Data analysis is all about getting answers to a question. However, this nearly always results in another question and another and another. In this video Matt introduces the analytic cycle and shows that an iterative approach to data analysis and visualization can be much more effect than a typical linear process.
- [Instructor] You might think of the typical approach to data analysis often follows a linear approach. You're sometimes given a question that somebody wants an answer to. So this, you haveta kind of find the data. You perform your analysis, and you might end up doing a visualization in order to provide the best answer to that question. You feel good about it. You send the results off. And time passes. You get an email, and somebody wants to know a little bit more about it. Inevitably, answers always lead to more questions.
More questions might need more data. And so this goes on. So maybe you were asked what was our busiest month? You find out that it was last month, and you send that information off. You then get asked, okay, well, how does that compare to other months? Was this a one off or are things trending up, trending down? Inevitably, people always want to know more. And what was that due to? How can we act on it to prevent this happening in the future? Or, maybe, it was a good thing. What we soon discover is that the linear process doesn't really work with how we typically have to work with data.
Instead, it's much more of a circular approach. A question needs data so we have to go and find the data to answer that question. Once we have that data, then we have to go and do our analysis on it. This provides some insight into the data, and we can provide an actionable insight. Then more questions arise; we might need more data, and so it goes round and round and round. What Tableau Prep allows you to do is something slightly different. We can move between each of these sections.
It's not a linear process. It's not a circular process. It's more of an intricate process. As we try to answer a question, we start with our original data set. But we might need to bring in additional data sets and combine those together. We then do our analysis, and we get our insight. But that might raise another question so maybe have to get some more data. We might want to act on that information we've got already, and that might provide more questions. And so it goes on and on. What we find is we end up answering more questions, we need more data. That allow us to do deeper analysis, provide greater insight, and more actions.
Prep allows you to move easily and quickly between each of these areas, providing the information that you need.
- The data prep cycle
- Connecting to data
- Examining data in the preview pane
- Cleaning data
- Combining data using joins and unions
- Reshaping and pivoting data
- Previewing and sharing data
- Data sampling to improve performance