space.discussion.Homework+1

Homework 1 (Problems) (GWU EMSE-271)
Index | Topics (Logical Lectures) | Lectures | Problems | Readings | References | Concepts


 * Problem 1: Graphing with Minitab**.
 * Copy the data from the [|spreadsheets provided] by Professor Van Dorp.
 * Paste into Minitab
 * Look up the desired graph on the Minitab menu Graph option
 * [[image:Minitab_Menu.JPG]]
 * Select features
 * Right click on graph for additional options


 * Problem 2: Quality Control Problem**
 * Use the material from the Session 1 lecture, Slides 22-26 and 29 to solve.
 * Set up tables 1-4 in Excel (see Slide 25 for summary)
 * Table 1 is the problem definition (Slide 22)
 * The Pr(Ai) are the probabilities of find each of the sets. For this problem, .3333 since they are equally likely.
 * The Ai are the "unknown" "states of nature." For this problem the probability of an item being defective.
 * The Bi are the number of times one tests before finding a defective item.
 * The 15 cells in the bottom center of Table 1 are the probabilites of finding a defective given the Ai's at the top of the each column. Consider using the Excel POWER function to compute the not defective part of the calucation for rows 2 and beyond.
 * [[image:Session_01_-_Problem_2.JPG]]
 * Table 2 is computed from Table 1. (Sldie 23)
 * Table 3 is computed from Table 2. (Slide 24)
 * Table 4 is computed from Table 2 and Table 3 (Slide 25)
 * Slide 29 provides a sample problem.
 * **Don't forget to provide the interpretation of the final values.**
 * ** Another learning point is that if the Bi's are collectively exclusive (CE), as in Slide 29, the columns in Tables 1, 2 and 3 sum to 1 (a great double check). **
 * If they are not CE, as in Problem 2 of the Homework, then they will sum to less than 1.


 * Additional Material**

Having solved a 3-by-n problem, it is not difficult to solve a n-by-n problem. My review notes from EMSE 271, Spring 2008 provide and example.
 * Note that we could use the Excel COMBIN function to create the Bi impact part of the values for Table 1 cells.
 * Also note that different distributions can be used to create the values for Table 1, **it depends on the problem.**
 * **I'm not sure if the Ai's have to be CE. I was thinking so, but now I guess not. (CHECK LATER)**