In the following pages we shall illustrate the features of our prototype SUSAM, which is presently being trialled within the course in linear regression run by our department.
In the following pages we shall illustrate the features of our prototype SUSAM, which is presently being trialled within the course in linear regression run by our department.
In 1974, one of the authors (GAFS) introduced a terminating first year service course in statistics, paper 26.181, for non-mathematics students. In the following years we observed that the service course 26.181 catered not only for students majoring in other subjects, but also for a substantial number of mathematics students who preferred a more practical approach than the traditional one. The numbers were also steadily growing (about 1500 in 1990). At that stage we realised that 26.181 provided a potential source of advancing statistics students who might be interested in taking a second year follow-up course. In the 1980s Alan Lee (and probably most of us that vintage) had been strongly influenced by works on exploratory data analysis by such people as Tukey , McNeil, Velleman and Hoaglin. Under Lee's direction a second year data analysis course 26.281 was launched in 1981. His aim was to provide further training in practical statistics and data analysis without requiring too much mathematical knowledge or statistical theory. He realised that the students needed easy computer access for a realistic approach to data analysis. Suitable access to a mainframe was out of the question at the time, but micros seemed a viable alternative. A suitable statistical package was therefore developed called STATCALC (Lee et al., 1984) which had partly evolved from several programs on exploratory data analysis adapted from McNeil (1977) by Ross Ihaka. The package runs on IBM and Macintosh personal computers, the latter being currently used in our department. Lee and Peter Mullins also wrote a manual to go with the package. The manual, with its extensive tutorial section, also serves as the text for the course.
Increased availability of computers and easy-to-use statistical software has greatly enhanced the ability to efficiently use large sets of real data for motivation and illustration of statistical concepts in applied courses. Several examples of such data and their use in a variety of courses are given below.
We have found spreadsheets very useful for teaching statistics, operations research, and other quantitative methods in commerce: students learn quickly when typing and debugging formulae and macros, and get a real feeling for the relation between computation and theory. However, after a while, it is necessary to turn to a more traditional statistical package to find both standard and advanced procedures needed for real-world data examination. WESTAT Associates has developed the MASS system for some years, a stand-alone statistical package of some size and power, and we have used it extensively for teaching, at second course but not at introductory level. In the past six months we (1.5 persons) have ported it to sit on the new spreadsheet WingZ (by Informix Inc) on the Mac to form a new program StatZ, which combines the advantages of a spreadsheet and a specialised program for the teaching and practice of statistics. (See Section 3(ii) for other computers.) Porting a program from one language or machine or system to another is usually a painful exercise. However, the power and flexibility of the "scripting" and of the facilities for linking external code provided by WingZ and its HyperScript macro language, which we have used for incorporating our MASS code into their highly commercial, well-tested base product, seems destined to have a profound effect on related areas of program development. This will yield new research and teaching software, as already pioneered by Apple's HyperCard, the technical precursor of WingZ's HyperScript.
These instructional methods are not to be construed as a guide to how introductory graduate statistics should be taught, but rather as an example of how one instructor tries to teach adult learners in an elementary statistics environment.
Regardless of the level of sophistication of one's students, or the exact content of one's course, it is important to think broadly about the general messages one wishes to convey, and then to formulate a number of explicit goals one would like to achieve in teaching statistics to sociology students. Five such goals are discussed: 1) overcoming fears, resistances, and tendencies to overmemorize; 2) the importance of intellectual honesty and integrity; 3) understanding the relationship between deductive and inductive inferences; 4) learning to play the role of reasonable critic; and 5) learning to handle complexities in a systematic fashion. Illustrative examples are given to show how exercises can be tailored to the course's contents and the level of student backgrounds.
The purpose of this volume, and others in the Addenda Series, is to provide instructional ideas and materials that will support implementation of the Curriculum and Evaluation Standards in local settings.
Students in elementary statistics traditionally see experiments and data as words and numbers in a text. They receive little exposure to the important statistical activities of sample selection, data collection, experimental design, development of statistical models, the need for randomization, selection of factors, etc. They often leave the first course without a firm understanding of the role of applied statistics or of the statistician in scientific investigations. In an attempt to improve elementary statistics education, we have developed a statistics laboratory similar to those of other elementary science courses. We will discuss our experiences in teaching the one-semester hour Elementary Statistics Laboratory course that can be taken with or after the traditional elementary statistics course. In each session students, working in teams, discuss the design of an experiment, carry out the experiment, and analyze their data using Minitab on a Macintosh. The students then individually either answer a series of short answer questions or write a formal scientific report. The labs are designed to be relatively inexpensive and do not require a prior background in science, statistics or computing.
The Quantitative Literacy (QL) project has affected how statistics is viewed and taught by high school mathematics teachers. Each summer since 1987 the ASA Center for Statistics Education has organized QL workshops at various places around the country. This movement has been in concert with the National Council of Teachers of Mathematics movement to revamp mathematics instruction with their Curriculum and Development Standards. Quantitative Literacy is now a major part of the thinking of national and local leaders in mathematics education. Unfortunately, few science teachers have been affected by the QL project. While mathematics teachers introduce boxplots in their algebra classes, the science teachers in the same building have each student complete a laboratory exercise and turn in a report, without ever considering how the results of the various students differ. The middle school or high school science laboratory is an excellent place in which to use statistical ideas, but rarely does this happen. In 1990 ASA organized a planning meeting that led to the formation of the SEAQL (Science Education And Quantitative Literacy) task force. This group of statisticians and science teachers is promoting the use of statistics in school science courses by focusing on common laboratory experiments that involve data collection. The task force will host a leadership conference in November for science curriculum supervisors. At the conference the task force will demonstrate some SEAQL laboratory activities and convey the philosophy of using data analysis as a science teaching tool.
We are proposing a statistical methods sequence, each having separate lecture-based and stat laboratory components. First, the lecture-based courses will allow a through examinations of the "when to" and "what-to", while the statistical lab will not only expose the student to the "how-to", but through simulations and discipline related problems motivate and demonstrate the underlying concepts discussed in the course lectures. The components should compliment each other, rather than be adversarial. Secondly, on many campuses across the country questions arise concerning how to incorporate writing in the undergraduate curriculum. The MTH 441/442 sequence is a perfect venue for incorporating student writing in a mathematics course. One objective of this revised MTH 441/442 sequence is to encourage student development of writing skills. The assignment in the stat lab sessions are to be completed in a report format. Using the edit options in the Primos System (if done on the mainframe) or the available text editing software on the PC, the student will be expected to "clean-up" the output from the statistical package and coherently express their analysis of the results in a written report. Not only should the statistical lab contribute to their mastery of the "how-to" of packages such as Minitab, SPSSx, and SAS, but the student should benefit from gaining stronger written communications skills.