Deck 13: Experimental Design

Question 1

Taguchi's loss function:

A) combines cost, target, and variation into one metric
B) has 8 different equations
C) is concerned with production losses in dollars due to waste of raw material, energy, and labor
D) none of the above

Answer 1

A

Question 2

Six samples of a quality characteristic give an average of 5.0 and a variance of 0.4. If the loss is $8.00 at a characteristic value of 6.0, what is the loss equation for smaller- the-better? What is the average loss?

Answer 2

L = 0.22 y², L_avg = $5.54

Question 3

Repetition is defined as:

A) a repeat of a treatment condition that requires a set up
B) multiple results of a treatment condition
C) treatment conditions run in a random order
D) a and b

Answer 3

B

Question 4

A simple method of working with interactions is:

A) linear graphs
B) Taguchi array
C) interaction table
D) none of the above

Question 5

Predict the outcome of the confirmation run for a four factor larger-the-better
experiment with an overall average of 18 for the data below:

Question 6

Using inferior-grade, low cost components, and raw materials in the design process is referred to as:

A) tolerance design
B) parameter design
C) Taguchi array
D) syatem design

Question 7

A range or band that the true value is presumed to occur is referred to as a(an):

A) point estimate
B) orthogonal design
C) interval estimate
D) t reference distribution

Question 8

Given the following ANOVA determine the percent contribution at the 0.05 level for F table values of 1/2 = 18.5, 1/3 = 10.1, and 1/4 = 7.71.

Question 9

A balanced experiment is efficient-it is called:

A) fraction factorial
B) orthogonal design
C) analysis of variance
D) full factorial

Question 10

The primary goal(s) of a designed experiment are(is) to:

A) determine the variable(s) and their(its) magnitude that influences the process
B) determine the levels for these variable(s)
C) determine the appropriate diagram
D) a and b

Question 11

A technique to compare the F value of the factor/level with a F distribution value is called:

A) fraction factorials
B) orthogonal design
C) analysis of variance
D) full factorials

Question 12

The process of selectively upgrading components to eliminate excessive variation is called:

A) tolerance design
B) parameter design
C) Taguchi array
D) syatem design

Question 13

The t test is used when:

A) the sample size is small
B) the distribution is skewed

Question 14

A new process is developed to reduce the number of defects. Three tests are made with results of 2, 4, and 1 defects. Determine the S/N and compare with a dB value for the existing process of -1.356.

Question 15

The concept of signal to noise was developed by:

A) Ishikawa
B) Fischer
C) Deming
D) Taguchi

Question 16

Given three two-level factors and two suspected interactions, determine the degrees of freedom and the OA?