We all know that correlation is not causation. It creates situations where a credulous person might say, “I know homeopathy works because I took onion juice tincture and then my broken leg healed.” Of course this leaves out a major piece of data – that the individual concerned also had their leg set and 6 weeks bed rest.
We also know it can be an incredibly useful way to formulate strategy and analyse results – and given that SEO is a constantly evolving process, it requires adaptability and constant analysis. There are useful tools that can help – for example Pearson’s Co-Efficient tables, which enable you to calculate the likelihood of a co-relation being accurate, rather than wishful thinking. Data visualisation can also help – even the charts in Excel can help you display your data for your own benefit, or to illustrate successes to your clients.
Statistics can be an imprecise, easily manipulable science, but useful when performed correctly.
