Correlation

Correlations are a way to describe relationships between two variables. A correlation coefficient, a number usually abbreviated as r, summarizes how strong the relationship is between two variables: it can range from 0 (there’s no relationship at all between the two variables), to 1 (as one changes, the other changes by a perfectly predictable, proportionate amount). Whether […]

Read more "Correlation"

A Note on Filetypes

It’s worth mentioning, Excel is to spreadsheets as Kleenex is to tissues; a giant in the category, but not the only game in town. Different types of files are marked by different extensions — Excel files end in a .xlsx or .xls. That tells you that data is stored in columns and rows, and also that you […]

Read more "A Note on Filetypes"

The Histogram

Really getting to know your data doesn’t stop at skimming through a spreadsheet at high speed, or making a few bar graphs. It’s important to know how many times each value shows up for some variables. This is called a variable’s distribution, and the shape of the distribution becomes very important when considering different kinds of tests down […]

Read more "The Histogram"

Why Do I Need To Learn This?

Short answer: You don’t, you have enough to do. Long answer: A critical mass of funding sources are demanding impact metrics. Now, you could hire a consultant to do this for you, but it would be nice to know what you’re getting. The move to quantify impact isn’t just about bean counting… it’s about making the best decisions for your […]

Read more "Why Do I Need To Learn This?"