If you are unable to evaluate the multiple factors affecting your research or hypothesis testing, you should opt for factorial design analysis. While it is easier than one-factor-at-one-time analysis, getting factorial design right is no way simple. But once you learn the technique of statistics, you would wonder why you never learnt it before!
In school, if you were one of those students who was involved doing experiments, you would have many a times learnt that all you needed was to hold all factors constant other than the one that you were studying. It is relatively simpler in comparison to a situation where you have to study many factors at the same point of time. It would be very impractical to individually devote time, effort and resources on each factor and you wouldn't even be able to study the cross effect of one on the other. This is the place where factorial design comes into picture. It minimises the number of experiments to run to obtain meaningful results and conclusions about the way many factors affect a response in an efficient manner.
A factorial design is needed when you have different factors that need to be studied simultaneously. If you set out to create the design by hand, it can be a herculean task and at the same time have a large scope of error. Fortunately, statistical software these days have simplified the task and we can customise the factorial design as per need. These tools help to create not just detailed factorial designs but also simplify them for easier understanding.
How do you choose the right design for your experiment? Well, the right design for your experiment will depend upon the:
1.Number of factors that you are studying
2.Number of levels in each factor that need to be studied
The least number of factors and levels required are 2 each. Which means a 2X2 factorial design is the simplest and then further factors can be added in a full factorial design. The good part is that your data can be either, categorical or continuous.
If you are a beginner and want to introduce yourself to the concept of factorial design, MiniTab is a great way to start. It is user friendly and will lead you through the factorial experiment right from the beginning to the end. It is quite a simple and uncomplicated way to get your hands on to a statistical technique which might otherwise appear daunting to you. Gradually when you become familiar, there are other softwares that offer a complete factorial design option that you could explore.