How Does A Data Analysis Work Proceed?
Any work, no matter how simple it seems, is composed of individual processes, the simpler the work is, the tinier the steps become. If you take the example of a routine work, say like car washing. You may not even notice that your action starts with data collection and mining. You examine the different parts of the car and decide which all require cleaning and what the cleaning methods to be adopted are. If any of the previous cleaning methods failed to show satisfactory results, it must be analyzed for improvement. The final result you wish to achieve will be a sparkling, fresh car.
You have a trading software working on the lines of QProfit System, but wish to make it more diverse by adding one or more features other than trading. Again, you have a series of steps, all pointing to data analysis methods.
Accomplishing a project with data analysis
Now, moving to a more professional example, let us take a case study in this post where you are given the task of designing and developing a drug delivery system for the buccal cavity.
- Obtaining raw data and traversing through them: You collect all viable literature and other market products in the same genre and carry out a thorough analysis by breaking them down and examining each.
- Processing: The next step is to segregate only the relevant data from credible sources and process them to form structured data for further interpretation. Data analysis offers several tools to realize this step such as spreadsheets and databases. Unwanted, incomplete or erroneous pieces of data are eliminated in repeated proof-reading. The positives of an existing delivery vehicle or the reason for the failure of another add caution to your design.
- Exploration: Now, that your data is ready to be played upon, this iterative step adds on any extra useful information, the inclusion of more recent sources and deletes unnecessary textual or pictorial data. This is also the step where you graphically and descriptively explore the data to elucidate a more compact data structure. A new product may have entered the market the last day or a new clinical study might have got published.
- Mathematical methods: Next comes the area where the calculative and predictive skills of the analyst are tested. Mathematical and statistical models related to progression and regression applied to the data brings out the nearest accurate inferential analysis of the structured data. A graphical model, when fed with the data will give you the first look at the delivery
- Display the result: The process of analyzing data and designing the ideal delivery system in your laptop does not complete the step. You have to display it and make the audience aware of the results through proper communication that may be mail or presentation or a file.