Vishwadeep Khatri

Simple answer - Go for the best - Calendar is here - http://www.benchmarksixsigma.com/calendar

That is DMAIC in another set of words.

Dear Mr. Agarwal, While you are head of operations, it should be feasible for you to carry out a project or two immediately after black belt training. Once you have done projects, you should have options to take up new job or consulting work. Let me know If you have any questions.

Hi Mahendra, When the objective is to find the critical X (root cause), multiple regression should be useful. Testing multi-collinearity is a key element in finding the right contributing factors. I can see that you have explained the usage of the tools quite well.

How many insect fragments or Rodent hair are permissible in food items. Just search for those terms here - http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/SanitationTransportation/ucm056174.htm Any comments?

Science can be bad. Data can be made to lie. Have a look - http://www.ted.com/talks/ben_goldacre_battling_bad_science

Dear Mahendra, One useful approach in fraud detection is the use of Benford Law. Here is a link that should help - http://www.statistics.gov.hk/wsc/CPS021-P2-S.pdf

Hi, Yes, it is just a rounding off error that causes the difference. The formula in excel uses normsdist and not normdist.

To state it simply, the more assessors you have, the harder it is to achieve an acceptable measurement system.

I agree to your points. Lean Six Sigma is not just about process improvement. It is also about process design. Designing a sales process for a new offering is an important challenge that companies face very often. This is area offers an immense opportunity for DMADV kind LSS projects. In recent times, I have worked with two well established organizations in digital marketing who found absolutely new opportunities while utilizing Lean Six Sigma.

Hi Syam, Hope you have benefited with your Lean Six Sigma learning. Do share your experiences

Hi Chandrasekar, I - MR chart is designed for continuous data but used often for discrete data as well. If one does not know which chart to use, I-MR chart is a good starting point.

Hi Rajiv, One option is to carry out a 5S audit and then have an internal brainstorming session to review the scores and worthwhile action areas.

Dear Pavan, Here are some answers. 1. Continuous data requires less number of samples as compared to discrete data in a similar sitiuation. 2. Sample size depends on the hypothesis that you are trying to test. Best Regards, VK

Dear Vaseem, Please note that Benchmark Six Sigma does not provide information on Industry Benchmark that you are seeking. We provide Business Improvement Programs of various kinds for which you can visit our website. Best Regards, VK

The difference with data analytics is highlighted by a very popular definition - by Michael L George. "Lean Six Sigma is a business improvement methodology that maximizes shareholder value by achieving the fastest rate of improvement in customer satisfaction, cost, quality, process speed, and invested capital" Lean Six Sigma provides an entire approach for business improvement including methods of identifying, defining, measuring, analyzing, improving and controlling. An "analytics" person may contribute in some of these phases (mostly (define and analyze) Hope this helps.

Hi Rajiv, Attrition is one of the most pressing problems in many companies. Here are some of my thoughts and suggestions. I presume that the exit interviews were done in a congenial atmosphere with promise of anonymity. There are sensitive causes that people will not expose unless they are comfortable. There should be multiple ways of doing Pareto's analysis (other than by causes - which you have done). For example you may do this analysis by tenure, by function, by job goals, by salary bracket etc. Please try some of those which you find meaningful. While doing Pareto's analysis with reasons, there are chances that the categories are not mutually exclusive - Someone may have left because of multiple reasons (though may have chosen to mention only one). Hope this helps. Best Regards, VK

Hi Srinivasan, I presume you are referring to something like productivity. If that is the case, it is not feasible to have upper control limit but you may not have an upper specification limit. Specification limit is a value beyond which the outcome is called defective. Hope this helps.

Dear Ashok, My email id is vk at benchmarksixsigma dot org. Best Regards, VK

Hi Srinivasan, good question. Your sample size to be audited shall depend on various factors, some of these are - Are you going to collect data for something like handling time (continuous) or for accuracy (attribute data) for calls? Sample size is smaller for continuous data and bigger for attribute data. How the performance has been in the past? If you have information from the past, it might help in reducing sample size. What is the objective if taking samples? You might be comparing the performance with an external standard, for example or you may like to compare different sets of data. Size depends on this. How much error you are willing to tolerate? Reducing error shall mean more samples needed. Finally, how much difference (from standard value or between sets of data) you wish to detect. If you are trying to detect smaller differences, you need more samples.

Hi Arun, a requirement not met is a defect. So, for Purchase Department, an incoming non conforming material is a defect, non-conformities missed in receiving inspection is a defect, good material rejected is a defect, not following the specified procedure is also a defect. It is hard to specify an action to reduce defects at this point as your action shall depend on the kind of defect and the surrounding information that you can gather when you are at the workplace. Hope this helps.

Hi Ashok, yes, that was a question posted in October 2013. It missed our attention and I responded it recently.

Hi Gopinath, I am not sure if your organisational circumstances provide an opportunity for customer to provide rating or appreciation for identified individual employees. If yes, is there a possibility of keeping it unbiased or beyond influence. Measurement of delight is highly dependent on perception and I feel we should get it from the horse's mouth. If the above is feasible, an "above expectation" feedback qualifies for customer delight while "met expectation" is satisfaction - I am just using hotel industry approach here. Hope this helps.