Hello, hello!
Apologies for the delay! Great choice of data set, @adamschroeder . There was so much to explore that it was hard for me to focus just on improving the visualization.
Improvements to the Original Visualization
The goal was to analyze the relationship between work experience and annual compensation. The heatmap in general is a great choice for showing patterns between two data categories (here: bins of numeric columns). However, deriving meaningful insights from a heatmap can sometimes be tricky. Aggregated results can complement the heatmap well.
To make insights clearer, I’ve added a simple bar chart that shows the median annual salary for different bins of work experience. This one-dimensional view makes it easier to understand statements like, “People with 3-4 years of experience, on average, earn XXX amount.” The downside is that outliers, which the heatmap reveals, are less visible. Thus, both charts together offer a more comprehensive analysis.
Additional Improvements
- Filters for Country and Developer Type: I’ve added these to ensure more accurate comparisons. Without these filters, the heatmap can be misleading due to varying salary levels across countries (e.g., USA vs. India) and roles (e.g., senior executive vs. student).
- Sequential Color Palette: I’ve chosen a more intuitive palette for the heatmap, where lighter colors represent smaller values and darker colors larger values. This is easier to understand than the yellow-purple palette, which requires checking which color corresponds to high or low values.
