SIE 330 - Engineering Experiment Design

SIE 330 - Engineering Experiment Design (3) Design and analysis of observational and factorial experiments employing numerical and graphical methods. Topics include control charts, probability plots, multiple regression analysis, confidence and prediction intervals and significance tests. 1.5ES, 1.5ED. P, 305, CR 330L.

Introduction to Probability and Statistics, J.S. Milton and J.C. Arnold Laboratory Notes for Statistics and Experiment Design, J.S. Ramberg Two Level Factorial and Fractional Factorial Experiments, J.S. Ramberg.

Statistical Thinking, Britz, Snee

John S. Ramberg, Professor of Systems and Industrial Engineering

1. Probability, conditional probability
2. Discrete and continuous probability distributions, including the binomial, Poisson, exponential and normal
3. Expectation, Population mean and variance
4. Sampling distributions, including the normal, student’s and chi square
5. Confidence intervals and significance tests for single population parameters - mean, variance and standard deviation
6. Graphical and numerical summarization of data, including histograms, box plots, and stem & leaf plots

Preliminary Homework (2 weeks) and selected questions on exam one.

To learn statistical thinking, beginning with the analysis of observational data, proceeding through the planning of experiments (engineering and statistical) for data collection, and including the analysis of this data with modern statistical software. Following completion of the course, the student will be able to analyze data for engineering decision making, understand the limitations of data collected in an unplanned fashion and design experiment to collect data useful for decision making.

1. Understand, conduct and analyze comparative experiments.
2. Understand and apply control charts for the analysis of observational data.
3. Design and conduct screening experiments, including graphical analysis.
4. Design, conduct and analyze complete factorial experiments using numerical and graphical methods.
5. Select fractional factorial experiment designs and conduct and analyze them.
6. Apply (multiple) regression analysis to historical data sets and planned experiments.

1. Review of SIE 305 through homework (3 hrs.)
2. Test of hypotheses - two populations - means and variances (3)
3. Graphical Portrayal of Data - with an emphasis on statistical software (3)
4. SQC and Control Charts (4.5)
5. Monte Carlo Methods (1.5)
6. Screening Designs (3)
7. Complete Factorial Experiments (4.5)
8. Simple Linear Regression (4.5)
9. Software for Regression (3)
10. Multiple Linear Regression (6)
11. Analyzing Factorial Experiments using regression (1.5)

Class Requirements:

1. Two lecture sessions (1.5 hrs.) per week
2. Two laboratory sessions (1.5 hrs) per week
3. Ten laboratories (with homework assignment) and two additional homework assignments
4. Three examinations and a final

1. Statistical software such as Systat or Minitab are required in nearly all of the laboratory assignments.
2. Spreadsheets such as Excel can be used for other assignments or as an alternative to the above.

Laboratories are scheduled each week, except during exam weeks. Experiments are carried out in our laboratory, employing paper helicopters and catapults to illustrate comparative experiments, screening experiments, complete and fractional factorial experiments, control charts, and Monte Carlo analysis. Data generated in these experiments is analyzed in our computational laboratory, using software such as Systat or Minitab.

1. Class examinations
2. Laboratory projects - individual and group
3. Graded homework

Contribution to program objectives: Goals 1, 2, 3, 4

Prepared by: John S. Ramberg    Date: April 14, 1998