Page 16 - Applied statistics and probability for engineers

xii Preface

COURSE SYLLABUS SUGGESTIONS
5IJT JT B WFSZ áYJCMF UFYUCPPL CFDBVTF JOTUSVDUPST JEFBT BCPVU XIBU TIPVME CF JO B àSTU
course on statistics for engineers vary widely, as do the abilities of different groups of stu-
dents. Therefore, we hesitate to give too much advice, but will explain how we use the book.
We believe that a irst course in statistics for engineers should be primarily an applied statistics
course, not a probability course. In our one-semester course we cover all of Chapter 1 (in one or
two lectures); overview the material on probability, putting most of the emphasis on the normal
distribution (six to eight lectures); discuss most of Chapters 6 through 10 on conidence intervals
and tests (twelve to fourteen lectures); introduce regression models in Chapter 11 (four lectures);
give an introduction to the design of experiments from Chapters 13 and 14 (six lectures); and
present the basic concepts of statistical process control, including the Shewhart control chart
from Chapter 15 (four lectures). This leaves about three to four periods for exams and review.
Let us emphasize that the purpose of this course is to introduce engineers to how statistics can
be used to solve real-world engineering problems, not to weed out the less mathematically gifted
students. This course is not the “baby math-stat” course that is all too often given to engineers.
If a second semester is available, it is possible to cover the entire book, including much of
the supplemental material, if appropriate for the audience. It would also be possible to assign
and work many of the homework problems in class to reinforce the understanding of the con-
cepts. Obviously, multiple regression and more design of experiments would be major topics
in a second course.

USING THE COMPUTER
In practice, engineers use computers to apply statistical methods to solve problems. Therefore,
we strongly recommend that the computer be integrated into the class. Throughout the book
we have presented typical example of the output that can be obtained with modern statistical
software. In teaching, we have used a variety of software packages, including Minitab, Stat-
graphics, JMP, and Statistica. We did not clutter up the book with operational details of these
different packages because how the instructor integrates the software into the class is ultimate-
ly more important than which package is used. All text data are available in electronic form
on the textbook Web site. In some chapters, there are problems that we feel should be worked
using computer software. We have marked these problems with a special icon in the margin.
In our own classrooms, we use the computer in almost every lecture and demonstrate how the
technique is implemented in software as soon as it is discussed in the lecture. Student versions
of many statistical software packages are available at low cost, and students can either purchase
their own copy or use the products available through the institution. We have found that this
greatly improves the pace of the course and student understanding of the material.
Users should be aware that inal answers may differ slightly due to different numerical preci-
sion and rounding protocols among softwares.