Reference document

Prepare Data
In an operational environment it is likely that data is gathered from a range of sources, and within each source individuals, processes and opinions will change, all of which affects the consistency and quality of the data supplied. Data can be considered of suitable quality if they are fit for their intended uses, whether that is short-term decision-making or longer term planning, or if they correctly represent the real-world situation, which they are intended to describe. The first step in the analysis is to figure out what data has been collected. Data requires sorting, putting into some type of order; otherwise it will be completely unmanageable. If data is disorganised it cannot possibly be reviewed to identify patterns; trends in the data will be missed and confidence in the in quality will be lost. Organising data will help with initial analysis and prepare for further analysis. A good start when organising data is to build an index of data sources and record extra information about each of them that may help to explain what they contain. This will help to retrieve the data easily later. The extra information recorded may provide some explanation of the trends and anomalies observed between data sources during exploration and analysis. It is likely that data will start all mixed up within each source and in a variety of formats that are not easy to compare or perform statistical analysis on. In order to see exactly what is in the data, it is handy to put it into a common format. Again, as data is re-formatted, it is useful for later if to mark in the data with which topic they relate to, if they seem to support or disprove a hypotheses and which questions they might help to answer..
Analytical Support Processes