Data Visualization

Evelyn Campbell, Ph.D.

Distilling large and complex sets of data into more easily digestible forms is an important component of data science. Through effective data visualization, we can concisely communicate the story that our data is telling. We do this by choosing appropriate visual depictions of our data to accurately represent what the data means.

The visual that we choose is dependent on the type of data. Two major data types that can be visualized graphically are numerical data and categorical data.

Numerical data is commonly visualized using histograms, scatter plots and line graphs. Categorical data can be depicted using bar graphs and pie charts. There are a vast number of other methods to visualize these data types, (e.g. box and whisker plots, heatmaps, etc.), but the aforementioned graphs can greatly distill complex datasets and are commonly used throughout multiple disciplines.