In: Statistics and Probability
Solution:
An important element in the growth of any business or business organisation is having a sound and effecient policy and decision making mechanism which drives revenue and growth. For efficient policy making, organisations need to rely on precise data driven mechanisms, Six Sigma tools are a group of disciplined, data-driven methodologies aligned with a statistical approach of analysing the data for aiding clearful business insights and enhancing policy and decision making.
Six Sigma methodology is an approach to convert a business problem into a statistical problem and applying statistical tools and techniques for analysing the information available in terms of data. The solution to the problem is in terms of statistical inference & outputs related to the data and finally the statistical solution is converted into a business solution for ease of understanding by the business domain experts.
The Six Sigma method involves breaking down the business problem in 5 stages, namely DMAIC (Define, Measure, Analyse, Improve and Control). But most of the statistical analysis is carried out in the 2nd and 3rd phases, namely Measure and Analyze phases.
1. Measure Phase: In a business, all major policies are done based on certain measures which the business planners and decision makers are interested to check, which known as "metrics". As the data needed to be worked upon are already clarified in the "Define" phase, the next procedure is the "Analyse" phase where we perform statistical analysis on the metrics calculated from the data available.
The "Measure" phase quantifies the business problem and provides the initial data analysis. In Statistics, the first phase of working with data is performing Descriptive Analysis, which gives a summarization report on the data available. The summary measure in statistics involves a 5-point summary, namely the minimum, 1st quartile, median, 3rd quartile and maximum values of the concerned variable(s) within the data. Along with the summary measures, the measures of central tendency and dispersion, like mean, range, variance, standard deviation, etc are also calculated to get an idea of the location and spread of the data.
The analysis of the descriptive measures give an idea how the problem behaves, what is the scope of the analysis to be carried out, and validation of the problem.Visualization of the problem based on the data and the measures calculated are also carried out. Some of the plots carried out (although it varies depending upon the nature of the data, the measures and the business domain worked out on) are histogram, pareto charts, frequency polygons etc..
For example, a histogram is used to measure a frequency distribution data into a plot. For example, if we are to improve patient servicing time by doctors in a clinic, a histogram may give us an idea how many patients are served within a time span of 5 minutes, 10 minutes, 15 minutes and 20 minutes intervals. We can have an understanding of the distribution of such event (normal or other) and see how many are served within the time limits and outside it. Data accuracy is another important factor to be checked upon to have a reliability on the measurements done. Six-sigma is a data driven process and hence measurements must be accurate to a very high level.
2. Analyze Phase: The analyze phase is the most statistics dependent phase of the DMAIC problem. The problem or the issue concerned is analysed using the statistical techniques like regression, hypothesis testing and other methodologies to identify the root cause of the problem. This is also known as Root Cause Analysis. Stability of the process is also checked using control charts to check whether the issue is under control or not or to detect any defects associated to the problem. Process Capability Analysis is a powerful collection of statistical analysis used to perform process improvement and capability. Some other statistical analysis used are:
a) Regression Analysis – This tool helps estimate the impact variables in a process have on each other and on the final product. It allows the project team to measure how well the theory fits the data.
b) ANOVA – This statistical technique tests three or more groups of data. It starts with a null hypothesis stating there is no significant difference between the groups. It then tests variation between the groups of data and variation within the groups of data. A high variation between groups of data indicates a possible root cause.
c) Chi Square – The Chi Square tests whether the difference between the expected and observed results is significant. This tool can determine whether differences in the expected and observed results are due to chance or there is an independent cause.
All measures mentioned above helps us to finally come to a conclusion to the root cause of the problem and take effective measures to reduce/eliminate the causes.
(Answer)