We live in a world in which we are constantly surrounded by data - counts, figures, lengths, colors, etcetera. With such an enormous amount of data penetrating our lives, how do we make sense of it? This is where the mathematical methods of statistics come in. Statistical methods allow us to analyze, evaluate and better understand the data that is all around us. However, before we can apply statistical methods to data analysis, the data must be collected: it must be put down in figures. This article details the process by which data suitable for statistical analysis can be collected form a system.
First of all, before actually collecting the data, one must decide upon a protocol for recording it. When considering how to organize data, ask yourself such questions as "What do certain data have in common?", as well as "What is the best way of making this data accessible in the future?" For instance, you would most certainly not begin data collection using a medium such as a paper napkin to record data - such a medium would be easily lost, would be easily damaged, and would be difficult to analyze. Think ahead: if you're planning on using a computer to perform data analysis, you may wish to use a computer to collect data, as it would then be immediately accessible for analysis after the collection process completes.
Next, you must decide upon an experimental setup - how will the data be physically measured? This could be as simple as a survey, or as complex as a gamma ray detector. It is imperative that the experimental setup allow for the collection of sufficient, relevant data. As well, it is good practice to record not only quantitative, numerical data, but qualitative observations. For instance, if you were measuring the temperature of a liquid over time using a thermometer, it would be wise to record details about the thermometer. For instance, was the thermometer cracked? Who manufactured the thermometer? Was it a mercury thermometer, or one that uses alcohol?
With these questions in mind, proceed to collect the data to be analyzed. Always be observant, and, above all, record all results that are obtained. Even if a result seems off, record it anyway - otherwise the validity of the statistical analysis will be compromised. By following these simple tips, you will be sure to succeed in collecting your data by experimentation.