All entries for Friday 03 May 2019

DECISION MAKING WITH STATISTICS

Today is the start of a very interesting day. We are doing decision making with statistics today. I was waiting for this module. The start was of the basic tools such as mean and mode. We learnt in Six Sigma that we should not rely just on the averages. We can use a combination of mean and modes to analyse the data further. The calculation of variation in the data in always beneficial to make the right decisions. Also it is very important to understand the variation with regard to the six sigma control limits to understand the process better.

It is interesting how various tools like geometric mean can be used to calculate the value of investments in future compared to the present value. But for that we need to have the values of the expected return on the investments. Again, nothing can work just on the assumptions. There has to be either probabilities or data to help in these calculations.

The other very interesting aspect is the use of normal distribution in statistics. The way vagelis portrayed it in the class it was very helpful. The data was collected from the production process and the weight of the product was measured. It was very beneficial to see how the concept of normal distribution that I learnt in my Undergrad was helpful to see the probability of the weights of the product. Calculating this way I can use this statistics to absolutely measure how many defected products I expect and what is the probability of getting that. Also, helping to know the aspects of the product that need to be worked on. This normal distribution can be used to model the whole production process and look at the normal deviation of it. Such accurate data, will help the company to understand their process and accurately tell their suppliers and stakeholders on what can be expected from the company. Such accurate awareness of own processes can act as a competitive advantage for the organisations and help win contracts.

Until now, what I am understanding is that more practical application of these tools is required. But it is something which is not much difficult, it will come with right guidance and experience. The main trick is to identify which distribution fits to solve what kind of a problem. I liked how hypergeometric distribution can be used to check the probability of the defective products. If we have agreed with a contract that only 5% of the products will be defective out of population of 1000 pieces. Then me as a customer, I can use the sample of the population to determine whether the claim given by the supplier is right or wrong. I have done the concept of confidence limits in my under graduation as well. But it is important to understand how this hypothesis testing will be used in the decision making in the organisations.

I can relate this application with the average width of the sole in shoes. This hypothesis testing can be used to measure the average of the sample and then try to prove the hypothesis whether this sample average is close to the average width of shoe population. This way we can make a judgement whether to sell the batch or rework on the whole batch. The decision making with statistics has been really beneficial for me to understand the practical implication of these tools.

Saurabh Kukreja : 03 May 2019 11:42 |  Comments (0) |  Close comments |  Report a problem