Ken Feldman

Would it be useful to have a simple metric for measuring how well your manufacturing process is working? That is exactly what Overall Equipment Effectiveness (OEE) is all about. Let’s learn a little more about OEE.

Despite the type of data your process generates, there are typically three descriptors or characteristics you would want to know to properly describe your process. Let’s learn what those descriptive statistics are.

It is not practical to calculate the statistical parameters of your population by measuring everything. The Central Limit Theorem will help you use samples to get to your answer.

How can you tell whether the average of your two sample data sets are statistically different and not just due to random chance? The T test is what you would want to use to answer the question. 

The concept of a prediction band is usually associated with regression analysis. We will also contrast a prediction band with a confidence band.

What kind of relationship do your X and Y variables have? Is it linear? Is it non-linear? Let’s learn more about how it can be determined graphically and statistically.

You have two sets of data and want to test whether the means of the populations they came from are the same but the data is not normally distributed. What can you do? Let’s find out.

A control chart is used to identify and alert you when your process has unexpectedly shifted. Traditional Shewhart charts are good for moderate to large shifts but not for small shifts. The CUSUM (Cumulative Sum) is better for identifying small shifts. Let’s see how.

You have compared your process data before and after an improvement. Is the difference real or just noise?  Statistical significance will help you answer that question.

Kanban is a tool used in Lean Manufacturing to visually manage the workflow through your processes. It is based on the concept of a pull system where things are only replaced as they are consumed. 

This consumer products company has a large portfolio of products. Working with key customers, the company develops and promotes brand marketing programs to help the customer improve their sales which increases the company’s revenue as well. The CEO of the company realized the percentage of successful promotions was less than he expected. At the same […]

You have retrofitted your major production machine to make it run faster. Did the retrofit really improve machine speed? A 2-Sample t-Test can be used to answer that question. 

When you collect data from your process, the values will vary. This dispersion or variation of data around the central tendency can be measured. Let’s explore the different ways to measure dispersion.  

Randomness and independence are two required characteristics for understanding and analyzing process variation. The Runs Test allows you to check for the randomness of a set of sequentially collected data. 

The word gating is often used in the context of Six Sigma and DMAIC and is usually referred to as tollgate reviews. The purpose is to evaluate your progress, and make sure you’re still heading in the right direction. 

ANOVA is a powerful statistical method for comparing several samples (3 or more) to each other to see if the populations that the means came from are statistically different.

In the context of marketing, segmentation is the process of dividing a target market into groups with similar needs and behaviors. This allows you to develop unique strategies to meet specific wants and needs. 

If your manufacturing process produces multiple products or SKUs in short runs, how many control charts will you need to monitor those different products? With short-run SPC, you might only need one. 

Does your organization use a pull or push system to allow your manufacturing function to meet your customers’ demand for your products? Let’s learn why a pull system might be a better approach for optimizing your organization.

When managing any organization, there are four voices you should be listening to; Voice of the Customer, Voice of the Process, Voice of the Business and Voice of the Employee. Let’s focus on learning more about the Voice of the Employee or VOE. 

We know that many hypothesis tests have an underlying assumption of normality. But, what if your data is not normal? Let’s see how the 1-sample sign test can help. 

If you have your three laboratory technicians measure the same samples using the same equipment and measurement process will they give you the same answers? Let’s learn more about the concept of reproducibility.

When you talk about Design of Experiments, you can discuss the 3 Big R’s; Randomization, Repetition and Replication. Let’s learn more about Replication and contrast that with the other R’s. 

The correlation coefficient (r) answers the question of how strong and in what direction is the relationship between two continuous variables. Let’s explore this statistical descriptor and what it tells you about your data.