You are on page 1of 7

Question 2: Explain with an example the Root Cause Analysis methodology.

Solution: Root Cause analysis (RCA) is a method of problem solving involving identification and evaluation of the basic reason for non-conformance, an undesirable condition, or a problem which (when solved) restores the status quo. It is a framework for determining possible causes of variation (families of variation)in your output variables and depicting the results of your analysis. It focuses on identifying and correcting the root causes of events, as opposed to simply addressing their symptoms. RCA is a reactive method of identifying causes, revealing problems and solving them. Analysis is done after an event has occurred. But sometimes the insights in RCA may make it useful as a preemptive method. In that event, RCA can be used to forecast or predict probable events even before they occur. And it takes multiple rounds of analysis for coming with the root cause and so RCA is an iterative process and a tool of continuous improvement. Steps Involved in RCA

Step 5 Step 4 Step 3 Step 2 Step 1


Define the NonConformity Methods for RCA Pareto Chart Investigate the Root Cause Create Proposed Action Plan & Define Timescales Implement Proposed Action Verification & Monitoring of Effectiveness

Pareto Analysis is a simple technique for prioritizing possible changes by identifying the problems that will be resolved by making these changes. By using this approach, you can prioritize the individual changes that will most improve the situation. Pareto Analysis uses the Pareto Principle also known as the "80/20 Rule" which is the idea that 20 percent of causes generate 80 percent of results. With this tool, we're trying to find the 20 percent of work that will generate 80 percent of the results that doing all of the work would deliver.

5 Why

By repeatedly asking the question why?' (use five as a rule of thumb), you can peel away the layers of an issue, just like the layers of an onion, which can lead you to the root cause of a problem. The reason for a problem can often lead into another question; you may need to ask the question fewer or more than five times before you get to the origin of a problem. The real key is to avoid assumptions and logic traps and encourage the team to keep drilling down to the real root cause.

Fish Bone

A commonly used method of root cause analysis, is the use of fishbone diagrams, or Ishikawa diagram. They are most useful when a complex issue needs to be broken in small size issues or where there is a lot of data that needs to be analysed. In the diagram, the various causes are grouped into categories (such as equipment, materials or processes) and the arrows in the image indicate how the causes cascade or flow toward the nonconformity.

The benefits of Root Cause Analysis include: Identification of permanent solutions Prevention of recurring failures Introduction of a logical problem solving process applicable to issues and nonconformities of all sizes

Question 4: The data given in worksheet Ph is ph value of a chemical collected from various batches. The specification is 4.005 0.005. Estimate the process performance in terms of sigma level. Ans: The process performance in PPM from the figure below is 40268.46 and from the appendix the corresponding sigma level is 3.25.

The process Capability Indices are also calculated to understand whether the process in under control or not: Specification Limit = 4.005 +/- 0.005 Mean of all Ph values = 4.00168 Standard Deviation = 0.005757525 Tolerance Limit = 4.00168 +/- 0.0057 UNTL: 4.018 , LTNL : 3.9844 Thus we can calculate the Cp,Cpk,Cpm values as mentioned below:

Cp= Cpk= Cpm= Yield=

0.28947625 0.09726402 0.277887832 92.51

Question 8: The following data refers to 4 locations doing the invoice processing. The number of invoices having errors and no errors was collected. Do a suitable analysis to find whether the number of errors is independent of the location? Which Location is contributing towards the error most? Location A Error No Error 18 121 B 17 240 C 28 134 D 36 237

Ans: As the independent and dependent variable both are continuouswe will use Chi-Square Test to check the dependency. Below attached is a snap-shot of the Minitab output when Chi-Square Test is performed on the above worksheet. Findings: 1. P-Value = 0.008 < 0.05,hence we reject the null hypothesis. The test-statistic carries a very high value of Chi-sq 11.892, which is also a reason to reject the null hypothesis. It is found that in location C the difference in the expected value and actual value of number of errors is high. Thus it can be safely concluded that Location C is mostly contributing towards the errors.

2.

Question5: The data in worksheet MPG contains information on 11 variables for n = 93 makes and models of automobiles. It is desired to construct a model to predict MPG Highway. 10 possible predictor variables were considered. They are 1. Horsepower, 2.Fuel tank, 3.Passengers, 4.Length, 5.Wheelbase, 6.Width, 7.U-Turn Space, 8.Rear seat, 9.Luggage and 10.Weight. Fit a parsimonious model, i.e., a model involving as few variables as possible, provided none of the omitted variables would add significantly to the predictive capability of the model. OUTPUT FROM MINITAB: Regression Analysis: MPG Highway versus Passengers, Length, ... The regression equation is MPG Highway = 14.3 - 0.976 Passengers + 0.0317 Length + 0.292 Wheelbase + 0.319 Width - 0.151 U Turn Space + 0.079 Rear seat + 0.106 Luggage - 0.00895 Weight - 0.0021 Horsepower - 0.511 Fueltank

Predictor Coef SE Coef T P Constant 14.25 11.24 1.27 0.208 Passengers -0.9758 0.5553 -1.76 0.083 Length 0.03168 0.04273 0.74 0.461 Wheelbase 0.2918 0.1222 2.39 0.019 Width 0.3193 0.2007 1.59 0.115 U Turn Space -0.1507 0.1691 -0.89 0.375 Rear seat 0.0789 0.1548 0.51 0.612 Luggage 0.1064 0.1473 0.72 0.472 Weight -0.008948 0.002068 -4.33 0.000 Horsepower -0.00210 0.01259 -0.17 0.868 Fueltank -0.5105 0.2143 -2.38 0.019

S = 2.84869 R-Sq = 74.6% R-Sq(adj) = 71.5%

Analysis of Variance Source DF SS MS F P Regression 10 1949.88 194.99 24.03 0.000 Residual Error 82 665.43 8.12 Total 92 2615.31

Source DF Seq SS Passengers 1 568.87 Length 1 342.83 Wheelbase 1 101.94 Width 1 160.65 U Turn Space 1 26.24 Rear seat 1 3.42 Luggage 1 27.25 Weight 1 670.62 Horsepower 1 1.99 Fueltank 1 46.07

Unusual Observations MPG Obs Passengers Highway 5 4.00 30.000 23.933 17 8.00 20.000 17.648 19 2.00 25.000 26.753 28 4.00 24.000 19.278 36 7.00 20.000 23.869 39 4.00 50.000 39.801 42 4.00 46.000 38.020 60 4.00 26.000 32.995

Fit SE Fit Residual St Resid 1.119 6.067 2.32R 1.780 2.352 1.06 X 1.939 -1.753 -0.84 X 1.666 4.722 2.04R 1.723 -3.869 -1.71 X 1.022 10.199 3.84R 1.034 7.980 3.01R 1.136 -6.995 -2.68R

R denotes an observation with a large standardized residual. X denotes an observation whose X value gives it large influence.
Histogram of MPG Highway
Normal 20
Mean StDev N 29.09 5.332 93

15
Frequency

10

18

24

30 36 MPG Highway

42

48

The R square value of 0.746 (74.6%) indicates that the model fits the data very well. Moreover, the very low P value (0.000), which is less than 0.05 shows that some of the independent variables being studied here are definitely significant. If we consider only Rear Seat and Horsepower as they have very high P values then the regression Output becomes: Regression Analysis: MPG Highway versus Passengers, Length, ... For the variables ignoring correlation >0.7 The regression equation is MPG Highway = 46.7 - 2.72 Passengers + 0.0007 Length + 0.176 Rear seat

+ 0.049 Luggage - 0.0663 Horsepower

Predictor Coef SE Coef T P Constant 46.733 5.817 8.03 0.000 Passengers -2.7202 0.4800 -5.67 0.000 Length 0.00069 0.04049 0.02 0.987 Rear seat 0.1762 0.1771 1.00 0.322 Luggage 0.0491 0.1696 0.29 0.773 Horsepower -0.066295 0.008881 -7.47 0.000

S = 3.45380 R-Sq = 60.3% R-Sq(adj) = 58.0%

Analysis of Variance Source DF SS MS F P Regression 5 1577.51 315.50 26.45 0.000 Residual Error 87 1037.80 11.93 Total 92 2615.31

Source DF Seq SS Passengers 1 568.87 Length 1 342.83 Rear seat 1 0.75 Luggage 1 0.29 Horsepower 1 664.77

Unusual Observations MPG Obs Passengers Highway 19 2.00 25.000 27.895 28 4.00 24.000 20.152 39 4.00 50.000 37.648 42 4.00 46.000 34.733 57 2.00 25.000 29.979 60 4.00 26.000 32.980

Fit SE Fit Residual St Resid 2.138 -2.895 -1.07 X 1.833 3.848 1.31 X 1.062 12.352 3.76R 0.756 11.267 3.34R 1.639 -4.979 -1.64 X 1.291 -6.980 -2.18R

R denotes an observation with a large standardized residual. X denotes an observation whose X value gives it large leverage. The equation below shows the best model without significantly affecting predictive capability of the model. The regression equation is MPG Highway = 46.7 - 2.72 Passengers + 0.0007 Length + 0.176 Rear seat + 0.049 Luggage 0.0663 Horsepower

You might also like