However simple or complex the chain of events in any given situation, when looked into it usually reveals a train of causal relationships - they are seen to be linked in some way. The methods of analysis aim to establish these relationships and provide a solid background for useful generalizations based on what at first appear to be separate events. The first step in this process is to collect facts and then see if any particular patterns emerge. If they do, it then becomes possible to form theories related to the facts, and this type of empirical theory forms a useful basis for analysis and prediction. However, on its own this theory is not enough; the essential second step is to test it by collecting more facts and by checking predictions against events. These new facts may mean you have to modify the theory, bearing in mind that new facts can only either disprove or support a theory - they cannot prove it to be right.