Question 8 . Control limits were originally defined at the three-sigma level because: This level provides a good balance between the risks of having “false alarms” and missing special causes This level establishes tight limits for the process to minimize variation This level provides only 3 false alarms out of 1,000 This level makes it difficult for special causes to occur None of the above
Question 9 . In evaluating data on our process outputs, four characteristics we might investigate are: central tendency, variation, shape of distribution, and stability. Which of the following tools would be most helpful to determine the shape of the distribution? Scatter plot Histogram Pareto chart Run chart None of the above
Question 10 . A process is said to be capable if: The process consistently meets the customers needs The common cause variation is less than +/- 3σ The common cause variation fits within the specification limits The process is experiencing no special causes None of the above
Question 11 . Tips for building useful models include: Reducing subjectivity by using computer algorithms to select model predictor variables Plotting the raw data and model residuals in a variety of ways Focus on keeping the models simple, not complex Validating the performance of the model A, B and C above B, C and D above None of the above
Question 12 . Results of regression analysis are often abused in the following ways: Using the model without understanding the context in which the model was developed and recommended for use Predicting outside the region of the data without noting the predictions are an extrapolation of the model Ignoring known scientific and economic theory regarding the model and its variables All of the above None of the above
Question 13 . An Adjusted R-square value is a correlation coefficient that has been modified to account for: Non-linearity of the model Number of degrees of freedom in the error term Magnitude of the residuals Number of predictor variables in the model None of the above
Question 14 . Which of the following is NOT a measure of model adequacy? Residuals have constant variance Residuals have mean = 0 Residuals are normally distributed Residuals show no patterns or trends None of the above
Question 15 . The model residuals are examined to: Check the adequacy of the model Check the assumptions of the analysis Identify ways to develop a better model if needed All of the above none of the above
Question 16 . Which of the following is the best description of randomization? A means of assuring parallel experimentation A technique used to increase the validity of an experiment The repetition of an observation or measurement The relationship between two or more variables None of the above
Question 17 . An aspect of good experimental strategy is to study the effects of the variables (Xs) over a wide range. This strategy increases the chances that effects will be found because: The sample size is increased The size of the effect relative to the experimental variation is increased The tests can be run faster Higher quality data are collected None of the above
Question 18 . Given the plot below, what might you suspect about factors A and B?
The effect of A is not influenced by the level of B The effect of A depends on the level of B A has a larger effect on the response than B B has a larger effect on the response than A None of the above
Question 19 . The basic reason for randomness of sampling in design of experiments is: Increase the chances that the sample represents the population Speed up the sampling process Guarantee to reduce the cost of inspection Guarantee correct interpretation of data None of the above
Question 20 . What should one consider when analyzing the results of an experiment? Understanding of the science, technology and operation of the process being studied Creating graphical displays to aid in the analysis and interpretation of the results of the experiment Performing a statistical analysis to test the significance of the effects studied in the experiment All the above None of the above
Question 21 . I have made a mistake in copying data into my computer to perform a large regression model with 20 independent variables. Basically, I ended up with random data in each of the columns. However, I accidentally ran the regression anyway. Making reasonable assumptions, which of the following outcomes is most likely? We find none of the regression coefficient p-values were below .05 We find one of the regression coefficient p-values was below .05 We find two of the regression coefficient p–values were below .05 We find several of the regression coefficient p-values were below .05 None of the above
Question 22 . After running an ANOVA comparing the average years of experience between five different job classifications, we obtained a p value of .02. Which of the following would be a reasonable conclusion concerning the population in this case? We have strong evidence to state that the average years of experience between these job classifications is the same We have insufficient evidence to state that the average years of experience between these job classifications is the same We have strong evidence to state that each of the job classifications has a different average age We have strong evidence to state that at least two of the job classifications have different average ages None of the above
Question 23 . Which of the following could be considered a prediction interval? An interval documenting uncertainty in an estimated regression coefficient An interval documenting uncertainty in the average of the next month’s daily revenues An interval documenting uncertainty in the difference between the average population cycle times in approving expense accounts between two business units An interval documenting uncertainty in the long-term return rate (%) of a consumer product ordered over the Internet None of the above
Question 24 . A local university ran a regression model on factors impacting starting job salary of employed graduates. After validating this as a reasonable model, the university noted that the regression coefficient for GPA had a p-value of .10. Which of the following would be a reasonable conclusion in this case? We have convincing evidence that GPA does have a relationship with starting salary We have convincing evidence that GPA does not have a relationship with starting salary We have insufficient evidence to conclude that GPA has a relationship with starting salary We have insufficient evidence to develop a confidence interval for the regression coefficient for GPA None of the above
Question 25 . We ran a taste-test to see which soft drink employees in our company prefer. We had 100 employees, selected as randomly as possible, taste test two brands and determine which they preferred. Which of the following would be a reasonable statistical analysis to determine if there is a clear preference among our employees? A two sample t-test A paired t test A confidence interval for a proportion A prediction interval for a proportion None of the above
