Several comments following the presentations concerned how difficult it was for students to understand the subtleties involved in correlation coefficients and confidence intervals. Related to this were comments about the importance of assumptions and the concepts of efficiency and power and how much should be discussed in introductory level courses with respect to these topics. There was also a discussion of the role of formulas and the practice of providing formulas only after students have had experience with the concepts via technology or other activities.

The dilemma of whether to use one large dataset for an entire course or several small datasets was discussed. One participant felt that the analysis of real data was not an end in itself, but a way to illustrate and motivate statistical concepts. A disadvantage of using the same dataset throughout a course is that there will be students who may not be interested in the context of that dataset. There was a discussion regarding the rapid improvements being made in technology, resulting in the fact that the number of people having access to the Internet is growing exponentially.

It was mentioned that, worldwide, better interfaces between amputees and telephones need to be developed and that language translators need to be developed in order to make Internet resources available in many languages. Discussions also centered around the use of "black box" systems for teaching and doing data analysis. Several participants expressed mixed feelings about this. One participant said that you can train a monkey to use a black box system, but the monkey cannot make intelligent decisions.

On the other hand, many aspects of statistical computing packages are already taught as black boxes without the users knowing where the formulas come from or without even being given the formulas ND/or algorithms that produced the computer output. It was also mentioned that black box models need to be readily understandable, especially when used to teach people with no background in statistics. Another participant pointed out that black boxes sometimes work better than people. One participant pointed out that there are not only white and black boxes but intermediate gray ones.The teacher needs to decide what should be white, gray, or black.

279 C. PLUMBER There were also some questions by participants as to what managers need to know in terms of statistics and quality control. In particular, do they need to know hypothesis testing procedures, such as t-tests, and how much do they need to know about control charting, capability studies, and trouble-shooting using statistical analyses? There was also some discussion of the advantages and disadvantages of resembling methods. The small group discussion on post-secondary issues took place the last morning of the conference.The six participants all teach or have taught at the university level.

Many of the ideas summarized below were unanimous, but several were by consensus and not unanimous. The main goal of the discussion was to make commendations, although the discussions covered a wide variety of topics. Before providing recommendations, the group felt that it was important to give a statement of the discussion group's philosophy. This philosophy statement helps explain some of the rationale behind the recommendations, and states: At present syllabuses tend to be overcrowded and ineffective.

We think that syllabuses should be as simple as possible given: The purposes of the students for enrolling in statistics courses. The available technology. The abilities of the students. The group also discussed ways in which technology can be used. In particular, technology can enable: ; Instructors to rethink content (basic ideas, new ideas).

Learning to be more active. Learning to be more utilitarian. Sharing of knowledge (Internet, etc. ). The group's recommendations in terms of teaching and learning were: ; We need to rethink content. Curricula need to be developed independent of a particular platform.

Although these recommendations are given here in the context of post-secondary education, many of them are applicable at the primary and secondary levels as well. The sources of these recommendations will be given in parentheses after each commendation. When the phrase "large group discussion" is used it refers to the discussions that took place at the conference after each of the individual presentations. Also, the word technology is used throughout these discussions to mean more than calculators and computers.It can include, but is not limited to, such diverse technologies as paper-and-pencil (see the paper by Jones), radios, videos, and compact discs. Recommendations on teaching and learning 1 .

It is important to always think about the customers being served (usually the students and their present or future employers) when designing curricula both in arms of topics taught and the use of technology. The teacher, the students, or both, may use technology. (Sources: All of the papers in this Section and the large and small group discussions. ) 2.

The actual topics taught and depth of instruction must be rethought in light of the available technology. In particular the role of probability needs to be examined carefully. (Sources: All of the papers in this Section and the large and small group discussions. ) 3. Curricula developers should try to develop the materials that use computers or calculators to be independent of a particular platform or calculator.

They must also take into account that students and teachers often will not have access to the latest technology. (Sources: Jones, Oarsman, Starlings, and the large and small group 4.