Vishwadeep Khatri

Defectives vs DPMO vs Yield vs Sigma Level Defectives as a data set does not make use of opportunities for error existing in a process. Defective is an entire unit while defects are the number of non conformities, which might be more then one in a unit. Customers generally prefer using defectives data to gauge the performance of a process. Yield (non defectives) is likely to be their prime metric. Process owners need to be detail oriented and generally prefer to use DPMO as their prime metric. A special case - If there is just one opportunity for error per unit, number of defects and defectives is the same. In such a case, DPMO method provides same sigma level as yield method and conversion is simple - if DPMO is 10000, yield is 99%. Hope this helps.

Hi. Mohini, Lean Six Sigma is a set of approaches that is used for business improvement. It commonly follows a DMAIC cycle and is a systematic sequence of problem solving. If you reword your question to highlight a business problem in a power plant, I shall be able to provide a more specific reply.

Please use the DPU method in this case. Errors are defects as there can be many of them in a single documents. You have correctly pointed out that the opportunities cannot be counted.

Interesting question. The flaw is - correlation and regression should only be used for variables that have a known cause and effect relationship. In this case we should not have used these tools.

Thanks for participating in this discussion. I agree with comments above in the context they are presented, but want to add a caution. And the caution is - we should just be talking only of occasional occurrences. An occasional false alarm is fine but too many of these can lead to loss of business or even closure of a company. As an extreme example, let us consider an automotive company that decides to do 100 percent inspections on all parts leading to high cost of product which finally leads to loss of market. We need to also consider that a false alarm can lead to process adjustments when none is required in an existing good process. Comments welcome.

With respect to complex process improvement projects - One argument goes like this - "Most of the times, it is not difficult to find a solution to a problem. The most difficult part is nailing the root cause. If the root cause is unclear or misunderstood, the attempts at solution do not work; it just seems that the solution is so hard to find. However, if the root cause is understood precisely, the solution shall just come and sit on your lap. (or you will know soon that there is no solution if the root cause is absolutely out of your span of control)". Another argument goes like this - "In most projects, the root cause of problems are found spending lesser time and effort. It is the sustainable solution that is tough to find. One needs to try various alternative solutions and this means some sort of trials/ pilots/ benchmarking/ brainstorming/ simulations are needed. More energy therefore needs to spent in evaluating alternative solutions so that we choose nothing but the best" If you dig into your experience and think hard on this, which point of view would you support over the other?

I assume that you have carried out these tests one by one (also called OFAT - one factor at a time). Due to feasibility of interactions (combined complex effects), it is not advisable to combine these results to reach a generalized conclusion. To be able to find a combination in a candidate that has the highest probability of passing, it is advisable to design experiments that use various combinations. Also, in such an experiment it is better to take the output as exam score instead of just a pass/fail outcome. Hope this helps.

Thanks for your response. Discussion in a forum is likely to have some risks of misunderstanding. To be sure that I understand clearly, I request you to provide details of the process and the explanation of categories.

Hi Ashish, A p value of zero indicates that the two extreme categories (the one with lowest success rate and the one with highest success rate) have a significant difference. You may like to remove one of these and run the test again to see if others show a different result. If you again get a zero p value, you can continue removing an extreme value and re-run the test one-by-one and draw conclusions. Hope this helps.

Hello Nagarajan, Here is some useful material to go through -http://globalnews.benchmarksixsigma.com/category/new-to-lean-six-sigma/lean-expert-basics/ Best Regards, VK

There has been no question so far on this article, here is one for the writer - "If data is used less frequently for decision making in a company where processes are mature and self adjusting in nature, the company may survive for a longer period. Comparatively, a company that uses data very frequently for decision making but the type of data is not forward looking or provides limited information, this company may be worse off." What is your comment on the above?

Dear Tarun, good article. Do you think there are situations where decision making should not use data? Are there situations where data analysis may be considered detrimental? Provide examples to support your view.

Kongkan, good work. Here is my question for you. If data driven decision making is so suitable for complex situations and it takes time to build intuition, why did Deep Blue - IBM's chess computer lose several games with Kasparov? After all, it had the capability of evaluating 200 million positions per second and more man years behind it as compared to Kasparov. http://en.wikipedia.org/wiki/Deep_Blue_(chess_computer)

Nachiketa, this is a very good write up. One query I have is this - You have mentioned that " D3M does not save time or money, at least not in the short term." Is this statement data driven? How?

Hi Suhani, very good article and good responses so far. Referring to your examples of stock markets, why is it that mostly those with deep pockets survive in markets? Why is it that most of the stock advisers/ experts like to spend their time making recommendations for others while the money at stake is not theirs? Don't you think that the data crunching experts should ideally be busy in higher value activity of multiplying their own existing money?

Hi Vignesh, Yours is a well written article. If you were involved in creation of a new product (in an entirely new category) where past data and predictions using past behaviors seem irrelevant, would you still recommend use of data in some way or just go with gut feel of the business leader?

Excellent article. I have a question - How do you suggest use of data in decision making when the product belongs to an entirely new category with no antecedents? Forecasts are unlikely to work and gut feel probably reigns supreme. The question is important as the biggest profits in the corporate world are derived from innovations. A response with examples will be appreciated.

Hi Raghavendran, If you do not have support or control in an area, it is better to look at other areas. So taking up a project in operations should be explored. Is there a pain area in operations that needs to be tackled, has good support and creates a good business case? Once you find one, you can progress further. Doing a project with a shop keeper is also fine. The lean six sigma methods do work for identified pain areas and here you have found a business case. Regards, VK

Tanmoy, How did team leaders or managers reach those positions? Many of them have been callers initially. I know several people who have risen from callers to the position of AVP. No one gets to a managerial position unless someone notices his aptitude. Aptitude shall not be noticed unless one thinks, communicates and works differently. Developing and demonstrating managerial aptitude requires some hard work. If learning and demonstrating managerial aptitude does not seem to help very soon, will not having it help at any time? It requires perseverance, patience, dedication and hard work to get to some hard goals. You have already taken the first step. Have a look at this - http://forum.benchmarksixsigma.com/page/media.html/_/ites-bpo-and-kpo/six-sigma-in-bpo-asad-siddique-r29

Hi Jayant, This is a good article and seems absolutely on track from start till end. If we look at the "how" in addition to the "why", there are several challenges. There are so many cases where right data was used but the inferences were wrong. Here is a link that highlights some examples - http://www.datavis.ca/gallery/missed.php Most business leaders do not know how to find the right sample size for carrying out a hypothesis test and end up concluding incorrectly. Many organizations rely on external statisticians who draw conclusions which may be right but carry little business value. There are other kinds of practical risks of promoting this approach. In a country like India - Researchers, Scientists, Artists, Designers, Chefs and the like tend to hide more than they share. At many occasions, they provide incorrect or incomplete information to others for the sake of job security. The question is this - If you wish to promote the right way of collecting and using data, what steps would you recommend for an organization? What could be the initial steps and what could be some of the simpler tools to use in the beginning? Who should be learning and using these methods initially?

Dear Rohang, The article is very well written. However, the last two sentences create a puzzle. Are you referring to some specific unit here? If yes, which unit is it and why is it being referred as "our unit"?

Hi Prateek, I am trying to understand your views of using both worlds at the same time. Are you trying to say that gut feel can be data driven, or that data driven decisions can be subjective? In my opinion, if your decision follows your intuition, you do not care about data. If you are a follower of data, you cannot go with your hunch. Your examples show that sometimes gut works and sometimes it does not. If you create hypotheses using gut feel, but finally take decision based on data, it is not an intuition based decision finally and shows that you are a supporter of DDDM. Or am I missing something?

I have not seen any post on this article so far. I thought it is apt to weave a response from the manager who just got your letter. Here is what the manager writes back - "Well, I can see certain comments in your letter which indicate that I have some inexpressible psychological factors and a feeling of insecurity. From the style you have adopted, it requires experience to indicate that your style of writing fails to create an environment of mutual respect. Being a believer in intent and final impact of actions, I have doubts over what you are trying to achieve. On the one hand, you mention my job as a manager being wonderful, and on the other hand you mention baselessness in my views. While your language may influence a reader searching for impressive english, it lacks the same element that you are trying to support, namely, evidence. While so many innovations and new products are created through ingenuity, I do not see great results coming out from data driven decisions. While I do see some merit in your arguments, my gut feel has resulted in good performance so far, as I KNOW the markets and its sentiments. There are numerous occasions where proponents of data usage have failed. Today, markets change with every competitive launch and data mostly shows things from the past and expectations derived from experience. I know for sure that companies like Nokia focused on data analysis while others like Apple and Samsung changed the entire market with forward thinking initiatives. I stick to my views as strongly as ever before and hope that with deeper insight, you shall understand that emotions drive people more than logic. It is because of emotional intelligence that I have led a big team successfully for such a long time. I hope my reply helps in putting things in right perspective. Let me know if you have any evidence driven arguments."

Team leaders benefit more by doing this program. For a caller, this program can help show the interest and aptitude in Management. It can help in initiating small improvements. You shall be able to bring bigger process changes only when you are allowed to. You may be allowed to work on bigger changes, if you have the potential and are able to present the aptitude well.

Dear Vishnu, LSSGB is a value added qualification which provides you a distinctive edge over other candidates for the job. It may not fetch you a higher salary than others but an edge over others vying for same position. For a job in Quality department, it enhances your awareness levels about most powerful methods in use today in corporate world and shows to the employer that you are willing and committed to building competence for such job. It also shows that you are more job-ready than others.