#### Problem:

I had a request to add the ability to perform ANOVA analysis to Stats Helper. At first, I thought it would be difficult to get Excel to calculate dynamically, but it turned out to be easier than I thought.

I had a request to add the ability to perform ANOVA analysis to Stats Helper. At first, I thought it would be difficult to get Excel to calculate dynamically, but it turned out to be easier than I thought.

I’m currently prepping a consulting project to instruct a Green Belt level course to an group that doesn’t have the budget for expensive statistical software. While Stats Helper can do most simple analysis, it was lacking a few things: namely making control charts and calculating confidence intervals for standard deviation.

Problem:

Recently, a friend and I were looking at some proportional data at work, using a tool that I had made to compare many groups of proportions with 95% confidence intervals graphically. You know, like, for fun…

Anyway, we noticed that, even with a smaller sample size, the confidence intervals for proportions that were closer to 100% were smaller than the confidence intervals for proportions significantly below 100%. I knew vaguely that this had something to do with the whole *p(1-p)* relationship in the equation for a confidence interval for a proportion, but I didn’t have a way to visualize the relationship.

Problem:

DOE is applicable almost anywhere, but it’s not applied nearly as often as it could be. As part of a course I took nearly a year ago on Corporate Finance, I saw an opportunity to apply DOE to save a lot of time and effort in planning a company’s finances. The opportunity is rampant, from pro-forma analysis to project analysis and decisions on major purchases. Basically, a model of a situation is created (usually in something like Excel) and calculations are made to calculate Net Income, Net Present Value, etc. Most of this type of analysis is complicated by the fact that all the results are based on assumptions, and we all know what happens when you assume. It turns out that DOE can save a lot of time when doing just about any type of financial analysis. In this post, I’ll lay out how DOE can be used in the “what if” portion of a project analysis.

Version 1.0 addresses a few problems with Stats Helper. First, it turns out that there’s a real need to work with non-summarized data, both in using samples as well as in working with proportions. Second, it’s not always immediately clear *what *goes *where* during data entry. Third, the capability worksheet really needs a graphical depiction of the results. Finally, there was an actual bug (gasp) in the “Power and Sample Size” worksheet (thanks to those who pointed this out–I guess people really are using this!)

Feedback I’ve been getting is showing me that there is a need for a simple way to do basic statistics, and that, while the previous version of Stats Helper accomplishes this, the product is still *too complex*. People have requested a few more basic features, such as visualizing a distribution and determining n0rmality, as well as basic linear regression. Feedback so far is positive on this new version: I think I’ve delivered what’s being requested in version 0.8 of Stats Helper.

It seems like whenever people learn about statistical problem solving, the sample size question comes up. Invariably, the number 30 is bandied about as a sweet spot that should get the job done. Astute learners generally want to understand why 30 seems to work. Read on to find out why.

Working with proportions can sometimes require huge sample sizes to distinguish between different treatments or products, making a hypothesis test unfeasible due to cost or time constraints. To illustrate, consider a ACME Widgets, a company that has fairly low margins, but nevertheless feels pressure to innovate its widget offering, in order to bolster sales. In order to qualify a new product, Read on »

Doing statistical analysis isn’t always as simple as it should be. While products like JMP and Minitab are extremely powerful and extremely useful, they are also more complex than is often needed. JMP specifically has a hard time dealing with summarized data, while Minitab can be a maze of menus and options. Read on »

**Problem:**

Six Sigma is a very powerful tool set for solving tough problems, and can be used in many different domains. However, the term “Six Sigma” seems to carry with it a certain “mystique” or “aura,” at least in some industries, that lends itself to misunderstanding, unapproachability, and even misuse. This may be one reason why, while continuous improvement remains very important in business today, Google Trends shows that Six Sigma interest may be in decline. Read on »