Lean Six Sigma Black Belt Sample Paper 2: Analyze 1. Fishbone diagrams cluster problems into categories. Which of the following is NOT normally used as a category? Method Market Man power Measurement2. A team’s goal is to improve information flow in a payroll function. They make 25 Post-It notes, each listing an issue for further investigation. After some discussion, they group them into four categories: mandated record keeping, privacy concerns, insurance concerns, and transfer concerns. This grouping process is best described by which approach to problem solving? a. Tree diagram b. Affinity diagram c. Fishbone diagram d. Interrelationship diagram3. A Six Sigma team has been formed to improve an existing process. Which of the following tools should the team use to determine all the potential pitfalls and the actual defects that might occur? PFMEA VSM OEE SIPOC4. Non-parametric statistical methods are preferred when: Knowledge of the underlying distribution of the population is limited Your area of study is better represented by the median You have a small sample and might not even be able to ascertain that you have normally distributed data All the above5. Which of the following statements is TRUE? A high detection rank indicates higher chance of detection Higher detection potential reduces Severity rankings High Severity means that the probability of failure occurring is high Improving Detection will bring down the Occurrence rankings6. Amool Milk Powder is sold in packets with an advertised mean weight of 500 gm. The standard deviation is known to be 20 grams. The weight data follows normal distribution. A consumer group wishes to check the accuracy of the advertised mean and takes a sample of 60 packets finding an average weight of 490 g. Which hypothesis test can they deploy? Sample t test 1 Sample z test 1 Sample sign test 1 Proportion test7. A financial services institution wants to determine if their average loan processing time differs from a competitor’s claim of 5 hours. The analyst randomly selects 20 loan applications from the past 1 month and calculates the time between loan initiation and notification to customer about the institution’s decision. (Assume that the processing time data follows normal distribution). Which hypothesis test can be deployed? 1 Proportion test 1 Sample t test Paired t test 1 Sample z test8. The Quality Assurance department at a call center wants to compare the call work times on incoming calls for two teams. For each team, 300 incoming calls are timed. What hypothesis test can be used to determine whether the teams differ in their call time variation and call time means respectively? Bartlett’s test and One-way ANOVA F-test and Two-way ANOVA Levene’s test and 2 proportion test F-test and 2-sample t test9. The customer support team claims that it typically takes about 2 days to respond to a ticket. The team recorded the resolution time for 100 similar issue tickets and checked for normality. The data was not normal. What hypothesis test can be used to check the validity of the claim? 1 Sample Sign test Moods Median test Mann Whitney test 1 Sample t test10. A research and development lab wanted to check the effectiveness of three types of fabric stain removers (A, B & C). They recorded the time required for 50 samples for each of the stain removers. What hypothesis test can the research team consider to understand if one of the products is better than the others? (The time required to remove stains follows normal distribution) One may ANOVA Two-way ANOVA Nested ANOVA MONOVA11. A production manager would like to know whether there is an association between the shift (1 /2/3) and the machine used (1 vs 2) with respect to defective items produced. An appropriate hypothesis test would be? 2 Proportion test Chi-Square test 2 Sample t test Mann Whitney test12. With respect to Pearson’s correlation coefficient, one of the following is NOT TRUE: The range is from -1 to 1 It is the square root of the coefficient of determination Value of ‘-0.9’ mean perfect negative correlation Value of ‘0’ means no correlation13. Which of the following hypothesis test cannot be used to compare pre and post improvement performance: 2- sample t test Paired t test 2 Proportion test None of the above14. For Kruskal-Wallis test, the alternate hypothesis is: All medians are equal Not all means are equal At least one mean is different At least one median is different15. The Alternate hypothesis usually represents: the theory the researcher would like to prove the preconceived ideas of the researcher the perceptions from the sample data All the above16. If the correlation coefficient is 0.9, the percentage of variation in the Y explained by the variation in X is 0.81% 81% 0.9% 90%17. We have marketing data that contains the impact of three advertising medias; facebook, youtube and instagram on sales. What approach can be used for predicting sales revenue from the budget invested in the three advertising medias? Multiple linear regression Binary logistic regression ANCOVA Design of Experiments18. To understand the relationship between time on a call (continuous variable) and response (Yes/No) to the question “Are you satisfied with the resolution?”, the best tool would be Ordinal logistic regression Binary logistic regression Nominal logistic regression Multiple Linear regression19. A physiologist wants to determine whether a particular treatment has an effect on resting heart rate. The heart rates of 25 randomly selected people were measured. The people were then put on the treatment and measured again 6 months later. If the data follows a normal distribution an appropriate test to compare the before after treatment condition would be 2 Sample t test Paired t test 2 proportion test Mann Whitney test20. When the market price of gold is high, so is the market price of silver indicating strong correlation. Best take away from this example is Market price of gold going up is causing market price of silver to go up Market price of silver going up is causing market price of gold to go up Correlation always implies causation Sale of both the precious metals might me caused by some other market variable21. The statements below are FALSE, except: As the difference to be detected becomes large, sample size required becomes small (Other things kept constant) As the standard deviation decreases, sample size increases (Other things kept constant) As power decreases, sample size required increases (Other things kept constant) As Alpha increases, sample size required increases (Other things kept constant)22. An economist wants to determine whether the monthly energy cost for families has increased from the previous year’s average of $50. An increase of at least $5 is considered meaningful and the estimated standard deviation is $3. To detect a difference of 5 with a power of 0.9, the economist needs to collect a sample of _______ observations for the 1-sample t-test. 7 5 26 3223. If the p-value is less or equal to the significance level, then: We do not have enough evidence to reject the null hypothesis We reject the alternate hypothesis The results are statistically significant The results are practically significant24. Rework, over-production, inventory, and motion are all examples of Noise factors Waste Value-adds Nuisance factors25. If beta=10%, then probability of finding out a real difference is: 90% 10% 5% 95%