Types of experimental design
There are many types of experimental design, but they can be broadly classified into two categories: factorial and fractional. Factorial designs test all possible combinations of factors and variables, while fractional designs test only a subset of them. Factorial designs are more comprehensive and accurate, but they require more resources and time. Fractional designs are more efficient and economical, but they may miss some important interactions or effects.
One common type of factorial design is the A/B test, which compares two versions of a product or feature (A and B) on a single factor or variable. For example, you can test whether a green or a red button leads to more clicks on your website. A/B testing is simple and easy to implement, but it can only test one factor at a time and it may not account for other factors that affect the outcome.
One common type of fractional design is the Taguchi method, which uses a special matrix called an orthogonal array to select a subset of combinations that cover the most variation and interaction among the factors and variables. For example, you can test four factors (headline, layout, color, and call to action) with three levels each (high, medium, and low) using only nine experiments instead of 81. The Taguchi method is more complex and sophisticated, but it can test multiple factors at once and it can estimate the optimal levels and settings for each factor.