Ken Feldman

When dealing with data and statistics, there is no certainty. As the researcher, you get to choose the level of confidence you need to make your decisions.

You can learn a lot about your process by monitoring its change over time. There are different trend charts which will allow you to do that. Let’s look at a few. 

Your data has failed the assumption of normality needed for a parametric test of hypotheses. If you have two or more groups you want to analyze, what do you do? Maybe the Mood’s Median Test will help. 

Sometimes you are interested in analyzing data that is not measurable. Nominal data consists of unordered categories which can be counted but not measured.  

Not all your data will have a normal distribution. Sometimes, the curve is taller or shorter than a theoretical  normal distribution. Does that mean you have a problem? We can answer that question by learning a little bit more about kurtosis and the platykurtic distribution.

When you hear the phrase Mallows’ Statistic (Cp) think of multiple regression analysis. The question you want to answer is which combination of predictor variables will give you the best prediction model. Let’s learn a little more. 

What is an independent variable and why is it important to understand? And how is it different from a dependent variable? Let’s answer those questions for you.

Normality of the data is an underlying assumption for the use of many statistical tools. When normality doesn’t exist, transforming the data may be necessary. Let’s see how one company did that.

Can you always believe data which is based on people’s opinions and self-reporting? The obvious answer is no. Let’s see how one company used Attribute Agreement Analysis to determine the reasons for speaker program cancellations.

Through the use of Design of Experiments (DOE), a pharmaceutical company was able to extend the shelf life for one of its life saving liquid oncology medicines. Let’s see how this saved the company money, lowered the price of the drug and helped save lives. 

The famous poet, Alexander Pope said, “To err is human … ” Using Error Mode Effects Analysis (EMEA) will help mitigate, eliminate and prevent those human errors.

Where should you focus your improvement efforts; the vital few or the trivial many? Let’s define these two terms so you can answer the question.

Six Sigma is a philosophy of business and a set of tools for improving all your organizational processes. There are three strategies or desired outcomes of deploying Six Sigma. Let’s explore them further.

As the name implies, process capability is the ability of your process to meet your customer’s expectations. There are a number of calculated metrics which can be used to indicate your process capability.

Chi-square analysis is a powerful tool for examining the relationship between variables formatted in a table. One consumer products company used it to look at its consumers in a different way, resulting in a significant increase in sales and profitability of its product portfolio.  Traditionally, an organization’s sales and marketing department makes advertising decisions based […]

After experiencing multiple adverse medical events, one large hospital system moved from the sole use of a reactive root cause analysis approach to include a proactive Failure Modes and Effects Analysis strategy, which significantly reduced adverse medical events.  As was the strategy at the time, a large tax supported public health system relied on root […]

Stop guessing what your customers value in your product or service. Use conjoint analysis to determine what features are most important to them and have the data to support your decisions. Conjoint analysis is often described as a design of experiments (DOE) for marketing. It is a structured way to help you identify and evaluate […]

Don’t you hate it when you make a mistake? A type 1 error is a mistake you can make when erroneously interpreting your data. Let’s learn what that mistake is and how to keep from making it.

You have heard the old saying that no two grains of sand are alike. Variation is everywhere. But, some variation is predictable, some is not? Special cause variation in the non predictable type of variation. Let’s learn more about it. 

Never-ending improvement is the heart of any continuous improvement effort. PDSA, developed by Dr. W. Edwards Deming, is one of the fundamental tools used today for continuous improvement.

The interquartile range is a measure of the variation in your data. While it is a calculated number, it is often shown in a graphical format on a box plot.

When someone says your data has kurtosis, do you think this is referring to some disease? Kurtosis can be a natural state of your data. Let’s find out more about kurtosis and what it means.

Dr. W. Edwards Deming is often quoted as saying, “It is not necessary to change. Survival is not mandatory.” But should the change be radical or incremental? Kaikaku is the Japanese word for “radical change”

How hard can it be to make orange juice? As one large global company found out, no single Six Sigma tool will solve all the problems in an organization. However, the integration of a number of tools such as correlation, Measurement System Analysis (MSA), control charts, and process capability will help solve many of them.  […]