Corner Six Online Statistics Courses: Examination and Review Jorge LARREAMENDY-JOERNS, Gaea LEINHARDT, and Javier CORREDOR We extend George W. Cobb’s evaluative framework of statistics textbooks to six online instructional materials that exemplify the diversity of introductory statistics materials available on the Internet. Materials range from course Web sites with limited in- teractive capabilities to courseware and electronic textbooks that make extensive use of interactive learning objects and environ- ments. Instructional materials are examined in light of recent cognitive research that underscores the robustness of learning from examples, the importance of authentic problem solving in promoting knowledge in use and skill acquisition, and the use of feedback to maximize learning opportunities. Selected units that focus on statistical tools (measures of central tendency, simple linear regression, and one-way analysis of variance) are analyzed in terms of authenticity and diversity of examples, authenticity and cognitive complexity of exercises, and use of interactive learning objects and feedback. General conclusions and sug- gestions for future directions for online statistics instruction are presented. KEY WORDS: Distance learning; Evaluation; Web-based in- struction. 1. INTRODUCTION In 1987, George W. Cobb examined 16 introductory statis- tics textbooks in the Journal of the American Statistical Asso- ciation. Cobb laid out an evaluative framework that considered technical level and quality of exposition, topics covered, and quality of exercises. He selected four standard topics: sample mean, sample standard deviation, normal distribution, and sam- pling distribution of the mean. He characterized explanations by identifying the extent to which the expositions relied on for- mulas and derivations. Cobb estimated the breadth and depth of explanations by comparing the content covered and the level of detail within an additional set of topics (regression, analysis of variance, exploratory data analysis, and computers). Finally, he addressed the quality of exercises by estimating the authen- Jorge Larreamendy-Joerns is Associate Professor, Department of Psychol- ogy, Universidad de los Andes, Bogota, Colombia. Gaea Leinhardt is Professor, and Javier Corredor is Doctoral Student, University of Pittsburgh, PA (E-mail: gaea@pitt.edu). Support for this article was provided by a grant to the second author from Carnegie Mellon University via The William and Flora Hewlett Foundation. The opinions expressed do not necessarily reflect the position or the policy of the foundation and no official endorsement should be inferred. ticity of datasets, the meaningfulness of the studies described in the problem statements, and the ratio of thinking to number crunching. In the 17 years since Cobb’s evaluative framework was pub- lished two major changes have occurred in the landscape of statistics education: First, there has been growing recognition of statistical knowledge as a crucial component of core scientific literacy (Utts 2003). As a result, we now see the teaching and learning of statistics in elementary, secondary, and higher educa- tion (NCTM 2000). Second, there has been a flowering of online technologies and courses that both support and teach statistics. The use of online technologies is often predicated under the as- sumptions that the Internet can contribute to making statistical knowledge accessible to vast audiences and that online mul- timedia environments can make learning more interactive and meaningful. Today an impressive variety of instructional mate- rials in statistics education is accessible on the Internet, from full stand-alone introductory courses and electronic textbooks to digital repositories of learning objects and datasets. This article extends Cobb’s evaluative framework to a set of online instructional materials. Like Cobb, we focus on the qual- ity of explanations and exercises. Yet, in identifying their critical features we draw on recent cognitive research that underscores the prevalence of learning from examples, the importance of gen- uine problem solving in promoting knowledge in use, and the use of feedback to maximize learning opportunities. Our goals are to provide a sense of the quality of some instructional materi- als available on the Internet and to suggest criteria for inspecting such materials; criteria that we imagine will be expanded and developed over time as more materials and resources emerge. 2. EVALUATIVE GOALS AND CRITERIA FOR ONLINE COURSEWARE An important goal for the evaluation of online multimedia courseware is the assessment of instructional explanations and learning opportunities. In order to assess explanations and learn- ing opportunities three analyses need to be conducted. First, conceptions of learning and teaching that underlie the design of instructional materials must be explicated. Second, the extent to which instructional materials comply with well-established prin- ciples of learning and teaching should be assessed. Third, learn- ing affordances and constraints linked to the technical imple- mentation of courseware need to be specified. [We use the term “affordances” following Gibson’s (1977) sense of the term as it has been adopted by psychologists and others (Greeno 1994). An affordance is an opportunity in an environment to make use of a physical or mental resource to accomplish a goal. But the op- 240 The American Statistician, August 2005, Vol. 59, No. 3 © 2005 American Statistical Association DOI: 10.1198/000313005X54162