![](https://www.thedataschool.co.uk/content/images/2023/12/_55204682-1f51-42e4-ac8f-7180da6c48fa-1.jpg)
Last week, our coach Val told us that the most important part of working with a new dataset is understanding what one row of data represents. Experience reinforced this lesson for me while working on my project this week.
Our assignment was to create a dashboard using the World Indicators dataset that's included in Tableau. My first step was to explore the data in Tableau to determine what story I wanted to tell, so I started building some simple charts to look for patterns.
As I was building, I expected some aggregation methods to return the same results--for example, I thought that when looking at a single country, summing or averaging the population shouldn't make a difference, since each country only has one population, and the sum of one thing is the same as the average of one thing. So I just left my aggregation set to sum as I started building.
Pretty quickly, I noticed some strange numbers. Life expectancy was reaching up to 1000 years! And everything was changing as I switched between aggregation methods, contrary to what I expected.
Eventually, I realized that these discrepancies were a result of my own misunderstanding. I thought that each row represented the indicators for a given country, but I was missing a critical piece: time. Each row represented the indicators for a given country in a given year. Since we had data from 2000 thru 2012, there were 12 rows of data per country!
Once I understood what a single row of data represented, I had no more trouble and was able to perform my exploratory analysis. Next time I'll remember Val's advice from the start and take the time to look at the data table and understand what one row of data means as the first step of my analysis.