The project for this week was to find a visualisation on the Makeover Monday website and recreate this visualisation based on improvements we thought would enhance the original visualisation.
I decided to try improve a visualisation based on five year cancer survival rates in the US. The visualisation and data set can be found Here:
![](https://www.thedataschool.co.uk/content/images/2022/04/image-98.png)
This vizz had some positives but could definitely be improved.
Positives:
Clearly outlines the type of cancer with label's, grouped up into time periods, the arrow points in the direction of the later time period.
Negatives:
The colour key is hidden within text so its difficult to interpret, the grid lines are unnecessary due to the percentages being labelled, difficult to tell when the arrow is pointed in the opposite direction.
Recommendations:
I would recommend highlighting the cancer types where survival rate has decreased as this is not clear and obvious (for uterus and cervix uteri cancer), I would then sort the data relative to the percentage difference between the year groups. The bottom axis isn't needed due to the percentages being labelled. I would also create a clear colour key.
Result:
This was more difficult than anticipated to achieve what I wanted to achieve, I had to manually create two year groups which made it difficult to find the percentage difference but I managed to include it within the tooltip. As you can see the colours key is easier to follow, grid lines have been removed as well as the labels on the x axis as percentages are already labelled on the chart. I decided to create a barbell chart as the arrow was confusing and then sorted descending by percentile of survival rate within cancer type. When recreating the vizz I never managed to find 2007 so I used 2006 instead resulting in no decreases in survival rate for uterus and cervix uteri cancer.
![](https://www.thedataschool.co.uk/content/images/2022/04/image-95.png)
Overall week 2 has gone by quickly and I have learnt about a lot of valuable chart types and had a great first week in the office.