From the course: Learning Data Analytics: 1 Foundations

Understanding truths

From the course: Learning Data Analytics: 1 Foundations

Understanding truths

“

- The moment I learned there are multiple truths around data in my career, I became a better analyst. To keep it simple, there are three truths, the stats truth, the data truth and the business truth. The stats truth at a minimum is the significance of your results. Not all my projects have a full data science team, but there is some work that absolutely requires that layer. Data is given to statisticians and researchers for that stats truth. Their truth is to present the statistical outcomes and the significance of the report and frankly tell you if your data is actually just a coincidence. To learn more about stats and their application to data, check out Statistics Foundations by Eddie Davila. The data truth can lead us to other truths. I once took a report to a manager of the entire organization. I was so proud of it. It was data from four different systems that had been consolidated together for the very first time and they had never seen their data like that. In those days that report took hours to run and when I delivered it to my boss, he said, this was great, but it's not accurate. Oh, I couldn't believe it. Did I just give my boss an inaccurate report? Did I tie the data wrong? But I said, it's in your systems. The data truth can be impacted by the business roles and in my story that's exactly what happened. Not all the data had been collected yet. And because of that, it was impacting the truth. Just remember, in data, timing is everything. Then there are moments where the data truth is skewed by the way the data was tied together. That was the first thing I checked. This is why joins are so important. You can never spend enough time learning about joins and practicing them. Then there are some basic truths. For example, we have a list of unique products. Let's say that there are 300 individual products and that's the truth, but that really just tells us the count not if they count because maybe not all of them have been ordered, which leads us to the business truth. Likely the entire reason you're working with data is to get to the business truth. The data truth and the stats truth inform the business truth. Companies that produce, their business truth is a measure of their production. The data truth can work with the business truth to inform and create better processes and give them the ability to see and create KPIs, which are key performance indicators. So for example, if we spend hours on the development of a product and hold that inventory and it never sells, the data and the business truth tell us it's likely not worth it to make that product. All of these truths together are meant to develop measures to improve outcomes. As an analyst, you need to be aware of all the truths and make sure you're speaking to the truth of not only the actual data, but the significance of the data and how it informs the truth of the business.

Contents