kirpa Shanker Tiwari

The sparsity of Effects principle is a concept in the field of Design of Experiments (DoE) that suggests that in most experiments, only a small number of factors have a significant impact on the outcome, while the majority of factors have little to no effect. This principle is beneficial to researchers in DoE as it helps them allocate their resources effectively by focusing on the factors that truly influence the response variable. By identifying and studying the influential factors, researchers can optimize their experiments, save time, and reduce costs. To illustrate this principle, let's consider an example. Imagine a researcher wants to determine the factors affecting the yield of a crop. Due to limited resources, it is impractical to test every possible factor. Instead, the researcher can employ the Sparsity of Effects principle to identify the key factors with significant impacts. The researcher selects several potential factors, such as fertilizer type, irrigation frequency, and temperature. By utilizing fractional factorial design, a method that allows for efficient testing of a subset of factors, the researcher can assess the main effects and interactions of the chosen factors while minimizing the number of experiments required. Upon analyzing the results of the experiment, the researcher discovers that fertilizer type and irrigation frequency significantly affect crop yield, while temperature does not have a substantial impact. This aligns with the Sparsity of Effects principle, which suggests that only a few factors (fertilizer type and irrigation frequency) are crucial in predicting and optimizing crop yield. By applying the Sparsity of Effects principle, the researcher can concentrate their efforts on refining the conditions related to fertilizer type and irrigation frequency in subsequent experiments, resulting in a more efficient allocation of resources and a better understanding of the factors that truly influence crop yield.

Is /is not study help the user to avoid untrue cause. Here the user will get a confirmed root cause which help further creating a plan to solve the problem and to avoid in future. It avoid confusion in root cause and help to get exact root cause in simple way. Other root cause analysis tools present like 5 why, fmea, fishbone In fishbone analysis: we generally draw a cause and impact diagram which look like bones of fish as branches. This diagram help to identify the cause for particular process /root cause. FMEA is failure mode and effect analysis where risk can be quantity using risk ranking system. 5 why is a tool to ask 5 question about why it happened to deep dive into root cause. Is/is not actually two column study one is for is and another "is" for " is not " And to ask what when where how big the problem is. Example: if there is a problem from vendor end, then it is not a problem of company . It is an external problem of vendor side. It is not a problem from delhi vendor but mumbai vendor. It is a problem with lock, it is not a problem with key, and it keeps going on. When there is need to find out exact root cause then this tool can be used

Transportation waste is movement of the product, information, or customer if it not have any value addition to the system. It occurs between the steps and this is visible type. Example: Returning back of unused raw material to store in manufacturing may be transportation waste. Trucks carrying items between two stations In hospitals patient's sometime waiting and move from one chamber to another is a kind of transportation waste. Unnecessary movement of public from one desk to another in govt. departments. Motion waste means unnecessary movement of man or machine. it is generally within step rather than between the process steps Example: Searching printer or searching for information or a device can be considered as motion waste. Stacking. repositioning, and clicking are examples of motion waste. Transferring information from one folder to another during consultation. Walking through office space to check if paper or file delivered or not.

There are multiple method to check data distribution 1. Histogram Histogram shows data visualization of a variable using bars. It present frequency per value of dataset. If bell shape curve is there we can say data follows normal distribution. In above figure first histogram looks normally distributed and second one looks uniformly distributed. 2. Box plot Box Plot is technique can be used to identify data type from samples. The Box Plot plots five variable: min, max, median, 1st & 3rd Quantiles. if at both sides of rectangle width is almost equal we can say data is normally distributed In the figure first plot has outliers and belongs to non –normal data with outliers and second (bottom) graph can be considered normally distributed. 3. QQ plot QQ plots need more understanding, QQ means Quantile vs quantile, and here we are plotting theoretical vs actual quantiles for variables, this graph is better than box plot to visualize deviation from normal distribution. Uniform distribution has many observation in both extreme however normal plot follow a straight line. 4. Kolmogorov Smirnov test Kolmogorov Smirnov test calculate the difference between the empirical distribution and the theoretical distribution and defines the test statistic based on those differences. This test can evaluate all type of distribution. 5. Lilliefors test This test is based on the above test. Here it is accepted that the mean and variance of the population distribution are estimated rather than pre-specified by the user. if p value is > 0.05 we assume normal distribution else non normal. 6. Shapiro Wilk test This test is the wonderful test and highly used when testing for a normal distribution. It has been developed specifically for the normal distribution and it cannot be used for testing against other distributions. if p value is > 0.05 we assume normal distribution else non normal 7. Anderson darling Method: This test also use widely along with Shapiro Wilk test and similarly used for detecting normal distribution. if p value is > 0.05 we assume normal distribution else non normal

Conjoint analysis is a form of statistical analysis where data generated using customer surveys that companies use in market research to know how customers rate different attributes of their products or services. for example : A car manufacturer want to know about the car from customers based on few attributes like mileage, Strength, looks, comfort, cost, space etc. Here first manufacturer need to define attributes and design survey questions and collect data and analyse survey results. Here we need to calculate relative importance of each attribute with respect to total utility attributes. This can be done by calculating difference of each attribute w.r.t. total utility of a product. Difference is called the range in the attribute’s utility values. This analysis sometime help to find out hidden drivers which is not easy to look and also it can measure preferences at customer’s level.

For 5 factors with 2 level in Full factorial need 32 runs . I would prefer full factorial if i need 3 factors interaction without any aliasing and if i have no resource or raw material constraint. However in practical situations we have limited resource and we are interested till 2 factor interaction (un-aliased) hence I will choose fractional factorial DoE with 16 trials where main factors aliased with four factor interaction , 2 factor interactions are aliased with 3 factor interactions so it will give good resolution even with fractional factorial DoE (considering 3 factor interaction are rarely exist in system). Using fractional factorial experiments will save time , manpower , resources, without much compromising resolution of design.

Cognitive Analytics is a kind of Predictive analytics, where we can analyse huge data base and other cognitive uses of data can lead to predictions for business objectives. While descriptive analytics is to describe the population data with some parameters like mean, median, mode etc. Predictive analytics is to forecast future outcomes and prescriptive analytics is to provide recommendation based on analysis. Examples: Cognitive analytics mimic human brain and based on AI. Here huge database including pictures text comments etc are used to create algorithms. Based on algorithm it predict or suggest outcome like Google assistance, SIRI and Chatbots are based on AI.

The Shainin System is develop by Dorian Shainin. It is a tool for statistical engineering and generally used in Automobile sector. Shainin also called Red –X strategy. This is typically used to high volume processes where huge database exist and ease of data availability. This system is used in parts and assembly manufacturing processes. This work on below underlying principles 1. Assumption that there are large cause of variations 2. Assumptions there is diagnostic processes and remedial actions. Steps of Shainin system 1. define the project 2. Establish Measurement system 3. generate hints 4. list probable factors 5. DoE 6. found Red –X 7. Check interactions 8. Irreversible corrective actions 9. SPC 10. monitor outcome 11. Consumer satisfaction How it is different than Six sigma Six sigma is more statistical however this is based on Statistics and more mechanistic. Shainin is systems that are developed to achieve six sigma targets Shainin systems are evidence based and covers maximum source of variations. Shainin systems generally used FACTUAL path while Six sigma used DMAIC kind of methodology. FACTUAL: Focus>>Approach>> converge>>Test >> understand>> Apply >> leverage DMAIC: Define>> Measure>> Analyse>> Improve>> Control.

Cohen's kappa is used for two raters considering same items will be rated by both raters while Fleiss Kappa used for multiple raters with a possibility of rating different items example: when in a study no of raters will be two and all the raters will rate for all the data points or observations like taste score (good , bad , neutral) then we can use Cohen's kappa

What is Abilene paradox Abilene paradox invented by J.B. Harvey, Professor Emeritus of Management at George Washington University in 1988. This explain how gap in team behaviour mainly to make consensus within team may lead to undesired outcomes. How it does adversely affects decision making Harvey shared example how in 1986 , NASA when ready to Launch a shuttle high wind forced its cancellation and delayed due to freezing temperature beyond expectation. NASA management decided that engineers must evaluate the impact of cold weather in Launch. Critical things were hided and not reported to NASA. On Jan 28, 1986 at 11:38 , NASA launch its shuttle however after 73 seconds shuttle is lost. As per Harvey this is Abilene paradox to describe the contradiction between actual desired purpose of cation and undesired outcomes. Individually , the member of group understand the problem they are trying to resolve. They know the steps necessary to achieve the desired outcomes, however they fail to transfer their individual belief to group. This lack of communication facilitates wrong decision and results in action which is contrary to the organisation desire. The results of the action lead to undesirable results, which causes team members to become dissatisfied with others and the company. This often leads to team members blaming each one. If the team does not step back and deal with the issue of agreement management, the paradox will continue to occur with much greater organizational consequences. What are ways to prevent it? He shared the idea that team must perform in a manner that contradicts what each individual believes is the right thing to do, often creating a giant issue for the organization than the original one being solved. he suggest the main issue of the paradox is the necessity of team members to avoid harm. 1. Action anxiety, which is the intense feelings one gets when thinking about acting in accordance with their beliefs. Action anxiety is often driven by what negative fantasies, which are the perceived negative consequences one believes they will receive if they follow their beliefs. Team members may think that going with the whole team group is less risky of two risks. He acknowledges that real risk is present in individual thought to follow the group or not, just as few risk is present in all conditions. 2. Separation anxiety, or the strong negative feeling of being separated from the rest of the group, according to Harvey. He suggest that we need for connection and engagement and the fear may drive some factors which lead to action anxiety, negative fantasies. Group Thinking or brainstorming is actually discussing the unfavourable events and allows room to questioning past decisions.